diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bSqkOH3Y80_00013_0_95.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bSqkOH3Y80_00013_0_95.json new file mode 100644 index 0000000000000000000000000000000000000000..48a82d06b5fbcdbf65c93b10c868af45feddf419 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bSqkOH3Y80_00013_0_95.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_-bSqkOH3Y80_00013_0_95", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/-bSqkOH3Y80_00013_0_95_fore", + "duration": 4.0 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/-bSqkOH3Y80_00013_0_95_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/-bSqkOH3Y80_00013_0_95_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bgPTgJJlv4_00019_8_72.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bgPTgJJlv4_00019_8_72.json new file mode 100644 index 0000000000000000000000000000000000000000..405b433a353fa13ed49326298d2da51bb8157f24 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-bgPTgJJlv4_00019_8_72.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_-bgPTgJJlv4_00019_8_72", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/-bgPTgJJlv4_00019_8_72_fore", + "duration": 2.169 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/-bgPTgJJlv4_00019_8_72_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/-bgPTgJJlv4_00019_8_72_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-gxw2PyiUJA_00045_573_723.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-gxw2PyiUJA_00045_573_723.json new file mode 100644 index 0000000000000000000000000000000000000000..345a22271fd4d95bcb9fb54009b188a24e1ba106 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-gxw2PyiUJA_00045_573_723.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_-gxw2PyiUJA_00045_573_723", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/-gxw2PyiUJA_00045_573_723_fore", + "duration": 5.072 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/-gxw2PyiUJA_00045_573_723_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/-gxw2PyiUJA_00045_573_723_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-ne-ss1e51c_00005_247_387.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-ne-ss1e51c_00005_247_387.json new file mode 100644 index 0000000000000000000000000000000000000000..4cff1c438619a3bf57569abfd7cfe04a047dcdbd --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_-ne-ss1e51c_00005_247_387.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_-ne-ss1e51c_00005_247_387", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/-ne-ss1e51c_00005_247_387_fore", + "duration": 4.739 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/-ne-ss1e51c_00005_247_387_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/-ne-ss1e51c_00005_247_387_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0-NOMvTGx3Q_00034_0_80.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0-NOMvTGx3Q_00034_0_80.json new file mode 100644 index 0000000000000000000000000000000000000000..28da5458abbde8e807a95f85399c15523fb83566 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0-NOMvTGx3Q_00034_0_80.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0-NOMvTGx3Q_00034_0_80", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0-NOMvTGx3Q_00034_0_80_fore", + "duration": 2.7 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0-NOMvTGx3Q_00034_0_80_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0-NOMvTGx3Q_00034_0_80_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_003q7lpJFGI_00003_934_1115.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_003q7lpJFGI_00003_934_1115.json new file mode 100644 index 0000000000000000000000000000000000000000..0015aa6e11096ae997dd10edbe2d5119df4f7155 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_003q7lpJFGI_00003_934_1115.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_003q7lpJFGI_00003_934_1115", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/003q7lpJFGI_00003_934_1115_fore", + "duration": 3.104 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/003q7lpJFGI_00003_934_1115_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/003q7lpJFGI_00003_934_1115_fore", + "rendered_video": "none", + "time_window": [ + 0.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0CyBZEykTu4_00001_59_258.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0CyBZEykTu4_00001_59_258.json new file mode 100644 index 0000000000000000000000000000000000000000..c46c70f8402dc95c0112b4a4f43d1eea96804e9e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0CyBZEykTu4_00001_59_258.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0CyBZEykTu4_00001_59_258", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0CyBZEykTu4_00001_59_258_fore", + "duration": 6.674 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0CyBZEykTu4_00001_59_258_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0CyBZEykTu4_00001_59_258_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 6.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0FIuy2pklxQ_00025_0_105.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0FIuy2pklxQ_00025_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..8944cefd8de9c8201d3f825354b32574f98cf2e7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0FIuy2pklxQ_00025_0_105.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0FIuy2pklxQ_00025_0_105", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0FIuy2pklxQ_00025_0_105_fore", + "duration": 4.375 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0FIuy2pklxQ_00025_0_105_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0FIuy2pklxQ_00025_0_105_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0Oxd5OiY258_00079_69_183.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0Oxd5OiY258_00079_69_183.json new file mode 100644 index 0000000000000000000000000000000000000000..73115fb1ee658b2b0b8fba4040314dcb7664c582 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0Oxd5OiY258_00079_69_183.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0Oxd5OiY258_00079_69_183", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0Oxd5OiY258_00079_69_183_fore", + "duration": 2.3 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0Oxd5OiY258_00079_69_183_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0Oxd5OiY258_00079_69_183_fore", + "rendered_video": "none", + "time_window": [ + 0.3, + 2.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0TVSPT_n_xk_00044_23_98.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0TVSPT_n_xk_00044_23_98.json new file mode 100644 index 0000000000000000000000000000000000000000..a668db98297a04695d5776a74fff015c55fe60b3 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0TVSPT_n_xk_00044_23_98.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0TVSPT_n_xk_00044_23_98", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0TVSPT_n_xk_00044_23_98_fore", + "duration": 3.13 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0TVSPT_n_xk_00044_23_98_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0TVSPT_n_xk_00044_23_98_fore", + "rendered_video": "none", + "time_window": [ + 2.2, + 3.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0XeuCK3PfqA_00024_87_316.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0XeuCK3PfqA_00024_87_316.json new file mode 100644 index 0000000000000000000000000000000000000000..dfcb3c2392ef40700d956d054cd87e0739883d43 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0XeuCK3PfqA_00024_87_316.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0XeuCK3PfqA_00024_87_316", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0XeuCK3PfqA_00024_87_316_fore", + "duration": 7.675 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0XeuCK3PfqA_00024_87_316_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0XeuCK3PfqA_00024_87_316_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 7.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0h7AlrKLslw_00056_0_313.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0h7AlrKLslw_00056_0_313.json new file mode 100644 index 0000000000000000000000000000000000000000..1437ddca1911321a261b5f4e8f611034cd956b6f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_0h7AlrKLslw_00056_0_313.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_0h7AlrKLslw_00056_0_313", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/0h7AlrKLslw_00056_0_313_fore", + "duration": 5.267 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/0h7AlrKLslw_00056_0_313_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/0h7AlrKLslw_00056_0_313_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CL4WdaOqcz4_00000_1019_1180.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CL4WdaOqcz4_00000_1019_1180.json new file mode 100644 index 0000000000000000000000000000000000000000..ca815fdd5bd767967685a859fa3a8edaa6f7d630 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CL4WdaOqcz4_00000_1019_1180.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_CL4WdaOqcz4_00000_1019_1180", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/CL4WdaOqcz4_00000_1019_1180_fore", + "duration": 5.539 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/CL4WdaOqcz4_00000_1019_1180_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/CL4WdaOqcz4_00000_1019_1180_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CivzEXLprZw_00005_835_2081.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CivzEXLprZw_00005_835_2081.json new file mode 100644 index 0000000000000000000000000000000000000000..ff5db579814eca30b409af6c6b26bee673d729ce --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CivzEXLprZw_00005_835_2081.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_CivzEXLprZw_00005_835_2081", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/CivzEXLprZw_00005_835_2081_fore", + "duration": 20.784 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/CivzEXLprZw_00005_835_2081_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/CivzEXLprZw_00005_835_2081_fore", + "rendered_video": "none", + "time_window": [ + 9.5, + 20.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Cn2s3-690wI_00004_105_199.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Cn2s3-690wI_00004_105_199.json new file mode 100644 index 0000000000000000000000000000000000000000..aff981f74e40d11df8d33ce39cddf5f1a5d714c0 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Cn2s3-690wI_00004_105_199.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_Cn2s3-690wI_00004_105_199", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/Cn2s3-690wI_00004_105_199_fore", + "duration": 3.17 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/Cn2s3-690wI_00004_105_199_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/Cn2s3-690wI_00004_105_199_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwVGkPsOM9g_00024_34_140.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwVGkPsOM9g_00024_34_140.json new file mode 100644 index 0000000000000000000000000000000000000000..0c561af7796cee65cb911a7789d93f19202eeed6 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwVGkPsOM9g_00024_34_140.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_CwVGkPsOM9g_00024_34_140", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/CwVGkPsOM9g_00024_34_140_fore", + "duration": 3.571 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/CwVGkPsOM9g_00024_34_140_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/CwVGkPsOM9g_00024_34_140_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwzjmP57_Jw_00008_1542_1744.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwzjmP57_Jw_00008_1542_1744.json new file mode 100644 index 0000000000000000000000000000000000000000..338047d765de926b57ee46d1d2ba13d3b5b8b53c --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_CwzjmP57_Jw_00008_1542_1744.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_CwzjmP57_Jw_00008_1542_1744", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/CwzjmP57_Jw_00008_1542_1744_fore", + "duration": 3.4 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/CwzjmP57_Jw_00008_1542_1744_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/CwzjmP57_Jw_00008_1542_1744_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_D0-b-ovg4VY_00031_183_276.