[
{
"id": 1,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 4,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 6,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 9,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 13,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 15,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 20,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 23,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/9848e52e-e6f3-4190-9c42-f85cd9847e10-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 26,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/9848e52e-e6f3-4190-9c42-f85cd9847e10-5.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 27,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/9848e52e-e6f3-4190-9c42-f85cd9847e10-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 29,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/9848e52e-e6f3-4190-9c42-f85cd9847e10-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 31,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 32,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 34,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-5.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 35,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 36,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 38,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 40,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-11.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 45,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-16.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 46,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 49,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Ego4D/d939955c-6db9-4406-8b7b-085494320bc7-20.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 50,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the stove. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-left? Answer Back-left or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the stove.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the stove and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/41069025-1.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Front-left"
]
},
{
"id": 52,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the sofa. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-left? Answer Front-right or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the sofa.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the sofa and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/41069025-3.mp4",
"answer": "Back-left",
"options": [
"Front-right",
"Back-left"
]
},
{
"id": 53,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the bathtub. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the bathtub.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the bathtub and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/41159504-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 54,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/41159504-2.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 55,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the stove. Please determine the final position of the target object relative to the camera. Is it Back-right or Back-left? Answer Back-right or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the stove.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the stove and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/41159525-1.mp4",
"answer": "Back-left",
"options": [
"Back-right",
"Back-left"
]
},
{
"id": 57,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the washer. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-left? Answer Front-left or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the washer.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the washer and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42444976-1.mp4",
"answer": "Back-left",
"options": [
"Front-left",
"Back-left"
]
},
{
"id": 58,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the table. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the table.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the table and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42444976-2.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 59,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42444976-3.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 62,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42445026-3.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 63,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Back-left or Back-right? Answer Back-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42445429-1.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Back-right"
]
},
{
"id": 65,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the washer. Please determine the final position of the target object relative to the camera. Is it Back-right or Back-left? Answer Back-right or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the washer.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the washer and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42446049-2.mp4",
"answer": "Back-left",
"options": [
"Back-right",
"Back-left"
]
},
{
"id": 69,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42446167-2.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 71,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42897528-1.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 76,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-57132587708734824_2_aapi.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 77,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-57132587708734824_2_gytz.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 78,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-141184560845819621_4_b_7r.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 80,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-169115044301335945_2_gvs3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 81,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-183829460855609442_1_p-9i.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 82,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-384975055665199088_2_d7yo.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 84,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-454855130179746819_3_tp7n.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 89,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1191788760630624072_1_hhbu.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 90,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1265122081809781363_3_5-ei.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 92,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1605912288178321742_2_9eqn.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 96,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1857377326903987736_1_gomb.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 98,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1918764220984209654_1_c-7c.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 99,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1921439581405198744_1_i7b9.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 102,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2101027554826767753_2_de7h.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 105,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2206505463279484253_1_1hn1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 106,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2323851946122476774_1_umam.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 107,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2547899409721197155_1_viwj.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 109,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2555987917096562599_1_cfs3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 110,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2581599794006798586_1_hnav.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 111,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-16102220208346880_1_0sfh_9fen.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 112,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-141184560845819621_1_lvv5_lphd.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 114,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-454855130179746819_1_tfj9_gqtk.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 115,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-634378055350569306_2_rchq_46la.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 117,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-972142630887801133_1_mkir_cbbz.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 118,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1005081002024129653_1_8efr_jt38.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 120,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1306458236359471795_1_h5mt_yx4p.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 123,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1352150727715827110_1_61rn_fgba.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 125,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1357883579772440606_1_ycmd_oxz-.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 127,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1416654423768444853_1_po7e_izn1.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 128,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-1416654423768444853_1_pwnk_ho1j.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 130,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-2209007584159204953_3_rnqw_mdz7.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 134,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-2330686858362435307_1_s0iq_61sk.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 136,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 137,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 138,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 140,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-5.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 142,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 144,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 145,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 146,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-11.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 148,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-13.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 150,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 152,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-17.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 153,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-18.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 156,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164002-21.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 165,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164105-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 168,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 170,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 171,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 172,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 175,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 177,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/20260115-164133-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 181,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171713-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 182,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171713-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 185,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171713-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 186,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171713-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 188,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171713-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 189,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 191,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 193,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 194,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 195,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 197,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-9.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 198,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260115-171746-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 201,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180731-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 202,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180731-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 204,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180731-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 205,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180737-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 206,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180737-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 209,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180745-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 210,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/20260119-180745-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 213,
"dataset": "YouTube",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/YouTube/3.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 215,
"dataset": "YouTube",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/YouTube/5.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 216,
"dataset": "YouTube",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/YouTube/6.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 218,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Video/20260115-161257.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 221,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Video/2026-01-19_154034_118.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 222,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Video/2026-01-19_154108_527.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 224,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42897538-2.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 226,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/42899461-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 228,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the dining-table. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the dining-table.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the dining-table and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/44358499-1.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 229,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/44358499-2.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 230,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the chair. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the chair.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the chair and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/45260928-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 233,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the stove. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the stove.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the stove and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/45663164-1.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 234,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the round table. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the round table.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the round table and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47204552-1.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 237,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-left? Answer Front-right or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47333457-1.mp4",
"answer": "Front-right",
"options": [
"Front-right",
"Back-left"
]
},
{
"id": 239,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the microwave oven. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the microwave oven.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the microwave oven and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47334096-1.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 240,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the sink. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-left? Answer Back-left or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the sink.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the sink and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47334103-1.mp4",
"answer": "Front-left",
"options": [
"Back-left",
"Front-left"
]
},
{
"id": 241,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47334239-1.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 243,
"dataset": "ARKitScenes",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ARKitScenes/47334380-2.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 249,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 250,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 251,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-9.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 253,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 256,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 257,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 258,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-16.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 259,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 260,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-18.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 267,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-25.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 269,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-27.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 270,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-28.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 272,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-30.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 273,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/114cdc4b-4ae5-4938-8218-151e1e35b3db-31.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 275,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-10485926982439064520_1_oqin.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 277,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-11070802577416161387_1_qahx.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 279,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-12251442326766052580_1_wd-d-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 280,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-12251442326766052580_1_wd-d-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 282,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-12321865437129862911_1_1ddr.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 284,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-13355317306876751663_1_t-jf.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 286,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-14018515129165961775_1_pegw.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 287,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-14753089714893635383_1_atlj-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 298,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-16473613811052081539_1_3y7p.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 299,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-16504318334867223853_1_pzcf.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 300,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-17750787536486427868_1_xp2h.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 302,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-18397511418934954408_1_sr9l.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 303,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2598465433001774398_1_bbon.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 307,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3068522656378006650_1_0zu_.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 309,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3363533094480067586_1_hgf9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 310,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3364861183015885008_1_l12e.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 313,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3385534893506316900_1_ivgh.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 314,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3543045673995761051_1_52sk.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 315,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-4164064449185492261_1_fulc.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 320,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-7120839653809570957_1_zstq.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 321,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-7466751345307077932_1_zocu.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 323,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-7761658966964621355_1_6mli.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 324,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-8148053503558757176_1_msp0.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 329,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-9509506420470671704_1_wtxc.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 330,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-9654060644653474834_1_6mxw.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 331,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-9747453753779078631_1_kp5-.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 332,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-10153695247769592104_1_bhtd_pfcs.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 334,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-10625026498155904401_1_ke-n_3xp7.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 335,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-10770759614217273359_1_stci_ll8s.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 336,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-10923963890428322967_1_skk2_whky.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 339,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-12273083120751993429_1_lbft_ckka.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 340,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-13182548552824592684_1_ofcd_riqp.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 341,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-13310437789759009684_1_2the_hvsq.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 342,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-14133920963894906769_1_jqsm_i3hj.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 343,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-14705303724557273004_1_pnzl_gtri.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 348,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-17295069199227237940_1_zwu3_q2pf.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 351,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-2415873247906962761_1_ejcd_hwht.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 354,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-3132521568089292927_1_d7ym_rwdq.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 355,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-3195159706851203049_1_fomg_eko9.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 356,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-3338044015505973232_1_2as6_4q98.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 360,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-3644145307034257093_1_w5cm_vbva.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 362,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-4191035366928259953_1_7sgy_2f6j.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 363,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-4604173119409817302_1_mfuz_rizh.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 364,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-5076950993715916459_1_xk0p_b_wg.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 365,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-5121298817582693383_1_r0dj_bxje.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 367,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-5835049423600303130_1_iltv_xlrj.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 368,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-6177474146670383260_1_u_-s_u11p.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 370,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-6625150143263637936_1_gsei_n8bw.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 371,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-6674547510992884047_1_ilbs_vo17.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 375,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-7466751345307077932_1_zocu_btp2.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 384,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-9509506420470671704_1__vpb_mpin.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 387,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/YouTube/videoplayback_1-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 389,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/YouTube/videoplayback_1-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 390,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/YouTube/videoplayback_1-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 394,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/YouTube/videoplayback_1-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 395,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 396,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 398,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 399,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 405,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 408,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 410,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 412,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-18.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 414,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-20.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 416,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-22.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 417,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-23.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 418,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-24.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 419,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-25.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 421,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-27.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 425,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-31.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 427,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-33.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 428,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-34.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 429,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-35.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 430,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/766427682_f33-1-30064-36.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 431,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/772973786-1-30064-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 436,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/1132324722-1-30080-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 438,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Golf_Shot",
"direct_prompting": "This is a video of a golf shot. Please predict whether the golf ball will go into the hole based on its observed trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of a golf shot.\n\nYour task is to predict whether the golf ball will go into the hole based on its observed trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Initial Motion Analysis: Observe the initial direction, speed, and motion pattern of the golf ball after it is hit.\n2. Trajectory Tracking: Track the ball's motion over time, including any rolling or bouncing behavior, and analyze its path relative to the hole.\n3. Outcome Prediction: Based on the observed trajectory and its alignment with the hole, predict whether the ball will go into the hole.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Golf_Shot/Bilibili/1133347881-1-30080-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 442,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 444,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-5.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 446,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 448,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-9.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 449,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 451,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 452,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-13.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 454,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-15.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 455,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 456,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 457,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-18.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 460,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1214765667_x1-1-100144-21.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 462,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1313286479-1-100144-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 463,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1313286479-1-100144-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 464,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1313286479-1-100144-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 467,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1313286479-1-100144-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 469,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 470,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 471,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 478,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 479,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 481,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-14.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 483,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 484,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Billiards_Shot",