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_D0-b-ovg4VY_00031_183_276.json new file mode 100644 index 0000000000000000000000000000000000000000..246b3754cb83151a675ca8a631382e308aeff9dd --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_D0-b-ovg4VY_00031_183_276.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_D0-b-ovg4VY_00031_183_276", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/D0-b-ovg4VY_00031_183_276_fore", + "duration": 3.963 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/D0-b-ovg4VY_00031_183_276_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/D0-b-ovg4VY_00031_183_276_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DKMRRpmgitc_00031_0_398.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DKMRRpmgitc_00031_0_398.json new file mode 100644 index 0000000000000000000000000000000000000000..39cbcad7eb92892a4faf0f18c41bf2f7a1e6f54b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DKMRRpmgitc_00031_0_398.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DKMRRpmgitc_00031_0_398", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DKMRRpmgitc_00031_0_398_fore", + "duration": 13.292 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DKMRRpmgitc_00031_0_398_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DKMRRpmgitc_00031_0_398_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 13.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DQGAcsxCzPw_00002_47_121.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DQGAcsxCzPw_00002_47_121.json new file mode 100644 index 0000000000000000000000000000000000000000..3b672a454a2bd7afd4284e1d249f834b262e8622 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DQGAcsxCzPw_00002_47_121.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DQGAcsxCzPw_00002_47_121", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DQGAcsxCzPw_00002_47_121_fore", + "duration": 2.536 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DQGAcsxCzPw_00002_47_121_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DQGAcsxCzPw_00002_47_121_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DTJjOysP-bI_00000_314_397.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DTJjOysP-bI_00000_314_397.json new file mode 100644 index 0000000000000000000000000000000000000000..2fba81fcaf2845c3905b443a6c4f1f42c8037fa4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DTJjOysP-bI_00000_314_397.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DTJjOysP-bI_00000_314_397", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DTJjOysP-bI_00000_314_397_fore", + "duration": 2.834 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DTJjOysP-bI_00000_314_397_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DTJjOysP-bI_00000_314_397_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DWgCNBhHih8_00005_234_474.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DWgCNBhHih8_00005_234_474.json new file mode 100644 index 0000000000000000000000000000000000000000..299331d55278b76f628e7098a75f9f6ec1ab47ce --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DWgCNBhHih8_00005_234_474.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DWgCNBhHih8_00005_234_474", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DWgCNBhHih8_00005_234_474_fore", + "duration": 9.603 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DWgCNBhHih8_00005_234_474_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DWgCNBhHih8_00005_234_474_fore", + "rendered_video": "none", + "time_window": [ + 7.5, + 9.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DjYdSs-Uhsw_00032_0_96.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DjYdSs-Uhsw_00032_0_96.json new file mode 100644 index 0000000000000000000000000000000000000000..f2f4c2eed534c41d97bdbc41fc1b828eb4ad1562 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DjYdSs-Uhsw_00032_0_96.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DjYdSs-Uhsw_00032_0_96", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DjYdSs-Uhsw_00032_0_96_fore", + "duration": 4.046 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DjYdSs-Uhsw_00032_0_96_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DjYdSs-Uhsw_00032_0_96_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 3.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DpVbta6xaUw_00001_647_1125.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DpVbta6xaUw_00001_647_1125.json new file mode 100644 index 0000000000000000000000000000000000000000..cd9c95a4fc14a0bacf24ffde31f909db07b31868 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DpVbta6xaUw_00001_647_1125.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DpVbta6xaUw_00001_647_1125", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DpVbta6xaUw_00001_647_1125_fore", + "duration": 20.02 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DpVbta6xaUw_00001_647_1125_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DpVbta6xaUw_00001_647_1125_fore", + "rendered_video": "none", + "time_window": [ + 8.0, + 19.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DsU0Cm6PmzE_00000_0_104.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DsU0Cm6PmzE_00000_0_104.json new file mode 100644 index 0000000000000000000000000000000000000000..d3417f8e2be0c5b5834ecc6fe81787e073b57ab1 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_DsU0Cm6PmzE_00000_0_104.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_DsU0Cm6PmzE_00000_0_104", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/DsU0Cm6PmzE_00000_0_104_fore", + "duration": 4.338 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/DsU0Cm6PmzE_00000_0_104_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/DsU0Cm6PmzE_00000_0_104_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EAu_JbniRVw_00041_0_206.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EAu_JbniRVw_00041_0_206.json new file mode 100644 index 0000000000000000000000000000000000000000..c79e30c1009ad28b006c8ab9f843a985e18dcb48 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EAu_JbniRVw_00041_0_206.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EAu_JbniRVw_00041_0_206", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EAu_JbniRVw_00041_0_206_fore", + "duration": 8.36 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EAu_JbniRVw_00041_0_206_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EAu_JbniRVw_00041_0_206_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 8.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EIibo7aTpys_00012_0_97.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EIibo7aTpys_00012_0_97.json new file mode 100644 index 0000000000000000000000000000000000000000..3f68ff019e5c1fac7f1167f330668039966b2c78 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EIibo7aTpys_00012_0_97.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EIibo7aTpys_00012_0_97", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EIibo7aTpys_00012_0_97_fore", + "duration": 4.088 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EIibo7aTpys_00012_0_97_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EIibo7aTpys_00012_0_97_fore", + "rendered_video": "none", + "time_window": [ + 0.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EJuVEQtoi7E_00013_0_77.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EJuVEQtoi7E_00013_0_77.json new file mode 100644 index 0000000000000000000000000000000000000000..364c8e7593cc873ceca5f48b95931fafce1785b4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EJuVEQtoi7E_00013_0_77.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EJuVEQtoi7E_00013_0_77", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EJuVEQtoi7E_00013_0_77_fore", + "duration": 2.67 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EJuVEQtoi7E_00013_0_77_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EJuVEQtoi7E_00013_0_77_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ELV9T7pISm0_00001_0_145.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ELV9T7pISm0_00001_0_145.json new file mode 100644 index 0000000000000000000000000000000000000000..d8c4681106537e02c8fa2dd9a96e9f7092b28bcb --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ELV9T7pISm0_00001_0_145.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_ELV9T7pISm0_00001_0_145", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/ELV9T7pISm0_00001_0_145_fore", + "duration": 4.872 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/ELV9T7pISm0_00001_0_145_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/ELV9T7pISm0_00001_0_145_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EUA3T4A_5yA_00000_0_48.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EUA3T4A_5yA_00000_0_48.json new file mode 100644 index 0000000000000000000000000000000000000000..9385dd25c59288b6746591c63864c0fb30b8bfee --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EUA3T4A_5yA_00000_0_48.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EUA3T4A_5yA_00000_0_48", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EUA3T4A_5yA_00000_0_48_fore", + "duration": 2.042 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EUA3T4A_5yA_00000_0_48_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EUA3T4A_5yA_00000_0_48_fore", + "rendered_video": "none", + "time_window": [ + 0.2, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EiYsUTfJAw0_00058_0_189.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EiYsUTfJAw0_00058_0_189.json new file mode 100644 index 0000000000000000000000000000000000000000..71c3e9dd446a407d19acaf7a2c374a39cbaa1702 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EiYsUTfJAw0_00058_0_189.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EiYsUTfJAw0_00058_0_189", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EiYsUTfJAw0_00058_0_189_fore", + "duration": 3.8 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EiYsUTfJAw0_00058_0_189_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EiYsUTfJAw0_00058_0_189_fore", + "rendered_video": "none", + "time_window": [ + 2.8, + 3.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Eif-QsyWL7o_00069_0_234.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Eif-QsyWL7o_00069_0_234.json new file mode 100644 index 0000000000000000000000000000000000000000..bf6c04ac2cf9082c493b5e16dfab4bdd9abc256b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Eif-QsyWL7o_00069_0_234.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_Eif-QsyWL7o_00069_0_234", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/Eif-QsyWL7o_00069_0_234_fore", + "duration": 4.7 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/Eif-QsyWL7o_00069_0_234_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/Eif-QsyWL7o_00069_0_234_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EkMEGFBoYPc_00001_0_93.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EkMEGFBoYPc_00001_0_93.json new file mode 100644 index 0000000000000000000000000000000000000000..0d04dc51278e00b09e9d456f47cf42afbf6059e9 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_EkMEGFBoYPc_00001_0_93.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_EkMEGFBoYPc_00001_0_93", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/EkMEGFBoYPc_00001_0_93_fore", + "duration": 3.137 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/EkMEGFBoYPc_00001_0_93_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/EkMEGFBoYPc_00001_0_93_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ElIQTOQgWNw_00022_0_113.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ElIQTOQgWNw_00022_0_113.json new file mode 100644 index 0000000000000000000000000000000000000000..2e9dd2f63087a776cb884f7e5ab8479f7d042363 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ElIQTOQgWNw_00022_0_113.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_ElIQTOQgWNw_00022_0_113", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/ElIQTOQgWNw_00022_0_113_fore", + "duration": 3.804 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/ElIQTOQgWNw_00022_0_113_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/ElIQTOQgWNw_00022_0_113_fore", + "rendered_video": "none", + "time_window": [ + 1.9, + 3.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F-bdFiGotmg_00012_0_123.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F-bdFiGotmg_00012_0_123.json new file mode 100644 index 0000000000000000000000000000000000000000..bb3ac5f26c6a32f79fccac7ba61f1563d095b54b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F-bdFiGotmg_00012_0_123.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_F-bdFiGotmg_00012_0_123", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/F-bdFiGotmg_00012_0_123_fore", + "duration": 4.171 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/F-bdFiGotmg_00012_0_123_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/F-bdFiGotmg_00012_0_123_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F6M_9Gb3L8g_00006_173_472.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F6M_9Gb3L8g_00006_173_472.json new file mode 100644 index 0000000000000000000000000000000000000000..91f57dcea694c381236d6ba0f9f8f752d0ed7b3d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_F6M_9Gb3L8g_00006_173_472.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_F6M_9Gb3L8g_00006_173_472", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/F6M_9Gb3L8g_00006_173_472_fore", + "duration": 10.0 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/F6M_9Gb3L8g_00006_173_472_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/F6M_9Gb3L8g_00006_173_472_fore", + "rendered_video": "none", + "time_window": [ + 7.5, + 9.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FLnjU2oyn_0_00022_0_105.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FLnjU2oyn_0_00022_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..679a86aaea3642089fc9031d661976107605315a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FLnjU2oyn_0_00022_0_105.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_FLnjU2oyn_0_00022_0_105", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/FLnjU2oyn_0_00022_0_105_fore", + "duration": 3.571 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/FLnjU2oyn_0_00022_0_105_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/FLnjU2oyn_0_00022_0_105_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FSxSuRaBLwQ_00063_0_225.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FSxSuRaBLwQ_00063_0_225.json new file mode 100644 index 0000000000000000000000000000000000000000..be42791fa1f7161baf26ef8face9c55b25ca87a6 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FSxSuRaBLwQ_00063_0_225.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_FSxSuRaBLwQ_00063_0_225", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/FSxSuRaBLwQ_00063_0_225_fore", + "duration": 9.427 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/FSxSuRaBLwQ_00063_0_225_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/FSxSuRaBLwQ_00063_0_225_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 9.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FcFGSUx2FPM_00009_0_72.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FcFGSUx2FPM_00009_0_72.json new file mode 100644 index 0000000000000000000000000000000000000000..d99aef55b3d56de0919bcaae12651fa30e933dc2 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_FcFGSUx2FPM_00009_0_72.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_FcFGSUx2FPM_00009_0_72", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/FcFGSUx2FPM_00009_0_72_fore", + "duration": 2.92 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/FcFGSUx2FPM_00009_0_72_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/FcFGSUx2FPM_00009_0_72_fore", + "rendered_video": "none", + "time_window": [ + 1.3, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HMYRSj3Rju4_00100_0_123.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HMYRSj3Rju4_00100_0_123.json new file mode 100644 index 0000000000000000000000000000000000000000..81aef4b4a75a7eddf37a28d8c55745a16643158f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HMYRSj3Rju4_00100_0_123.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_HMYRSj3Rju4_00100_0_123", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/HMYRSj3Rju4_00100_0_123_fore", + "duration": 4.144 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/HMYRSj3Rju4_00100_0_123_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/HMYRSj3Rju4_00100_0_123_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 4.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HXBUzJigzrg_00028_0_960.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HXBUzJigzrg_00028_0_960.json new file mode 100644 index 0000000000000000000000000000000000000000..9f3f839206f418a971835e36f67ff58d2d8d9b91 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HXBUzJigzrg_00028_0_960.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_HXBUzJigzrg_00028_0_960", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/HXBUzJigzrg_00028_0_960_fore", + "duration": 16.306 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/HXBUzJigzrg_00028_0_960_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/HXBUzJigzrg_00028_0_960_fore", + "rendered_video": "none", + "time_window": [ + 9.0, + 16.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HbGLl85N7Cw_00001_0_65.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HbGLl85N7Cw_00001_0_65.json new file mode 100644 index 0000000000000000000000000000000000000000..a65c94596650950be1a339629ecdcbb1e0bce5a8 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HbGLl85N7Cw_00001_0_65.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_HbGLl85N7Cw_00001_0_65", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/HbGLl85N7Cw_00001_0_65_fore", + "duration": 2.64 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/HbGLl85N7Cw_00001_0_65_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/HbGLl85N7Cw_00001_0_65_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HfM2BjLPm1w_00023_0_175.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HfM2BjLPm1w_00023_0_175.json new file mode 100644 index 0000000000000000000000000000000000000000..a954894af7e231f7600064f4d55a7ff65ebf65db --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_HfM2BjLPm1w_00023_0_175.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_HfM2BjLPm1w_00023_0_175", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/HfM2BjLPm1w_00023_0_175_fore", + "duration": 5.873 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/HfM2BjLPm1w_00023_0_175_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/HfM2BjLPm1w_00023_0_175_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 5.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IHSaR35NJuM_00001_1062_1406.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IHSaR35NJuM_00001_1062_1406.json new file mode 100644 index 0000000000000000000000000000000000000000..d1b4ba7599bd3792cb2d26e95d4b9851d5c493b9 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IHSaR35NJuM_00001_1062_1406.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_IHSaR35NJuM_00001_1062_1406", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/IHSaR35NJuM_00001_1062_1406_fore", + "duration": 11.545 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/IHSaR35NJuM_00001_1062_1406_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/IHSaR35NJuM_00001_1062_1406_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 11.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IRJ014GJbYw_00073_0_102.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IRJ014GJbYw_00073_0_102.json new file mode 100644 index 0000000000000000000000000000000000000000..85353e9c3af046a373123e28725401bbc23c0279 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IRJ014GJbYw_00073_0_102.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_IRJ014GJbYw_00073_0_102", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/IRJ014GJbYw_00073_0_102_fore", + "duration": 4.338 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/IRJ014GJbYw_00073_0_102_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/IRJ014GJbYw_00073_0_102_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IXdgnxngfs4_00007_16_98.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IXdgnxngfs4_00007_16_98.json new file mode 100644 index 0000000000000000000000000000000000000000..69bdd16d09a3fc863f4216849ca1c0c3268351c7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IXdgnxngfs4_00007_16_98.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_IXdgnxngfs4_00007_16_98", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/IXdgnxngfs4_00007_16_98_fore", + "duration": 3.32 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/IXdgnxngfs4_00007_16_98_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/IXdgnxngfs4_00007_16_98_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IbuZuIWQ16k_00016_0_823.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IbuZuIWQ16k_00016_0_823.json new file mode 100644 index 0000000000000000000000000000000000000000..bfd9d02301b110199fe45accc7e943d0af3e29af --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IbuZuIWQ16k_00016_0_823.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_IbuZuIWQ16k_00016_0_823", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/IbuZuIWQ16k_00016_0_823_fore", + "duration": 13.764 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/IbuZuIWQ16k_00016_0_823_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/IbuZuIWQ16k_00016_0_823_fore", + "rendered_video": "none", + "time_window": [ + 10.0, + 13.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Ir83IWRVTOw_00004_0_242.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Ir83IWRVTOw_00004_0_242.json new file mode 100644 index 0000000000000000000000000000000000000000..0ed33e89c3fb4d9afa54aa4b9d09ea11245488ec --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Ir83IWRVTOw_00004_0_242.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_Ir83IWRVTOw_00004_0_242", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/Ir83IWRVTOw_00004_0_242_fore", + "duration": 10.084 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/Ir83IWRVTOw_00004_0_242_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/Ir83IWRVTOw_00004_0_242_fore", + "rendered_video": "none", + "time_window": [ + 7.5, + 10.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IvUYGasf4-g_00053_0_113.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IvUYGasf4-g_00053_0_113.json new file mode 100644 index 0000000000000000000000000000000000000000..522adfe00f0b13fc288c080e9bca7cd78e9ffe8b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_IvUYGasf4-g_00053_0_113.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_IvUYGasf4-g_00053_0_113", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/IvUYGasf4-g_00053_0_113_fore", + "duration": 3.904 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/IvUYGasf4-g_00053_0_113_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/IvUYGasf4-g_00053_0_113_fore", + "rendered_video": "none", + "time_window": [ + 1.4, + 3.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Iy5prfkaOAY_00001_238_354.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Iy5prfkaOAY_00001_238_354.json new file mode 100644 index 0000000000000000000000000000000000000000..7cea5e49fe09f7fa88110289ab04cf82f10e7d14 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_Iy5prfkaOAY_00001_238_354.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_Iy5prfkaOAY_00001_238_354", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/Iy5prfkaOAY_00001_238_354_fore", + "duration": 3.938 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/Iy5prfkaOAY_00001_238_354_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/Iy5prfkaOAY_00001_238_354_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J27DCHIJauw_00027_0_82.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J27DCHIJauw_00027_0_82.json new file mode 100644 index 0000000000000000000000000000000000000000..150a9fa45737657be51c2621b198179cb022f1a7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J27DCHIJauw_00027_0_82.