"direct_prompting": "This is a video of billiards. Please predict whether the target ball will go into the pocket based on its current trajectory. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and trajectory-based motion prediction. I will provide you with a video of billiards.\n\nYour task is to predict whether the target ball will go into the pocket based on its current trajectory.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the target ball and observe its current position relative to the pocket.\n2. Motion Direction Analysis: Analyze the direction and speed of the target ball based on its recent movement.\n3. Outcome Prediction: Based on the ball's trajectory and its alignment with the pocket, predict whether it will go into the pocket.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Billiards_Shot/Bilibili/1315826455-1-100144-17.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 490,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/440512299-1-30112-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 491,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/440512299-1-30112-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 492,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/440512299-1-30112-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 493,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/440512299-1-30112-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 496,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/440512299-1-30112-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 499,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 500,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 504,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 505,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-8.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 506,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 507,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 509,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 511,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 513,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-16.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 516,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-19.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 520,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/752224128-1-30112-23.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 522,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 528,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 529,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-8.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 530,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 532,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-11.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 533,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 534,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-13.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 536,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/818035305-1-30112-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 540,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 544,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 546,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 548,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 549,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 550,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 552,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-13.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 554,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 555,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/863143396-1-30112-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 556,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 557,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 562,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 564,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 566,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 567,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-12.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 569,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 570,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 571,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-16.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 572,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 573,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/894285769-1-30112-18.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 575,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 576,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 577,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 578,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-4.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 579,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 580,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/923644535-1-30112-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 583,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/1392731680-1-30112-1-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 585,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/1392731680-1-30112-1-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 588,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/Bilibili/1392731680-1-30112-1-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 594,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 595,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 597,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 600,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 601,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 602,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 604,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-1-13.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 607,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 608,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 609,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 613,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 614,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-8.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 615,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 616,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 618,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-12.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 621,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 623,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 624,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-18.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 625,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-19.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 627,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-21.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 628,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-22.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 629,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-2-23.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 633,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-3-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 635,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-3-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 638,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-3-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 642,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Soccer_Shot",
"direct_prompting": "This is a video of a soccer free kick. Please predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a soccer free kick.\n\nYour task is to predict whether the ball will go into the goal based on its trajectory, ignoring the goalkeeper and any defensive interference.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Kick Analysis: Observe the initial direction, height, and apparent force of the ball immediately after the free kick.\n2. Trajectory Tracking: Track the ball's flight path through the air and analyze its curvature and alignment relative to the goal frame.\n3. Outcome Prediction: Based only on the ball's trajectory and its alignment with the goal, predict whether it will go into the goal, ignoring the goalkeeper and any other players.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Soccer_Shot/YouTube/videoplayback-4-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 650,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-1.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 652,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-3.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 653,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-4.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 654,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-5.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 657,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-8.mp4",
"answer": "Slip",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 663,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-14.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 665,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-16.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 666,
"dataset": "Bilibili",
"dimension": "Physical_Dynamics",
"task": "Knot_Type",
"direct_prompting": "This is a video of knot manipulation. Please determine whether the knot is a slip knot (can be undone by pulling one end) or a fixed knot (cannot be undone and remains tight when pulled). Answer Slip or Fixed.",
"manual_cot_prompting": "You are an expert in physical reasoning and knot behavior analysis. I will provide you with a video of knot manipulation.\n\nYour task is to determine whether the knot is a slip knot or a fixed knot.\n\nA slip knot is a knot that can be undone or released by pulling one end of the rope.\nA fixed knot is a knot that cannot be undone by pulling one end and instead becomes tighter or remains fixed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Structure Observation: Observe the rope configuration, including loops, crossings, and how the knot is formed.\n2. Pulling Behavior Analysis: Analyze what happens when one end of the rope is pulled, especially whether the knot loosens, releases, or tightens.\n3. Knot Type Determination: Based on the observed behavior, determine whether the knot can be undone by pulling one end.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Slip or Fixed]\n",
"video": "data/Physical_Dynamics/Knot_Type/Bilibili/152337032-1-30102-17.mp4",
"answer": "Fixed",
"options": [
"Slip",
"Fixed"
]
},
{
"id": 668,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 7. Answer 3 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/59598339-1-30077-2.mp4",
"answer": "7",
"options": [
"3",
"7"
]
},
{
"id": 669,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 7. Answer 3 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/59598339-1-30077-3.mp4",
"answer": "7",
"options": [
"3",
"7"
]
},
{
"id": 670,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 7. Answer 3 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/59598339-1-30077-4.mp4",
"answer": "3",
"options": [
"3",
"7"
]
},
{
"id": 671,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 6 and lane 7. Answer 6 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 6 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 6 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n6 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/238757918-1-30077-1.mp4",
"answer": "7",
"options": [
"6",
"7"
]
},
{
"id": 672,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 4. Answer 3 or 4.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 4.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 4 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 4\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/238757918-1-30077-1.mp4",
"answer": "3",
"options": [
"3",
"4"
]
},
{
"id": 673,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 5. Answer 4 or 5.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 5.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 5 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 5\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/238757918-1-30077-1.mp4",
"answer": "5",
"options": [
"4",
"5"
]
},
{
"id": 678,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 6. Answer 4 or 6.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 6.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 6 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 6\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/406834878_x1-1-30102-1.mp4",
"answer": "6",
"options": [
"4",
"6"
]
},
{
"id": 679,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 7. Answer 4 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/406834878_x1-1-30102-1.mp4",
"answer": "4",
"options": [
"4",
"7"
]
},
{
"id": 680,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 7. Answer 4 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/406834878_x1-1-30102-2.mp4",
"answer": "4",
"options": [
"4",
"7"
]
},
{
"id": 682,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 6. Answer 4 or 6.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 6.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 6 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 6\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/406834878_x1-1-30102-4.mp4",
"answer": "6",
"options": [
"4",
"6"
]
},
{
"id": 684,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 5 and lane 6. Answer 5 or 6.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 5 and lane 6.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 5 and lane 6 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n5 or 6\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/29407381406-1-30102-1.mp4",
"answer": "6",
"options": [
"5",
"6"
]
},
{
"id": 685,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 5 and lane 6. Answer 5 or 6.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 5 and lane 6.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 5 and lane 6 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n5 or 6\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/29407381406-1-30102-2.mp4",
"answer": "5",
"options": [
"5",
"6"
]
},
{
"id": 696,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 8 and lane 9. Answer 8 or 9.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 8 and lane 9.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 8 and lane 9 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n8 or 9\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-4.mp4",