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_J27DCHIJauw_00027_0_82", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/J27DCHIJauw_00027_0_82_fore", + "duration": 2.77 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/J27DCHIJauw_00027_0_82_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/J27DCHIJauw_00027_0_82_fore", + "rendered_video": "none", + "time_window": [ + 1.6, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J5p4Zmu7zDQ_00023_84_361.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J5p4Zmu7zDQ_00023_84_361.json new file mode 100644 index 0000000000000000000000000000000000000000..0084db405e7da1244db85ce5bbc2d7d0217c3615 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J5p4Zmu7zDQ_00023_84_361.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_J5p4Zmu7zDQ_00023_84_361", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/J5p4Zmu7zDQ_00023_84_361_fore", + "duration": 9.377 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/J5p4Zmu7zDQ_00023_84_361_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/J5p4Zmu7zDQ_00023_84_361_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 9.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JY1mxnXCeF8_00003_511_598.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JY1mxnXCeF8_00003_511_598.json new file mode 100644 index 0000000000000000000000000000000000000000..dfbe0f4e639a5b3050563ce8958e371b5af56fa4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JY1mxnXCeF8_00003_511_598.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_JY1mxnXCeF8_00003_511_598", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/JY1mxnXCeF8_00003_511_598_fore", + "duration": 2.904 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/JY1mxnXCeF8_00003_511_598_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/JY1mxnXCeF8_00003_511_598_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JYL3s66-0W4_00056_288_374.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JYL3s66-0W4_00056_288_374.json new file mode 100644 index 0000000000000000000000000000000000000000..8036f2f39d70d19a667ee02c8a2f4846b74de24a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JYL3s66-0W4_00056_288_374.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_JYL3s66-0W4_00056_288_374", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/JYL3s66-0W4_00056_288_374_fore", + "duration": 3.56 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/JYL3s66-0W4_00056_288_374_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/JYL3s66-0W4_00056_288_374_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J_4NWbKHP0I_00012_0_82.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J_4NWbKHP0I_00012_0_82.json new file mode 100644 index 0000000000000000000000000000000000000000..722fb65fc760b044e5a328d8056f189870b0001b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_J_4NWbKHP0I_00012_0_82.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_J_4NWbKHP0I_00012_0_82", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/J_4NWbKHP0I_00012_0_82_fore", + "duration": 2.77 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/J_4NWbKHP0I_00012_0_82_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/J_4NWbKHP0I_00012_0_82_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JvV640xrJ28_00026_0_84.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JvV640xrJ28_00026_0_84.json new file mode 100644 index 0000000000000000000000000000000000000000..fb56acc6f0716769b22a97e72c7450b4cb162224 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_JvV640xrJ28_00026_0_84.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_JvV640xrJ28_00026_0_84", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/JvV640xrJ28_00026_0_84_fore", + "duration": 3.503 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/JvV640xrJ28_00026_0_84_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/JvV640xrJ28_00026_0_84_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 3.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_K8hCBLTuVyk_00045_0_137.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_K8hCBLTuVyk_00045_0_137.json new file mode 100644 index 0000000000000000000000000000000000000000..0ce355dbdeeec38604db8b298603ba4191a62bf9 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_K8hCBLTuVyk_00045_0_137.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_K8hCBLTuVyk_00045_0_137", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/K8hCBLTuVyk_00045_0_137_fore", + "duration": 4.638 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/K8hCBLTuVyk_00045_0_137_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/K8hCBLTuVyk_00045_0_137_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRY1aqBGTCE_00057_0_57.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRY1aqBGTCE_00057_0_57.json new file mode 100644 index 0000000000000000000000000000000000000000..1ba86b2fbf6747d027d5d7af53b02939cb75c153 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRY1aqBGTCE_00057_0_57.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_KRY1aqBGTCE_00057_0_57", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/KRY1aqBGTCE_00057_0_57_fore", + "duration": 2.375 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/KRY1aqBGTCE_00057_0_57_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/KRY1aqBGTCE_00057_0_57_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00001_423_720.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00001_423_720.json new file mode 100644 index 0000000000000000000000000000000000000000..a0e7cb4b03d716b22251fe0f2e21710bf4f0ba45 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00001_423_720.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_KRheAHqm4SU_00001_423_720", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/KRheAHqm4SU_00001_423_720_fore", + "duration": 9.918 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/KRheAHqm4SU_00001_423_720_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/KRheAHqm4SU_00001_423_720_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 9.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00006_477_633.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00006_477_633.json new file mode 100644 index 0000000000000000000000000000000000000000..937160141840079a01b882b3eecf7683589b67a3 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KRheAHqm4SU_00006_477_633.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_KRheAHqm4SU_00006_477_633", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/KRheAHqm4SU_00006_477_633_fore", + "duration": 5.206 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/KRheAHqm4SU_00006_477_633_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/KRheAHqm4SU_00006_477_633_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KU5e1HH-j-c_00018_0_135.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KU5e1HH-j-c_00018_0_135.json new file mode 100644 index 0000000000000000000000000000000000000000..889510aef81dbe2463e8edf87bd1cf46404920da --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_KU5e1HH-j-c_00018_0_135.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_KU5e1HH-j-c_00018_0_135", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/KU5e1HH-j-c_00018_0_135_fore", + "duration": 4.634 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/KU5e1HH-j-c_00018_0_135_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/KU5e1HH-j-c_00018_0_135_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L1rESfCgUZo_00010_0_117.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L1rESfCgUZo_00010_0_117.json new file mode 100644 index 0000000000000000000000000000000000000000..560b0a90eb6f2207fb5f56c8a2c7821201e28952 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L1rESfCgUZo_00010_0_117.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_L1rESfCgUZo_00010_0_117", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/L1rESfCgUZo_00010_0_117_fore", + "duration": 3.938 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/L1rESfCgUZo_00010_0_117_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/L1rESfCgUZo_00010_0_117_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L44lwaZ-qAM_00004_0_170.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L44lwaZ-qAM_00004_0_170.json new file mode 100644 index 0000000000000000000000000000000000000000..f2689b1d7ac705df71d827bf203f9ffe67e5fb2c --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L44lwaZ-qAM_00004_0_170.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_L44lwaZ-qAM_00004_0_170", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/L44lwaZ-qAM_00004_0_170_fore", + "duration": 7.435 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/L44lwaZ-qAM_00004_0_170_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/L44lwaZ-qAM_00004_0_170_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 7.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L72G8TLtTCk_00014_0_105.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L72G8TLtTCk_00014_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..303dd3cd03312057ae2ae53cc0c561195d950c7c --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_L72G8TLtTCk_00014_0_105.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_L72G8TLtTCk_00014_0_105", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/L72G8TLtTCk_00014_0_105_fore", + "duration": 4.421 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/L72G8TLtTCk_00014_0_105_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/L72G8TLtTCk_00014_0_105_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LALTJJyO2js_00000_0_66.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LALTJJyO2js_00000_0_66.json new file mode 100644 index 0000000000000000000000000000000000000000..6ff87c032b9a1b0e2805ba75d1e3e9730f724a83 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LALTJJyO2js_00000_0_66.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LALTJJyO2js_00000_0_66", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LALTJJyO2js_00000_0_66_fore", + "duration": 2.68 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LALTJJyO2js_00000_0_66_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LALTJJyO2js_00000_0_66_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LDM9aXVAoOw_00002_0_694.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LDM9aXVAoOw_00002_0_694.json new file mode 100644 index 0000000000000000000000000000000000000000..f8cf048c0caa83936e25ae9fba0053ac65e6bd18 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LDM9aXVAoOw_00002_0_694.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LDM9aXVAoOw_00002_0_694", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LDM9aXVAoOw_00002_0_694_fore", + "duration": 13.9 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LDM9aXVAoOw_00002_0_694_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LDM9aXVAoOw_00002_0_694_fore", + "rendered_video": "none", + "time_window": [ + 10.0, + 13.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LEppqung-TQ_00006_0_109.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LEppqung-TQ_00006_0_109.json new file mode 100644 index 0000000000000000000000000000000000000000..98df01391eb314abab273863f5506b859142afff --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LEppqung-TQ_00006_0_109.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LEppqung-TQ_00006_0_109", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LEppqung-TQ_00006_0_109_fore", + "duration": 4.4 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LEppqung-TQ_00006_0_109_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LEppqung-TQ_00006_0_109_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LGoX7vNg3IM_00026_592_719.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LGoX7vNg3IM_00026_592_719.json new file mode 100644 index 0000000000000000000000000000000000000000..d43f1cf7ca18f172d289b5bf54ecfad4b02b7fad --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LGoX7vNg3IM_00026_592_719.