"answer": "9",
"options": [
"8",
"9"
]
},
{
"id": 697,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 6 and lane 7. Answer 6 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 6 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 6 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n6 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-4.mp4",
"answer": "7",
"options": [
"6",
"7"
]
},
{
"id": 698,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 5 and lane 7. Answer 5 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 5 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 5 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n5 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-4.mp4",
"answer": "5",
"options": [
"5",
"7"
]
},
{
"id": 703,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 8 and lane 10. Answer 8 or 10.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 8 and lane 10.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 8 and lane 10 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n8 or 10\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-7.mp4",
"answer": "8",
"options": [
"8",
"10"
]
},
{
"id": 704,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 2 and lane 3. Answer 2 or 3.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 2 and lane 3.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 2 and lane 3 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n2 or 3\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-7.mp4",
"answer": "3",
"options": [
"2",
"3"
]
},
{
"id": 705,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 8. Answer 3 or 8.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 8.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 8 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 8\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-7.mp4",
"answer": "3",
"options": [
"3",
"8"
]
},
{
"id": 707,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 9 and lane 10. Answer 9 or 10.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 9 and lane 10.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 9 and lane 10 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n9 or 10\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-9.mp4",
"answer": "9",
"options": [
"9",
"10"
]
},
{
"id": 708,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 4. Answer 3 or 4.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 4.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 4 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 4\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-9.mp4",
"answer": "3",
"options": [
"3",
"4"
]
},
{
"id": 709,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 1 and lane 2. Answer 1 or 2.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 1 and lane 2.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 1 and lane 2 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n1 or 2\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-9.mp4",
"answer": "2",
"options": [
"1",
"2"
]
},
{
"id": 710,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 8 and lane 9. Answer 8 or 9.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 8 and lane 9.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 8 and lane 9 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n8 or 9\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-10.mp4",
"answer": "9",
"options": [
"8",
"9"
]
},
{
"id": 711,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 6. Answer 4 or 6.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 6.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 6 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 6\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-10.mp4",
"answer": "6",
"options": [
"4",
"6"
]
},
{
"id": 712,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 1 and lane 2. Answer 1 or 2.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 1 and lane 2.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 1 and lane 2 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n1 or 2\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-10.mp4",
"answer": "2",
"options": [
"1",
"2"
]
},
{
"id": 717,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 4. Answer 3 or 4.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 4.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 4 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 4\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-12.mp4",
"answer": "4",
"options": [
"3",
"4"
]
},
{
"id": 718,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 5 and lane 7. Answer 5 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 5 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 5 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n5 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-12.mp4",
"answer": "5",
"options": [
"5",
"7"
]
},
{
"id": 719,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 8. Answer 4 or 8.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 8.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 8 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 8\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-12.mp4",
"answer": "4",
"options": [
"4",
"8"
]
},
{
"id": 720,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 7 and lane 8. Answer 7 or 8.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 7 and lane 8.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 7 and lane 8 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n7 or 8\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-13.mp4",
"answer": "7",
"options": [
"7",
"8"
]
},
{
"id": 721,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 7. Answer 4 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-13.mp4",
"answer": "7",
"options": [
"4",
"7"
]
},
{
"id": 722,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 2 and lane 3. Answer 2 or 3.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 2 and lane 3.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 2 and lane 3 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n2 or 3\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-13.mp4",
"answer": "2",
"options": [
"2",
"3"
]
},
{
"id": 723,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 3 and lane 9. Answer 3 or 9.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 3 and lane 9.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 3 and lane 9 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n3 or 9\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-13.mp4",
"answer": "3",
"options": [
"3",
"9"
]
},
{
"id": 734,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 2 and lane 5. Answer 2 or 5.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 2 and lane 5.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 2 and lane 5 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n2 or 5\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-19.mp4",
"answer": "2",
"options": [
"2",
"5"
]
},
{
"id": 737,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 7 and lane 9. Answer 7 or 9.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 7 and lane 9.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 7 and lane 9 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n7 or 9\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-21.mp4",
"answer": "7",
"options": [
"7",
"9"
]
},
{
"id": 738,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 2 and lane 7. Answer 2 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 2 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 2 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n2 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-21.mp4",
"answer": "7",
"options": [
"2",
"7"
]
},
{
"id": 740,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 4 and lane 7. Answer 4 or 7.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 4 and lane 7.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 4 and lane 7 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n4 or 7\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-23.mp4",
"answer": "4",
"options": [
"4",
"7"
]
},
{
"id": 741,
"dataset": "Bilibili",
"dimension": "Outcome_Prediction",
"task": "Swimming_Race",
"direct_prompting": "This is a video of a swimming race. The swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom. Please determine which swimmer will reach the finish line first between lane 2 and lane 8. Answer 2 or 8.",
"manual_cot_prompting": "You are an expert in video-based motion analysis and temporal reasoning. I will provide you with a video of a swimming race.\n\nThe swimmer at the top of the frame is in lane 1, and the lane numbers increase from top to bottom.\n\nYour task is to determine which swimmer will reach the finish line first between lane 2 and lane 8.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Lane Identification: Identify the swimmers in lane 2 and lane 8 based on their vertical positions in the frame.\n2. Motion Analysis: Observe and compare their movement over time, focusing on their speed, stroke consistency, and forward progress.\n3. Temporal Order Prediction: Based on their motion trends, determine which swimmer will reach the finish line first.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n2 or 8\n",
"video": "data/Outcome_Prediction/Swimming_Race/Bilibili/37241685933-1-30116-23.mp4",
"answer": "8",
"options": [
"2",
"8"
]
},
{
"id": 743,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 744,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 749,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-8.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 750,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 751,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 753,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 754,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-13.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 756,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 757,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 762,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-21.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 764,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-23.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 765,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-24.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 767,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-26.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 768,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-27.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 772,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-31.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 773,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-32.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 775,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-34.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 776,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-35.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 778,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-37.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 780,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-39.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 781,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-40.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 782,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-41.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 783,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-42.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 784,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-43.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 785,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-44.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 788,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/652f0590-aab5-4de1-b578-238adb54fa0e-1-47.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 791,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 792,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 793,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 795,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 796,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 798,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 799,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 800,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/775acd8e-086f-48cf-adf3-c154f0d0bd2d-1-11.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 812,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 813,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 814,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 815,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 817,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 818,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-11.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 819,
"dataset": "Ego4D",
"dimension": "Outcome_Prediction",
"task": "Basketball_Shot",
"direct_prompting": "This is a video of a basketball shot. Please predict whether the basketball will go into the hoop. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in sports video analysis and physics trajectory prediction. I will provide you with a video of a basketball shot.\n\nYour task is to predict whether the basketball will go into the hoop.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Release Analysis: Observe the initial release of the basketball, including its direction, angle, and apparent force.\n2. Trajectory Tracking: Track the motion of the basketball through the air and analyze its trajectory relative to the hoop and backboard.\n3. Outcome Prediction: Based on the observed trajectory and alignment, predict whether the basketball will go into the hoop.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Outcome_Prediction/Basketball_Shot/Ego4D/c48a70f7-44a3-44aa-ac14-baf35e696e5c-1-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 820,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-89454214745557131_2_ruh5_oqk4.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 821,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-933621182106051783_1_mawx_kjam.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 822,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-10837554759555844344_1_3vvl_2pvk.mp4",
"answer": "Green",
"options": [
"Green",
"Blue"
]
},
{
"id": 823,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Relative_Velocity",
"direct_prompting": "This is a driving video. Please determine which vehicle is moving faster based on their motion over time: the vehicle in the green box or the vehicle in the blue box. Answer Green or Blue.",
"manual_cot_prompting": "You are an expert in driving video analysis and motion comparison. I will provide you with a driving video.\n\nYour task is to determine which vehicle is moving faster: the vehicle in the green box or the vehicle in the blue box.\n\nPlease base your decision on their motion over time, not just their positions in a single frame.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Identification: Identify the vehicle in the green box and the vehicle in the blue box.\n2. Motion Analysis: Observe and compare their movement across frames, focusing on how much each vehicle moves over time.\n3. Speed Comparison: Based on their displacement and motion consistency, determine which vehicle is moving faster.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Green or Blue]\n",
"video": "data/Motion_Perception/Relative_Velocity/Waymo/segment-11048712972908676520_1_shs-_sb_t.mp4",
"answer": "Blue",
"options": [
"Green",
"Blue"
]
},
{
"id": 828,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-933621182106051783_1_amxp.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 829,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-11616035176233595745_1_amto.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 831,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-14081240615915270380_1__17g.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 832,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-14956919859981065721_1_uxig.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 833,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1943605865180232897_1_gmbk.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 834,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-1943605865180232897_1_vu9z.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 835,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-2094681306939952000_1_bwh5.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 836,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3577352947946244999_1__nlb.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 837,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-3915587593663172342_1_l6dm.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 838,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-5302885587058866068_1_33d6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 840,
"dataset": "Waymo",
"dimension": "Motion_Perception",
"task": "Vehicle_Movement",
"direct_prompting": "This is a driving video. Please determine whether the vehicle in the green box shows subtle movement during the video. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving video analysis and subtle motion detection. I will provide you with a driving video.\n\nYour task is to determine whether the vehicle in the green box shows subtle movement during the video.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Vehicle Identification: Identify the vehicle in the green box and observe its position relative to nearby lane markings, surrounding vehicles, or other static background references.\n2. Motion Comparison: Compare the target vehicle's position across frames and analyze whether there is any slight but consistent displacement over time.\n3. Motion Determination: Based on the observed frame-to-frame changes, determine whether the vehicle in the green box shows subtle movement during the video.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Motion_Perception/Vehicle_Movement/Waymo/segment-5832416115092350434_1_tgem.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 851,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the electric fan. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the electric fan.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the electric fan and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/7b6477cb95-1.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 852,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the telephone. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the telephone.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the telephone and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/7b6477cb95-2.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 855,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the door. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the door.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the door and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/acd95847c5-2.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 856,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the bookshelf. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the bookshelf.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the bookshelf and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/fb5a96b1a2-1.mp4",
"answer": "Front-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 857,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the monitor. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the monitor.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the monitor and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/a24f64f7fb-1.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 863,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the laundry machine. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-right? Answer Back-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the laundry machine.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the laundry machine and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/45b0dac5e3-2.mp4",
"answer": "Front-right",
"options": [
"Back-left",
"Front-right"
]
},
{
"id": 866,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-right? Answer Back-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/286b55a2bf-1.mp4",
"answer": "Front-right",
"options": [
"Back-left",
"Front-right"
]
},
{
"id": 867,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the electric hair drier. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the electric hair drier.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the electric hair drier and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/5942004064-1.mp4",
"answer": "Front-right",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 868,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/5942004064-2.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 871,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the heating. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the heating.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the heating and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/bcd2436daf-1.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 873,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the monitor. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the monitor.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the monitor and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/0d2ee665be-1.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 874,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the microwave oven. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the microwave oven.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the microwave oven and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/0d2ee665be-2.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 875,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-left? Answer Front-left or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/3db0a1c8f3-1.mp4",
"answer": "Back-left",
"options": [
"Front-left",
"Back-left"
]
},
{
"id": 876,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the laundry machine. Please determine the final position of the target object relative to the camera. Is it Back-left or Back-right? Answer Back-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the laundry machine.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the laundry machine and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/3db0a1c8f3-2.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Back-right"
]
},
{
"id": 878,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the suitcase. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the suitcase.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the suitcase and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/09c1414f1b-1.mp4",
"answer": "Front-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 880,
"dataset": "ScanNet++",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet++/6115eddb86-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 884,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the laundry machine. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the laundry machine.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the laundry machine and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0307_02-1.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 885,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the stove. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the stove.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the stove and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0328_00-1.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 887,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the chair. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the chair.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the chair and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0356_00-1.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 892,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the stove. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-left? Answer Front-right or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the stove.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the stove and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0651_02-1.mp4",
"answer": "Back-left",
"options": [
"Front-right",
"Back-left"
]
},
{
"id": 893,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the chair. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the chair.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the chair and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0658_00-1.mp4",
"answer": "Back-right",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 894,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0697_01-1.mp4",
"answer": "Front-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 899,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0550_00-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 900,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the sofa. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-left? Answer Back-left or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the sofa.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the sofa and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0568_02-1.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Front-left"
]
},
{
"id": 901,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the bookshelf. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the bookshelf.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the bookshelf and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0568_02-2.mp4",
"answer": "Front-left",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 902,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the door. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the door.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the door and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0575_01-1.mp4",
"answer": "Front-left",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 904,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the table lamp. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-left? Answer Back-left or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the table lamp.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the table lamp and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0580_01-1.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Front-left"
]
},
{
"id": 906,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the sofa. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-right? Answer Back-right or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the sofa.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the sofa and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0591_01-1.mp4",
"answer": "Front-right",
"options": [
"Back-right",
"Front-right"
]
},
{
"id": 908,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the backpack. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-left? Answer Front-left or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the backpack.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the backpack and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0593_00-2.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Back-left"
]
},
{
"id": 909,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the tv. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the tv.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the tv and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0608_00-1.mp4",
"answer": "Back-right",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 910,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the world map. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-left? Answer Front-left or Back-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the world map.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the world map and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0608_00-2.mp4",
"answer": "Back-left",
"options": [
"Front-left",
"Back-left"
]
},
{
"id": 911,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the guitar. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the guitar.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the guitar and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0608_00-3.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 917,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the wall clock. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the wall clock.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the wall clock and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0050_01-2.mp4",
"answer": "Front-right",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 918,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0146_01-1.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 919,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the backpack. Please determine the final position of the target object relative to the camera. Is it Back-left or Back-right? Answer Back-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the backpack.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the backpack and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0146_01-2.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Back-right"
]
},
{
"id": 920,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Back-left or Front-right? Answer Back-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0149_00-1.mp4",
"answer": "Back-left",
"options": [
"Back-left",
"Front-right"
]
},
{
"id": 923,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the refrigerator. Please determine the final position of the target object relative to the camera. Is it Front-right or Back-right? Answer Front-right or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the refrigerator.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the refrigerator and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0164_02-2.mp4",
"answer": "Back-right",
"options": [
"Front-right",
"Back-right"
]
},
{
"id": 926,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Front-left or Back-right? Answer Front-left or Back-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Back-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Back-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Back-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0207_02-3.mp4",