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LGoX7vNg3IM_00026_592_719", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LGoX7vNg3IM_00026_592_719_fore", + "duration": 2.156 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LGoX7vNg3IM_00026_592_719_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LGoX7vNg3IM_00026_592_719_fore", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LNvgFIUv9hI_00023_0_59.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LNvgFIUv9hI_00023_0_59.json new file mode 100644 index 0000000000000000000000000000000000000000..25b485836e8802205f51495f7a8bd159e0a11857 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LNvgFIUv9hI_00023_0_59.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LNvgFIUv9hI_00023_0_59", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LNvgFIUv9hI_00023_0_59_fore", + "duration": 2.44 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LNvgFIUv9hI_00023_0_59_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LNvgFIUv9hI_00023_0_59_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LOWyPvH0210_00002_0_103.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LOWyPvH0210_00002_0_103.json new file mode 100644 index 0000000000000000000000000000000000000000..9344e8893080a66376052b3a15ff1f24cf7fe164 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LOWyPvH0210_00002_0_103.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LOWyPvH0210_00002_0_103", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LOWyPvH0210_00002_0_103_fore", + "duration": 3.537 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LOWyPvH0210_00002_0_103_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LOWyPvH0210_00002_0_103_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LTozxHr17ME_00024_0_119.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LTozxHr17ME_00024_0_119.json new file mode 100644 index 0000000000000000000000000000000000000000..80ad96c525f6f0d2b927da70176256eade8d8728 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LTozxHr17ME_00024_0_119.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LTozxHr17ME_00024_0_119", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LTozxHr17ME_00024_0_119_fore", + "duration": 4.004 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LTozxHr17ME_00024_0_119_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LTozxHr17ME_00024_0_119_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LX5G4VkRH3o_00035_0_74.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LX5G4VkRH3o_00035_0_74.json new file mode 100644 index 0000000000000000000000000000000000000000..e613ca0431f62dd8f88fdbfd64a8ecbeabd66a2f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LX5G4VkRH3o_00035_0_74.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LX5G4VkRH3o_00035_0_74", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LX5G4VkRH3o_00035_0_74_fore", + "duration": 3.167 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LX5G4VkRH3o_00035_0_74_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LX5G4VkRH3o_00035_0_74_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LZ9kV0wNDrY_00011_393_503.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LZ9kV0wNDrY_00011_393_503.json new file mode 100644 index 0000000000000000000000000000000000000000..4f3d44eccb1c4d3bdc3708ebcd4973b89cf295ff --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LZ9kV0wNDrY_00011_393_503.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LZ9kV0wNDrY_00011_393_503", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LZ9kV0wNDrY_00011_393_503_fore", + "duration": 3.641 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LZ9kV0wNDrY_00011_393_503_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LZ9kV0wNDrY_00011_393_503_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LtVkTfmcyOw_00048_0_99.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LtVkTfmcyOw_00048_0_99.json new file mode 100644 index 0000000000000000000000000000000000000000..9327cff5fce41719fe98c47dccd7df08d40f3427 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_LtVkTfmcyOw_00048_0_99.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_LtVkTfmcyOw_00048_0_99", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/LtVkTfmcyOw_00048_0_99_fore", + "duration": 3.337 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/LtVkTfmcyOw_00048_0_99_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/LtVkTfmcyOw_00048_0_99_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_M3oVDmDIHek_00095_0_134.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_M3oVDmDIHek_00095_0_134.json new file mode 100644 index 0000000000000000000000000000000000000000..f7d668c683bd73524c850447ebcb356a75bd9b10 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_M3oVDmDIHek_00095_0_134.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_M3oVDmDIHek_00095_0_134", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/M3oVDmDIHek_00095_0_134_fore", + "duration": 2.701 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/M3oVDmDIHek_00095_0_134_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/M3oVDmDIHek_00095_0_134_fore", + "rendered_video": "none", + "time_window": [ + 1.6, + 2.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__1tzdCdC4UY_00005_77_203.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__1tzdCdC4UY_00005_77_203.json new file mode 100644 index 0000000000000000000000000000000000000000..212830d9f986d998a32f8173da702c98c54a1c5b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__1tzdCdC4UY_00005_77_203.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__1tzdCdC4UY_00005_77_203", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_1tzdCdC4UY_00005_77_203_fore", + "duration": 5.259 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_1tzdCdC4UY_00005_77_203_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_1tzdCdC4UY_00005_77_203_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__MXhQXP5hEI_00002_0_662.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__MXhQXP5hEI_00002_0_662.json new file mode 100644 index 0000000000000000000000000000000000000000..1996dc5442ad5e24e03e8d5ab9e1cfc2c12f5aa5 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__MXhQXP5hEI_00002_0_662.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__MXhQXP5hEI_00002_0_662", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_MXhQXP5hEI_00002_0_662_fore", + "duration": 11.08 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_MXhQXP5hEI_00002_0_662_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_MXhQXP5hEI_00002_0_662_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 10.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__dFAYYMJbWk_00193_268_382.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__dFAYYMJbWk_00193_268_382.json new file mode 100644 index 0000000000000000000000000000000000000000..6effa3a4c5682614afae77f6d3abd55757886f14 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__dFAYYMJbWk_00193_268_382.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__dFAYYMJbWk_00193_268_382", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_dFAYYMJbWk_00193_268_382_fore", + "duration": 4.6 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_dFAYYMJbWk_00193_268_382_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_dFAYYMJbWk_00193_268_382_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__eyx3-8oIr0_00009_0_234.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__eyx3-8oIr0_00009_0_234.json new file mode 100644 index 0000000000000000000000000000000000000000..74cdc91a33b9e9024e339dad65a26c83afb0b093 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__eyx3-8oIr0_00009_0_234.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__eyx3-8oIr0_00009_0_234", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_eyx3-8oIr0_00009_0_234_fore", + "duration": 7.842 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_eyx3-8oIr0_00009_0_234_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_eyx3-8oIr0_00009_0_234_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 7.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__fIznC6ZIX0_00032_0_67.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__fIznC6ZIX0_00032_0_67.json new file mode 100644 index 0000000000000000000000000000000000000000..c0ebb5a8f483864830e670058cfe0538429f79de --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__fIznC6ZIX0_00032_0_67.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__fIznC6ZIX0_00032_0_67", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_fIznC6ZIX0_00032_0_67_fore", + "duration": 2.267 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_fIznC6ZIX0_00032_0_67_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_fIznC6ZIX0_00032_0_67_fore", + "rendered_video": "none", + "time_window": [ + 1.4, + 2.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__oYIT7I3avY_00018_947_1296.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__oYIT7I3avY_00018_947_1296.json new file mode 100644 index 0000000000000000000000000000000000000000..d57b1d5fb2e32823a984194cb7c7cd59885e6783 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN__oYIT7I3avY_00018_947_1296.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN__oYIT7I3avY_00018_947_1296", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/_oYIT7I3avY_00018_947_1296_fore", + "duration": 11.679 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/_oYIT7I3avY_00018_947_1296_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/_oYIT7I3avY_00018_947_1296_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 11.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_a500SZyQFxQ_00014_0_65.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_a500SZyQFxQ_00014_0_65.json new file mode 100644 index 0000000000000000000000000000000000000000..c8bc29fb21b504113f123fb7f10b3c0127c0c25e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_a500SZyQFxQ_00014_0_65.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_a500SZyQFxQ_00014_0_65", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/a500SZyQFxQ_00014_0_65_fore", + "duration": 2.711 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/a500SZyQFxQ_00014_0_65_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/a500SZyQFxQ_00014_0_65_fore", + "rendered_video": "none", + "time_window": [ + 0.6, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aDuB1arf8aE_00030_0_111.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aDuB1arf8aE_00030_0_111.json new file mode 100644 index 0000000000000000000000000000000000000000..860614b8d7d535280d6087625f0022946f9b5d0a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aDuB1arf8aE_00030_0_111.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_aDuB1arf8aE_00030_0_111", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/aDuB1arf8aE_00030_0_111_fore", + "duration": 3.771 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/aDuB1arf8aE_00030_0_111_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/aDuB1arf8aE_00030_0_111_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aEFB3NGhx9E_00034_0_146.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aEFB3NGhx9E_00034_0_146.json new file mode 100644 index 0000000000000000000000000000000000000000..2e90171d10306de83315f5bb45f8c0e4d5fd8f36 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_aEFB3NGhx9E_00034_0_146.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_aEFB3NGhx9E_00034_0_146", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/aEFB3NGhx9E_00034_0_146_fore", + "duration": 4.905 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/aEFB3NGhx9E_00034_0_146_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/aEFB3NGhx9E_00034_0_146_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 4.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ae4jA0iRvtg_00030_0_120.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ae4jA0iRvtg_00030_0_120.json new file mode 100644 index 0000000000000000000000000000000000000000..33f8f9667791a830fa01ca5a43f9e5ef61b13c1f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ae4jA0iRvtg_00030_0_120.