"answer": "Front-left",
"options": [
"Front-left",
"Back-right"
]
},
{
"id": 927,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the backpack. Please determine the final position of the target object relative to the camera. Is it Front-right or Front-left? Answer Front-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the backpack.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the backpack and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0222_01-1.mp4",
"answer": "Front-left",
"options": [
"Front-right",
"Front-left"
]
},
{
"id": 931,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the ceiling lamp. Please determine the final position of the target object relative to the camera. Is it Front-left or Front-right? Answer Front-left or Front-right.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the ceiling lamp.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Front-left and Front-right.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the ceiling lamp and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Front-left or Front-right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Front-left or Front-right]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0246_00-1.mp4",
"answer": "Front-right",
"options": [
"Front-left",
"Front-right"
]
},
{
"id": 932,
"dataset": "ScanNet",
"dimension": "Spatial_Relations",
"task": "Ego_Motion",
"direct_prompting": "This is a video recorded by a moving camera in an indoor scene. The target object is the toilet. Please determine the final position of the target object relative to the camera. Is it Back-right or Front-left? Answer Back-right or Front-left.",
"manual_cot_prompting": "You are an expert in video-based spatial reasoning. I will provide you with a video recorded by a moving camera in an indoor scene.\n\nThe target object is the toilet.\n\nYour task is to determine the final position of the target object relative to the camera.\n\nThe two candidate answers are: Back-right and Front-left.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Object Identification: Identify the toilet and observe its position in the scene.\n2. Relative Motion Analysis: Track how the target object's position changes relative to the moving camera throughout the video.\n3. Final Position Determination: Based on the relative motion, determine whether the final position is Back-right or Front-left.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Back-right or Front-left]\n",
"video": "data/Spatial_Relations/Ego_Motion/ScanNet/scene0246_00-2.mp4",
"answer": "Front-left",
"options": [
"Back-right",
"Front-left"
]
},
{
"id": 960,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 962,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 963,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 967,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 969,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 970,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-11.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 971,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-12.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 973,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 976,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-17.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 979,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-20.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 980,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163455-21.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 982,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 985,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 989,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 992,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-12.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 995,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 996,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419163247-16.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1002,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1003,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1004,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-7.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1006,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1007,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-10.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1010,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Mikado_Dependency",
"direct_prompting": "This is a video of a Mikado game. Please predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical interaction analysis and video-based contact reasoning. I will provide you with a video of a Mikado game.\n\nYour task is to predict whether the stick indicated by the pointing stick can be picked up without touching any other sticks.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Stick Identification: Identify the pointing stick and determine which stick it is indicating as the target stick.\n2. Contact Analysis: Observe whether the target stick is touching, supporting, or entangled with any other sticks.\n3. Pickup Feasibility Prediction: Based on the contact relationships and arrangement, predict whether the target stick can be picked up without touching any other sticks.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Mikado_Dependency/Video/VID20260419162948-13.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1011,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1012,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1013,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-3.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1020,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1022,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-12.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1023,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID_20260419_165720-13.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1024,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1025,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1029,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-6.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1030,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-7.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1031,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-8.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1036,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-13.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1039,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-16.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1040,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-17.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1041,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164244-18.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1050,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419170751-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1053,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419170751-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1054,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419170751-5.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1058,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-4.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1060,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-6.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1062,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-8.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1063,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-9.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1064,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-10.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1067,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-13.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1068,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-14.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1069,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-15.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1071,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-17.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1072,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-18.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1073,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-19.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1074,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-20.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1076,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-22.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1077,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-23.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1078,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-24.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1079,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-25.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1080,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-26.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1081,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-27.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1082,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-28.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1083,
"dataset": "Video",
"dimension": "Physical_Dynamics",
"task": "Jenga_Stability",
"direct_prompting": "This is a video of a Jenga game. Please predict whether the tower will remain stable after the block that the hand is trying to pull out is removed. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in physical reasoning and video-based stability analysis. I will provide you with a video of a Jenga game.\n\nYour task is to predict whether the tower will remain stable after the block that the hand is trying to pull out is removed.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Block Identification: Identify which block the hand is trying to pull out, and determine its position within the tower.\n2. Structural Support Analysis: Analyze how this block contributes to the support and balance of the surrounding blocks and upper layers.\n3. Stability Prediction: Based on the tower structure and the role of the target block, predict whether the tower will remain stable after this block is removed.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Yes or No]\n",
"video": "data/Physical_Dynamics/Jenga_Stability/Video/VID20260419164948-29.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1084,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_LEFT_2026_04_30_16_50_18_320-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1085,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_LEFT_2026_04_30_16_50_18_320-2.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1086,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_LEFT_2026_04_30_16_53_17_880-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1087,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_LEFT_2026_04_30_16_53_17_880-2.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1090,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_RIGHT_2026_04_30_16_47_18_810-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1092,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/DVR_RIGHT_2026_04_30_16_47_18_810-3.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1094,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5362-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1095,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5364-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1096,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5365-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1097,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5366-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1099,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5369-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1107,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5377-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1109,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5379-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1112,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5384-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1113,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5386-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1114,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5387-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1115,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/IMG_5389-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1118,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430161739-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1121,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430162224-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1124,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430162658-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1125,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430162741-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1126,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430163012-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1128,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430163641-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1129,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430164456-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1131,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430164708-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1133,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430164807-1.mp4",
"answer": "Yes",
"options": [
"Yes",
"No"
]
},
{
"id": 1134,
"dataset": "Video",
"dimension": "Spatial_Relations",
"task": "Passage_Feasibility",
"direct_prompting": "This is a video of a vehicle and traffic cones placed near its driving path. Please predict whether the vehicle can pass the cone-constrained area without touching any cone. Answer Yes or No.",
"manual_cot_prompting": "You are an expert in driving scene analysis and vehicle-obstacle clearance reasoning. I will provide you with a video showing a vehicle moving near traffic cones.\n\nYour task is to predict whether the vehicle can complete the current driving maneuver without touching any cone.\n\nPlease observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Cone Layout Analysis: Observe the positions of the traffic cones relative to the vehicle's driving path, including whether they form a narrow passage or are placed on only one side of the vehicle.\n2. Vehicle Motion Analysis: Analyze the vehicle's current position, orientation, moving direction, and steering tendency relative to the nearby cones.\n3. Clearance Prediction: Based on the cone layout and vehicle motion, predict whether the vehicle can complete the maneuver without touching any cone.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\nYes or No\n",
"video": "data/Spatial_Relations/Passage_Feasibility/Video/VID20260430164949-1.mp4",
"answer": "No",
"options": [
"Yes",
"No"
]
},
{