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_ae4jA0iRvtg_00030_0_120", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/ae4jA0iRvtg_00030_0_120_fore", + "duration": 2.42 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/ae4jA0iRvtg_00030_0_120_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/ae4jA0iRvtg_00030_0_120_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_b9-gJuDgKnU_00012_0_86.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_b9-gJuDgKnU_00012_0_86.json new file mode 100644 index 0000000000000000000000000000000000000000..8a65d0d831c2b8a9074c353b8d3da408b7fefea4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_b9-gJuDgKnU_00012_0_86.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_b9-gJuDgKnU_00012_0_86", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/b9-gJuDgKnU_00012_0_86_fore", + "duration": 3.48 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/b9-gJuDgKnU_00012_0_86_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/b9-gJuDgKnU_00012_0_86_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bFX5Jd2bHS8_00001_0_135.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bFX5Jd2bHS8_00001_0_135.json new file mode 100644 index 0000000000000000000000000000000000000000..3a3d524bda62e09388168a1e6ed4030544c6050f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bFX5Jd2bHS8_00001_0_135.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bFX5Jd2bHS8_00001_0_135", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bFX5Jd2bHS8_00001_0_135_fore", + "duration": 4.538 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bFX5Jd2bHS8_00001_0_135_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bFX5Jd2bHS8_00001_0_135_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bN-o8SRUdC4_00010_67_186.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bN-o8SRUdC4_00010_67_186.json new file mode 100644 index 0000000000000000000000000000000000000000..aacda00be314731081e7d61dbb294619241f4a93 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bN-o8SRUdC4_00010_67_186.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bN-o8SRUdC4_00010_67_186", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bN-o8SRUdC4_00010_67_186_fore", + "duration": 4.0 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bN-o8SRUdC4_00010_67_186_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bN-o8SRUdC4_00010_67_186_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bQoRtsNwRCo_00026_711_840.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bQoRtsNwRCo_00026_711_840.json new file mode 100644 index 0000000000000000000000000000000000000000..cfd413e2b36062cccd4d41afcad50c6ec20cbb4e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bQoRtsNwRCo_00026_711_840.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bQoRtsNwRCo_00026_711_840", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bQoRtsNwRCo_00026_711_840_fore", + "duration": 4.372 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bQoRtsNwRCo_00026_711_840_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bQoRtsNwRCo_00026_711_840_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bS5ve8rG-Ik_00052_0_93.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bS5ve8rG-Ik_00052_0_93.json new file mode 100644 index 0000000000000000000000000000000000000000..df6e1f532f7f76087095f1696899c109aededb7e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bS5ve8rG-Ik_00052_0_93.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bS5ve8rG-Ik_00052_0_93", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bS5ve8rG-Ik_00052_0_93_fore", + "duration": 3.137 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bS5ve8rG-Ik_00052_0_93_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bS5ve8rG-Ik_00052_0_93_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bSdMZBG-tDg_00016_0_359.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bSdMZBG-tDg_00016_0_359.json new file mode 100644 index 0000000000000000000000000000000000000000..7c8ea0afe2896883b5b15c08f14137a41bafc060 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bSdMZBG-tDg_00016_0_359.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bSdMZBG-tDg_00016_0_359", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bSdMZBG-tDg_00016_0_359_fore", + "duration": 6.011 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bSdMZBG-tDg_00016_0_359_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bSdMZBG-tDg_00016_0_359_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bohbtpZJkXo_00042_38_96.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bohbtpZJkXo_00042_38_96.json new file mode 100644 index 0000000000000000000000000000000000000000..872dc044a4a7d9b017f7e07af4bc0a63b2cf35b8 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bohbtpZJkXo_00042_38_96.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bohbtpZJkXo_00042_38_96", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bohbtpZJkXo_00042_38_96_fore", + "duration": 2.36 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bohbtpZJkXo_00042_38_96_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bohbtpZJkXo_00042_38_96_fore", + "rendered_video": "none", + "time_window": [ + 0.7, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bwr6x5xT9nE_00001_120_711.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bwr6x5xT9nE_00001_120_711.json new file mode 100644 index 0000000000000000000000000000000000000000..25163f9899057122b25411e40181c1ca0962bb90 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_bwr6x5xT9nE_00001_120_711.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_bwr6x5xT9nE_00001_120_711", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/bwr6x5xT9nE_00001_120_711_fore", + "duration": 9.888 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/bwr6x5xT9nE_00001_120_711_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/bwr6x5xT9nE_00001_120_711_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 9.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c03xSP9PzyA_00044_0_101.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c03xSP9PzyA_00044_0_101.json new file mode 100644 index 0000000000000000000000000000000000000000..d77401447d062a3bb7b56d5ae7913c81dd695378 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c03xSP9PzyA_00044_0_101.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_c03xSP9PzyA_00044_0_101", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/c03xSP9PzyA_00044_0_101_fore", + "duration": 4.08 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/c03xSP9PzyA_00044_0_101_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/c03xSP9PzyA_00044_0_101_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c3Dh8n1qyCY_00009_0_66.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c3Dh8n1qyCY_00009_0_66.json new file mode 100644 index 0000000000000000000000000000000000000000..cf15be8938d22eb9178186c8364609d73324998b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_c3Dh8n1qyCY_00009_0_66.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_c3Dh8n1qyCY_00009_0_66", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/c3Dh8n1qyCY_00009_0_66_fore", + "duration": 2.236 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/c3Dh8n1qyCY_00009_0_66_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/c3Dh8n1qyCY_00009_0_66_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cAEYHc5dzfs_00003_297_392.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cAEYHc5dzfs_00003_297_392.json new file mode 100644 index 0000000000000000000000000000000000000000..f332222da2a94165ac1caa57f128704e586ddb56 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cAEYHc5dzfs_00003_297_392.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_cAEYHc5dzfs_00003_297_392", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/cAEYHc5dzfs_00003_297_392_fore", + "duration": 3.204 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/cAEYHc5dzfs_00003_297_392_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/cAEYHc5dzfs_00003_297_392_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cEdo6STidC0_00023_361_458.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cEdo6STidC0_00023_361_458.json new file mode 100644 index 0000000000000000000000000000000000000000..e7cabd55a63dde6f7c12d5ef225bda3ef42876f7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cEdo6STidC0_00023_361_458.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_cEdo6STidC0_00023_361_458", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/cEdo6STidC0_00023_361_458_fore", + "duration": 3.27 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/cEdo6STidC0_00023_361_458_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/cEdo6STidC0_00023_361_458_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 3.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cUUQtoBnAJU_00000_2062_2222.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cUUQtoBnAJU_00000_2062_2222.json new file mode 100644 index 0000000000000000000000000000000000000000..94475d508e18f2d56ed52062e93ccda91a0a6e39 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_cUUQtoBnAJU_00000_2062_2222.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_cUUQtoBnAJU_00000_2062_2222", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/cUUQtoBnAJU_00000_2062_2222_fore", + "duration": 2.683 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/cUUQtoBnAJU_00000_2062_2222_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/cUUQtoBnAJU_00000_2062_2222_fore", + "rendered_video": "none", + "time_window": [ + 1.6, + 2.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_d3uu7C26flM_00001_0_83.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_d3uu7C26flM_00001_0_83.json new file mode 100644 index 0000000000000000000000000000000000000000..0036ddc63ae6aaf689c123b9d051932b8725a1e7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_d3uu7C26flM_00001_0_83.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_d3uu7C26flM_00001_0_83", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/d3uu7C26flM_00001_0_83_fore", + "duration": 2.803 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/d3uu7C26flM_00001_0_83_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/d3uu7C26flM_00001_0_83_fore", + "rendered_video": "none", + "time_window": [ + 1.7, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00045_0_209.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00045_0_209.json new file mode 100644 index 0000000000000000000000000000000000000000..10689df195665ab2dc88a07d092e1600fdf508c0 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00045_0_209.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_htSmYNfT8QA_00045_0_209", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/htSmYNfT8QA_00045_0_209_fore", + "duration": 6.983 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/htSmYNfT8QA_00045_0_209_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/htSmYNfT8QA_00045_0_209_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 7.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00054_0_195.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00054_0_195.json new file mode 100644 index 0000000000000000000000000000000000000000..65f400692140c422ef248687c0b3ef07c94305dc --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_htSmYNfT8QA_00054_0_195.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_htSmYNfT8QA_00054_0_195", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/htSmYNfT8QA_00054_0_195_fore", + "duration": 6.541 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/htSmYNfT8QA_00054_0_195_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/htSmYNfT8QA_00054_0_195_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_httbW7TujP0_00057_0_170.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_httbW7TujP0_00057_0_170.json new file mode 100644 index 0000000000000000000000000000000000000000..c3164f4aace6b459175dad4c249b3256777f9154 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_httbW7TujP0_00057_0_170.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_httbW7TujP0_00057_0_170", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/httbW7TujP0_00057_0_170_fore", + "duration": 5.673 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/httbW7TujP0_00057_0_170_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/httbW7TujP0_00057_0_170_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_iIVJN0Yz1Y0_00027_0_84.