"id": 1139,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/Ways_to_Jump_Sit_Fall_3608_clip1_00_00_00_000_00_00_02_286.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1143,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/Ways_to_Jump_Sit_Fall_Heel_Click_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1144,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/Ways_to_Jump_Sit_Fall_karate_front_fall_clip5_00_00_00_000_00_00_02_214.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1146,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backflip_4_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1147,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backflip_5_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1148,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backflip_8_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1150,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backward_roll_4_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1152,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backward_roll_6_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1154,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backward_roll_11_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1155,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/backward_roll_19_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1159,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/balancebeam_rotate_10_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1160,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/balancebeam_rotate_11_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1161,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/balancebeam_rotate_13_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1163,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/balancebeam_rotate_16_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1165,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/balancebeam_rotate_19_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1170,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the person's own perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the person's own perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the person's own perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/bandaging_4_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1172,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/bandaging_12_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1173,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/bandaging_14_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1175,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/cartwheeling_3_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1177,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/cartwheeling_10_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1178,
"dataset": "MotionX",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/MotionX/cartwheeling_12_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1180,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/body_dance_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1182,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/dance_1_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1185,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/dance_7_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1188,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/dance_14_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1193,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/dance_15_clip1_Trim3.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1194,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Peking_opera_performance_1_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1195,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/peking_opera_performance_man_clip2_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1199,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_1_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1200,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_1_clip1_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1202,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_5_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1204,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_10_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1205,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_11_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1206,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_15_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1207,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_15_clip1_Trim2.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1208,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Aerial_Kick_Kungfu_wushu_22_clip2.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1210,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Back_Flip_Kungfu_wushu_Trim3_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1214,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Back_Flip_Kungfu_wushu_Trim10_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1217,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Dragon_Sword_Shaolin_Kung_Fu_Wushu_Da_Mo_Sword_1_clip1_Trim2.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1218,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Dragon_Sword_Shaolin_Kung_Fu_Wushu_Da_Mo_Sword_2_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1220,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Dragon_Sword_Shaolin_Kung_Fu_Wushu_Da_Mo_Sword_5_clip1_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1221,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Dragon_Sword_Shaolin_Kung_Fu_Wushu_Da_Mo_Sword_6_clip1_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1222,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Dragon_Sword_Shaolin_Kung_Fu_Wushu_Training_1_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1225,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Flying_Kick_kung_fu_wushu_1_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1227,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Flying_Kick_kung_fu_wushu_1_clip1_clip2_Trim2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1228,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Kung_Fu_Nunchucks_1_clip1_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1232,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Kung_Fu_Wushu_Chain_Whip_2_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1234,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Basic_Tiger_Sword_Training_1_clip2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1239,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_2_clip2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1240,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_3_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1241,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_3_clip1_Trim2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1243,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_6_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1244,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_7_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1245,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_8_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1246,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From the viewer's perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From the viewer's perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From the viewer's perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Cartwheel_Training_9_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1248,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Kicks_Flying_Kicks_1_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1249,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Kicks_Flying_Kicks_1_clip3_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1253,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_1_clip3_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1255,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_1_clip4_Trim2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1256,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_5_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1257,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_6_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1260,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_10_clip1.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1263,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_11_clip2.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1264,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Lotus_Kick_12_clip1_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1266,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Selfdefense_Sword_Form_Session_1_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1267,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Selfdefense_Sword_Form_Session_1_clip2.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1268,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Selfdefense_Sword_Form_Session_19_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1269,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Selfdefense_Sword_Form_Session_27_clip2_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1271,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Selfdefense_Sword_Form_Session_30_clip2_Trim.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1277,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Tornado_Kick_Training_2_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1278,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Tornado_Kick_Training_4_clip1_clip1.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1281,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Tornado_Kick_Training_13_clip3.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1283,
"dataset": "YouTube",
"dimension": "Motion_Perception",
"task": "Rotation_Direction",
"direct_prompting": "This is a video of a person performing a rotational action. From a top-down perspective, please determine whether the motion is clockwise or counterclockwise. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and rotational motion analysis. I will provide you with a video of a person performing a rotational action. From a top-down perspective, your task is to determine whether the observed motion is clockwise or counterclockwise. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Motion Identification: Identify the person and locate the main rotational action to be judged in the video.\n2. Perspective-Based Motion Tracking: From a top-down perspective, track how the person's body or relevant moving part changes orientation across frames.\n3. Rotation Direction Determination: Based on the temporal change in orientation, determine whether the motion is clockwise or counterclockwise from the specified perspective.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Motion_Perception/Rotation_Direction/YouTube/Shaolin_Kung_Fu_Wushu_Tornado_Kick_Training_16_clip1_Trim.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1286,
"dataset": "MotionX",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Get up or Lie down.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Get up and Lie down. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Get up or Lie down.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Get up or Lie down]\n",
"video": "data/Outcome_Prediction/Fall_Direction/MotionX/Ways_to_Go_to_Bed_Sleep_Wake_Up_Big_Stretch_clip1_00_00_00_000_00_00_03_088.mp4",
"answer": "Lie down",
"options": [
"Get up",
"Lie down"
]
},
{
"id": 1288,
"dataset": "MotionX",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left-hand side or Right-hand side.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left-hand side and Right-hand side. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left-hand side or Right-hand side.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left-hand side or Right-hand side]\n",
"video": "data/Outcome_Prediction/Fall_Direction/MotionX/Ways_to_Jump_Sit_Fall_Basketball_Fake_Out_clip1_00_00_00_000_00_00_03_013.mp4",
"answer": "Right-hand side",
"options": [
"Left-hand side",
"Right-hand side"
]
},
{
"id": 1289,
"dataset": "MotionX",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left-hand side or Right-hand side.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left-hand side and Right-hand side. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left-hand side or Right-hand side.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left-hand side or Right-hand side]\n",
"video": "data/Outcome_Prediction/Fall_Direction/MotionX/Ways_to_Jump_Sit_Fall_Duel_clip1_00_00_00_000_00_00_02_490.mp4",
"answer": "Left-hand side",
"options": [
"Left-hand side",
"Right-hand side"
]
},
{
"id": 1290,
"dataset": "MotionX",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Forward or Backward.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Forward and Backward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Forward or Backward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Forward or Backward]\n",
"video": "data/Outcome_Prediction/Fall_Direction/MotionX/bike_fall_12_clip1_00_00_00_000_00_00_00_595.mp4",
"answer": "Forward",
"options": [
"Forward",
"Backward"
]
},
{
"id": 1297,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_03_52_Trim_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1298,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_05_43_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1299,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_06_26_Trim_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1300,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Forward or Backward.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Forward and Backward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Forward or Backward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Forward or Backward]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_08_15_Trim.mp4",
"answer": "Backward",
"options": [
"Forward",
"Backward"
]
},
{
"id": 1301,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_09_25_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1302,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_10_26_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1303,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_11_28_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1304,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_12_17.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1305,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_12_57_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1307,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_14_04_Trim_Trim.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1308,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_15_20_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1309,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_16_04.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1311,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_18_00.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1312,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_19_44_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1318,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_25_21.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1321,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_27_45_Trim_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1324,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_32_14_Trim_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1325,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_33_32_Trim.mp4",