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_iIVJN0Yz1Y0_00027_0_84.json new file mode 100644 index 0000000000000000000000000000000000000000..1937cea5f07c16c7c3567162223336ef832f66a4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_iIVJN0Yz1Y0_00027_0_84.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_iIVJN0Yz1Y0_00027_0_84", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/iIVJN0Yz1Y0_00027_0_84_fore", + "duration": 3.4 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/iIVJN0Yz1Y0_00027_0_84_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/iIVJN0Yz1Y0_00027_0_84_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_is-juN3Pkfg_00008_0_101.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_is-juN3Pkfg_00008_0_101.json new file mode 100644 index 0000000000000000000000000000000000000000..b0845e961c933f3150b81c696c3da72ac9283007 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_is-juN3Pkfg_00008_0_101.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_is-juN3Pkfg_00008_0_101", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/is-juN3Pkfg_00008_0_101_fore", + "duration": 4.255 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/is-juN3Pkfg_00008_0_101_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/is-juN3Pkfg_00008_0_101_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ish8GggzGsA_00006_0_109.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ish8GggzGsA_00006_0_109.json new file mode 100644 index 0000000000000000000000000000000000000000..e6ec555c0c9c234a1c65f3eb885055aebc4ebb0f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ish8GggzGsA_00006_0_109.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_ish8GggzGsA_00006_0_109", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/ish8GggzGsA_00006_0_109_fore", + "duration": 3.704 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/ish8GggzGsA_00006_0_109_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/ish8GggzGsA_00006_0_109_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_j88cdZCpDEM_00008_0_150.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_j88cdZCpDEM_00008_0_150.json new file mode 100644 index 0000000000000000000000000000000000000000..c437a12731321b8c402501e1b731fccb1d00394a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_j88cdZCpDEM_00008_0_150.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_j88cdZCpDEM_00008_0_150", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/j88cdZCpDEM_00008_0_150_fore", + "duration": 2.514 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/j88cdZCpDEM_00008_0_150_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/j88cdZCpDEM_00008_0_150_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jA5Xq4Bs2yE_00065_0_87.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jA5Xq4Bs2yE_00065_0_87.json new file mode 100644 index 0000000000000000000000000000000000000000..228afb04dd56f58cad499d12a3d78c434fcbadec --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jA5Xq4Bs2yE_00065_0_87.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jA5Xq4Bs2yE_00065_0_87", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jA5Xq4Bs2yE_00065_0_87_fore", + "duration": 2.97 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jA5Xq4Bs2yE_00065_0_87_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jA5Xq4Bs2yE_00065_0_87_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jD8dLyj_7-U_00011_0_95.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jD8dLyj_7-U_00011_0_95.json new file mode 100644 index 0000000000000000000000000000000000000000..b9164414b561b599c8d300454b8cdf332ef923fc --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jD8dLyj_7-U_00011_0_95.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jD8dLyj_7-U_00011_0_95", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jD8dLyj_7-U_00011_0_95_fore", + "duration": 3.204 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jD8dLyj_7-U_00011_0_95_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jD8dLyj_7-U_00011_0_95_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_je-1dG6dBbM_00008_0_114.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_je-1dG6dBbM_00008_0_114.json new file mode 100644 index 0000000000000000000000000000000000000000..2e66bb9eca42c7471b6e7b7332018d469e0e336a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_je-1dG6dBbM_00008_0_114.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_je-1dG6dBbM_00008_0_114", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/je-1dG6dBbM_00008_0_114_fore", + "duration": 4.797 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/je-1dG6dBbM_00008_0_114_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/je-1dG6dBbM_00008_0_114_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jfxw2B0uBX0_00031_293_436.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jfxw2B0uBX0_00031_293_436.json new file mode 100644 index 0000000000000000000000000000000000000000..9eaf2289a033e96e24eaa63efb3b5bf8d53b621b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jfxw2B0uBX0_00031_293_436.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jfxw2B0uBX0_00031_293_436", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jfxw2B0uBX0_00031_293_436_fore", + "duration": 2.41 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jfxw2B0uBX0_00031_293_436_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jfxw2B0uBX0_00031_293_436_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jhlZItT7mmo_00005_235_320.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jhlZItT7mmo_00005_235_320.json new file mode 100644 index 0000000000000000000000000000000000000000..1d20f9e0e1736e73ee47f659cba14e505c30fc55 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jhlZItT7mmo_00005_235_320.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jhlZItT7mmo_00005_235_320", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jhlZItT7mmo_00005_235_320_fore", + "duration": 2.838 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jhlZItT7mmo_00005_235_320_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jhlZItT7mmo_00005_235_320_fore", + "rendered_video": "none", + "time_window": [ + 0.7, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jmw0Kz5K1ms_00014_0_162.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jmw0Kz5K1ms_00014_0_162.json new file mode 100644 index 0000000000000000000000000000000000000000..6404d47ff5b0edfff4daf2ce93f439d68bf14e42 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jmw0Kz5K1ms_00014_0_162.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jmw0Kz5K1ms_00014_0_162", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jmw0Kz5K1ms_00014_0_162_fore", + "duration": 5.413 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jmw0Kz5K1ms_00014_0_162_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jmw0Kz5K1ms_00014_0_162_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 5.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jnVb7_CVh6U_00003_173_269.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jnVb7_CVh6U_00003_173_269.json new file mode 100644 index 0000000000000000000000000000000000000000..4ce5cc4bb765346c0dc7864e36380680a131660a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jnVb7_CVh6U_00003_173_269.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jnVb7_CVh6U_00003_173_269", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jnVb7_CVh6U_00003_173_269_fore", + "duration": 3.88 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jnVb7_CVh6U_00003_173_269_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jnVb7_CVh6U_00003_173_269_fore", + "rendered_video": "none", + "time_window": [ + 0.7, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jqNk4ExKLXI_00033_14_100.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jqNk4ExKLXI_00033_14_100.json new file mode 100644 index 0000000000000000000000000000000000000000..57034aee970bdb2660822eeb32fcb0bf1aa947cf --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jqNk4ExKLXI_00033_14_100.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jqNk4ExKLXI_00033_14_100", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jqNk4ExKLXI_00033_14_100_fore", + "duration": 3.56 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jqNk4ExKLXI_00033_14_100_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jqNk4ExKLXI_00033_14_100_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jz8L6htl0yA_00038_311_415.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jz8L6htl0yA_00038_311_415.json new file mode 100644 index 0000000000000000000000000000000000000000..f5453010d9e20e7779bf698deeadc233d52ebc00 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_jz8L6htl0yA_00038_311_415.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_jz8L6htl0yA_00038_311_415", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/jz8L6htl0yA_00038_311_415_fore", + "duration": 3.504 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/jz8L6htl0yA_00038_311_415_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/jz8L6htl0yA_00038_311_415_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_k0X8Mz1ksEg_00013_0_154.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_k0X8Mz1ksEg_00013_0_154.json new file mode 100644 index 0000000000000000000000000000000000000000..f1ec6726f6582c4771d6072fd86952dfd479b1b0 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_k0X8Mz1ksEg_00013_0_154.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_k0X8Mz1ksEg_00013_0_154", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/k0X8Mz1ksEg_00013_0_154_fore", + "duration": 6.2 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/k0X8Mz1ksEg_00013_0_154_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/k0X8Mz1ksEg_00013_0_154_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHQ8ib3EQb8_00014_0_556.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHQ8ib3EQb8_00014_0_556.json new file mode 100644 index 0000000000000000000000000000000000000000..1d9c736cb32d327d97aeaf850e1e915647c24718 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHQ8ib3EQb8_00014_0_556.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kHQ8ib3EQb8_00014_0_556", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kHQ8ib3EQb8_00014_0_556_fore", + "duration": 11.14 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kHQ8ib3EQb8_00014_0_556_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kHQ8ib3EQb8_00014_0_556_fore", + "rendered_video": "none", + "time_window": [ + 6.5, + 10.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHU2bS89gVI_00010_178_246.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHU2bS89gVI_00010_178_246.json new file mode 100644 index 0000000000000000000000000000000000000000..0f71c57da46ff2ac2107d30523bba4edbec09884 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kHU2bS89gVI_00010_178_246.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kHU2bS89gVI_00010_178_246", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kHU2bS89gVI_00010_178_246_fore", + "duration": 2.336 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kHU2bS89gVI_00010_178_246_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kHU2bS89gVI_00010_178_246_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kN8xl4KgjNE_00000_29_192.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kN8xl4KgjNE_00000_29_192.json new file mode 100644 index 0000000000000000000000000000000000000000..ad287a585451c8d06a2fbb00c9c4cb55cb78884f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kN8xl4KgjNE_00000_29_192.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kN8xl4KgjNE_00000_29_192", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kN8xl4KgjNE_00000_29_192_fore", + "duration": 6.793 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kN8xl4KgjNE_00000_29_192_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kN8xl4KgjNE_00000_29_192_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kSm_3Y8jnrA_00007_113_301.