"answer": "Right",
"options": [
"Left",
"Right"
]
},
{
"id": 1329,
"dataset": "YouTube",
"dimension": "Outcome_Prediction",
"task": "Fall_Direction",
"direct_prompting": "This is a video of a person falling. Please determine the person's fall direction. Answer Left or Right.",
"manual_cot_prompting": "You are an expert in video-based human motion analysis and fall direction reasoning. I will provide you with a video of a person falling. Your task is to determine the person's fall direction. The two candidate answers are: Left and Right. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Person Identification: Identify the person who is falling and observe their initial body posture and position.\n2. Falling Motion Analysis: Track how the person's body moves over time, focusing on the direction of displacement and body orientation during the fall.\n3. Fall Direction Determination: Based on the person's temporal motion and final body movement, determine whether the fall direction is Left or Right.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Left or Right]\n",
"video": "data/Outcome_Prediction/Fall_Direction/YouTube/2026_06_24_18_36_28.mp4",
"answer": "Left",
"options": [
"Left",
"Right"
]
},
{
"id": 1330,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mat. Please determine the person's movement direction relative to a mat. Answer Approaching or Moving away.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mat. Your task is to determine the person's movement direction relative to a mat. The two candidate answers are: Approaching and Moving away. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mat, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mat across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mat is Approaching or Moving away.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Approaching or Moving away]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/Ways_to_Catch_Dive_clip1_camera_motion.mp4",
"answer": "Approaching",
"options": [
"Approaching",
"Moving away"
]
},
{
"id": 1331,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a wall. Please determine the person's movement direction relative to a wall. Answer Downward or Upward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a wall. Your task is to determine the person's movement direction relative to a wall. The two candidate answers are: Downward and Upward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a wall, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a wall across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a wall is Downward or Upward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Downward or Upward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/abseiling_2_clip1_camera_motion.mp4",
"answer": "Downward",
"options": [
"Downward",
"Upward"
]
},
{
"id": 1332,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a wall. Please determine the person's movement direction relative to a wall. Answer Downward or Upward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a wall. Your task is to determine the person's movement direction relative to a wall. The two candidate answers are: Downward and Upward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a wall, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a wall across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a wall is Downward or Upward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Downward or Upward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/abseiling_3_clip1_camera_motion.mp4",
"answer": "Downward",
"options": [
"Downward",
"Upward"
]
},
{
"id": 1333,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a wall. Please determine the person's movement direction relative to a wall. Answer Downward or Upward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a wall. Your task is to determine the person's movement direction relative to a wall. The two candidate answers are: Downward and Upward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a wall, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a wall across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a wall is Downward or Upward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Downward or Upward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/abseiling_5_clip1_camera_motion.mp4",
"answer": "Downward",
"options": [
"Downward",
"Upward"
]
},
{
"id": 1334,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a wall. Please determine the person's movement direction relative to a wall. Answer Downward or Upward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a wall. Your task is to determine the person's movement direction relative to a wall. The two candidate answers are: Downward and Upward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a wall, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a wall across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a wall is Downward or Upward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Downward or Upward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/abseiling_18_clip1_camera_motion.mp4",
"answer": "Downward",
"options": [
"Downward",
"Upward"
]
},
{
"id": 1335,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a calf. Please determine the person's movement direction relative to a calf. Answer Moving away or Approaching.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a calf. Your task is to determine the person's movement direction relative to a calf. The two candidate answers are: Moving away and Approaching. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a calf, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a calf across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a calf is Moving away or Approaching.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Moving away or Approaching]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/calfrope_catch_15_clip1_camera_motion.mp4",
"answer": "Moving away",
"options": [
"Moving away",
"Approaching"
]
},
{
"id": 1336,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_1_clip1_camera_motion.mp4",
"answer": "Horizontal",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1337,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_2_clip1_camera_motion.mp4",
"answer": "Vertical",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1338,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_4_clip1_camera_motion.mp4",
"answer": "Vertical",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1339,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_4_clip2_camera_motion.mp4",
"answer": "Vertical",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1341,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_9_clip1_camera_motion.mp4",
"answer": "Horizontal",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1342,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Horizontal or Vertical.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Horizontal and Vertical. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Horizontal or Vertical.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Horizontal or Vertical]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_10_clip1_camera_motion.mp4",
"answer": "Vertical",
"options": [
"Horizontal",
"Vertical"
]
},
{
"id": 1344,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Clockwise or Counterclockwise.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_13_clip1_camera_motion.mp4",
"answer": "Counterclockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1345,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a mirror. Please determine the person's movement direction relative to a mirror. Answer Clockwise or Counterclockwise.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a mirror. Your task is to determine the person's movement direction relative to a mirror. The two candidate answers are: Clockwise and Counterclockwise. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a mirror, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a mirror across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a mirror is Clockwise or Counterclockwise.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Clockwise or Counterclockwise]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/cleaning_mirror_15_clip1_camera_motion.mp4",
"answer": "Clockwise",
"options": [
"Clockwise",
"Counterclockwise"
]
},
{
"id": 1350,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climb_pole_6_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1352,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climb_pole_9_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1355,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climb_pole_12_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1358,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climb_pole_18_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1359,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climb_pole_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1360,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rock_1_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1361,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rock_3_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1365,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and the man's body. Please determine the person's movement direction relative to the man's body. Answer on the man's Left-hand side or on the man's Right-hand side.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and the man's body. Your task is to determine the person's movement direction relative to the man's body. The two candidate answers are: on the man's Left-hand side and on the man's Right-hand side. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and the man's body, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to the man's body across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to the man's body is on the man's Left-hand side or on the man's Right-hand side.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[on the man's Left-hand side or on the man's Right-hand side]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rock_13_clip1_camera_motion.mp4",
"answer": "on the man's Left-hand side",
"options": [
"on the man's Left-hand side",
"on the man's Right-hand side"
]
},
{
"id": 1368,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a climbing surface. Please determine the person's movement direction relative to a climbing surface. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a climbing surface. Your task is to determine the person's movement direction relative to a climbing surface. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a climbing surface, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a climbing surface across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a climbing surface is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rock_19_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1370,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a rope. Please determine the person's movement direction relative to a rope. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a rope. Your task is to determine the person's movement direction relative to a rope. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a rope, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a rope across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a rope is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rope_2_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1372,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a rope. Please determine the person's movement direction relative to a rope. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a rope. Your task is to determine the person's movement direction relative to a rope. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a rope, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a rope across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a rope is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rope_4_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1377,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a rope. Please determine the person's movement direction relative to a rope. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a rope. Your task is to determine the person's movement direction relative to a rope. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a rope, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a rope across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a rope is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_rope_18_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1381,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a tree. Please determine the person's movement direction relative to a tree. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a tree. Your task is to determine the person's movement direction relative to a tree. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a tree, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a tree across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a tree is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_tree_10_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
},
{
"id": 1383,
"dataset": "MotionX",
"dimension": "Spatial_Relations",
"task": "Interaction_Direction",
"direct_prompting": "This is a video of an interaction between a person and a tree. Please determine the person's movement direction relative to a tree. Answer Upward or Downward.",
"manual_cot_prompting": "You are an expert in video-based spatial-temporal reasoning and person-object interaction analysis. I will provide you with a video of an interaction between a person and a tree. Your task is to determine the person's movement direction relative to a tree. The two candidate answers are: Upward and Downward. Please observe the video carefully, strictly follow the analysis steps below, and output your final answer in a fixed format.\n\n[Analysis Steps]\n1. Target Interaction Identification: Identify the person and a tree, and observe their initial spatial relationship in the video.\n2. Relative Motion Analysis: Track how the person's position or motion changes relative to a tree across frames, ignoring camera-induced motion when necessary.\n3. Movement Direction Determination: Based on the relative displacement over time, determine whether the person's movement direction relative to a tree is Upward or Downward.\n\n[Output Format Requirements]\nPlease strictly adhere to the following XML tag format for your response. Do not add any extra text outside the tags.\n\n\nWrite your detailed reasoning process based on the three steps above here.\n\n\n[Upward or Downward]\n",
"video": "data/Spatial_Relations/Interaction_Direction/MotionX/climbing_tree_18_clip1_camera_motion.mp4",
"answer": "Upward",
"options": [
"Upward",
"Downward"
]
}
]