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kSm_3Y8jnrA_00007_113_301.json new file mode 100644 index 0000000000000000000000000000000000000000..743c821f7c1861117020eba384ee59363a2d32a4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kSm_3Y8jnrA_00007_113_301.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kSm_3Y8jnrA_00007_113_301", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kSm_3Y8jnrA_00007_113_301_fore", + "duration": 3.157 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kSm_3Y8jnrA_00007_113_301_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kSm_3Y8jnrA_00007_113_301_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 3.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kgmq3A6RzIw_00000_325_677.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kgmq3A6RzIw_00000_325_677.json new file mode 100644 index 0000000000000000000000000000000000000000..4e5230fd8a650ff64abbc3379a9e9b63ad5ee70f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kgmq3A6RzIw_00000_325_677.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kgmq3A6RzIw_00000_325_677", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kgmq3A6RzIw_00000_325_677_fore", + "duration": 11.779 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kgmq3A6RzIw_00000_325_677_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kgmq3A6RzIw_00000_325_677_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 11.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_km7u5M35xb4_00025_0_105.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_km7u5M35xb4_00025_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..db89fe6eb66f2e6f497cfb224d09945602ef643b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_km7u5M35xb4_00025_0_105.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_km7u5M35xb4_00025_0_105", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/km7u5M35xb4_00025_0_105_fore", + "duration": 4.24 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/km7u5M35xb4_00025_0_105_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/km7u5M35xb4_00025_0_105_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kvfNBJXHugg_00048_549_824.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kvfNBJXHugg_00048_549_824.json new file mode 100644 index 0000000000000000000000000000000000000000..e4602e2019bb32ebedc323dd107951f5921affc2 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_kvfNBJXHugg_00048_549_824.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_kvfNBJXHugg_00048_549_824", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/kvfNBJXHugg_00048_549_824_fore", + "duration": 9.276 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/kvfNBJXHugg_00048_549_824_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/kvfNBJXHugg_00048_549_824_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 8.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qqANXkZiVoU_00027_0_254.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qqANXkZiVoU_00027_0_254.json new file mode 100644 index 0000000000000000000000000000000000000000..cb83afc740df925e5bb0536f86bbe43ed4a6fd5b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qqANXkZiVoU_00027_0_254.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_qqANXkZiVoU_00027_0_254", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/qqANXkZiVoU_00027_0_254_fore", + "duration": 4.271 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/qqANXkZiVoU_00027_0_254_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/qqANXkZiVoU_00027_0_254_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 4.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qvQSt_eqISY_00002_103_188.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qvQSt_eqISY_00002_103_188.json new file mode 100644 index 0000000000000000000000000000000000000000..28a88aaf45d6378faf31e3614b6ef6b682518503 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_qvQSt_eqISY_00002_103_188.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_qvQSt_eqISY_00002_103_188", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/qvQSt_eqISY_00002_103_188_fore", + "duration": 2.87 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/qvQSt_eqISY_00002_103_188_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/qvQSt_eqISY_00002_103_188_fore", + "rendered_video": "none", + "time_window": [ + 0.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_r8gxkZynhvk_00008_0_104.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_r8gxkZynhvk_00008_0_104.json new file mode 100644 index 0000000000000000000000000000000000000000..d36cf317a92438a29e271157446694734b09606b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_r8gxkZynhvk_00008_0_104.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_r8gxkZynhvk_00008_0_104", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/r8gxkZynhvk_00008_0_104_fore", + "duration": 3.537 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/r8gxkZynhvk_00008_0_104_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/r8gxkZynhvk_00008_0_104_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rEtmuByf-0E_00006_1922_2189.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rEtmuByf-0E_00006_1922_2189.json new file mode 100644 index 0000000000000000000000000000000000000000..f65fe22056abe266afb9207a15b952ef7fa7c019 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rEtmuByf-0E_00006_1922_2189.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rEtmuByf-0E_00006_1922_2189", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rEtmuByf-0E_00006_1922_2189_fore", + "duration": 5.4 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rEtmuByf-0E_00006_1922_2189_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rEtmuByf-0E_00006_1922_2189_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rGsF1notX2A_00020_0_88.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rGsF1notX2A_00020_0_88.json new file mode 100644 index 0000000000000000000000000000000000000000..49c80b821b317666b00fd22cb51dce7926304167 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rGsF1notX2A_00020_0_88.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rGsF1notX2A_00020_0_88", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rGsF1notX2A_00020_0_88_fore", + "duration": 3.003 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rGsF1notX2A_00020_0_88_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rGsF1notX2A_00020_0_88_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rLoHXD9dxUU_00042_97_166.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rLoHXD9dxUU_00042_97_166.json new file mode 100644 index 0000000000000000000000000000000000000000..05c972384101fbb553b6701bccfcbc9ad94ee22c --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rLoHXD9dxUU_00042_97_166.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rLoHXD9dxUU_00042_97_166", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rLoHXD9dxUU_00042_97_166_fore", + "duration": 2.917 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rLoHXD9dxUU_00042_97_166_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rLoHXD9dxUU_00042_97_166_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rPvkdU5kkqA_00010_0_59.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rPvkdU5kkqA_00010_0_59.json new file mode 100644 index 0000000000000000000000000000000000000000..5cab7cd7c394c0f9d743699c13f0ccc9060a1196 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rPvkdU5kkqA_00010_0_59.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rPvkdU5kkqA_00010_0_59", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rPvkdU5kkqA_00010_0_59_fore", + "duration": 2.002 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rPvkdU5kkqA_00010_0_59_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rPvkdU5kkqA_00010_0_59_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 1.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rVq_-sLnHY4_00004_0_76.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rVq_-sLnHY4_00004_0_76.json new file mode 100644 index 0000000000000000000000000000000000000000..0fb8c06a4e57745d21e07e067673bf77af7cf514 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rVq_-sLnHY4_00004_0_76.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rVq_-sLnHY4_00004_0_76", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rVq_-sLnHY4_00004_0_76_fore", + "duration": 2.57 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rVq_-sLnHY4_00004_0_76_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rVq_-sLnHY4_00004_0_76_fore", + "rendered_video": "none", + "time_window": [ + 1.7, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_re5DQem4pGg_00001_0_68.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_re5DQem4pGg_00001_0_68.json new file mode 100644 index 0000000000000000000000000000000000000000..e644a0be5673ce5b2ce54ba23ce9027b1b4806e0 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_re5DQem4pGg_00001_0_68.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_re5DQem4pGg_00001_0_68", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/re5DQem4pGg_00001_0_68_fore", + "duration": 2.37 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/re5DQem4pGg_00001_0_68_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/re5DQem4pGg_00001_0_68_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rguBtTP7mw4_00022_0_406.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rguBtTP7mw4_00022_0_406.json new file mode 100644 index 0000000000000000000000000000000000000000..f0a2e09f8dbcde1515f893308309ef39741c267d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rguBtTP7mw4_00022_0_406.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rguBtTP7mw4_00022_0_406", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rguBtTP7mw4_00022_0_406_fore", + "duration": 6.817 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rguBtTP7mw4_00022_0_406_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rguBtTP7mw4_00022_0_406_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 6.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rz1r2f68qx8_00002_220_296.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rz1r2f68qx8_00002_220_296.json new file mode 100644 index 0000000000000000000000000000000000000000..30e025cca29e027ea72c7f22a1fd0b1c71f2c602 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_rz1r2f68qx8_00002_220_296.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_rz1r2f68qx8_00002_220_296", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/rz1r2f68qx8_00002_220_296_fore", + "duration": 3.175 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/rz1r2f68qx8_00002_220_296_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/rz1r2f68qx8_00002_220_296_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3L1T34Zrdo_00047_0_236.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3L1T34Zrdo_00047_0_236.json new file mode 100644 index 0000000000000000000000000000000000000000..51ca6bcf981576df0d2f31183feb2681e1fb34ba --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3L1T34Zrdo_00047_0_236.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_s3L1T34Zrdo_00047_0_236", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/s3L1T34Zrdo_00047_0_236_fore", + "duration": 7.913 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/s3L1T34Zrdo_00047_0_236_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/s3L1T34Zrdo_00047_0_236_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 7.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3PvzlrzuQ4_00017_83_141.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3PvzlrzuQ4_00017_83_141.json new file mode 100644 index 0000000000000000000000000000000000000000..51ea48069d3ab5a84cf5863ae5bfa4729748bd8a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s3PvzlrzuQ4_00017_83_141.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_s3PvzlrzuQ4_00017_83_141", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/s3PvzlrzuQ4_00017_83_141_fore", + "duration": 2.545 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/s3PvzlrzuQ4_00017_83_141_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/s3PvzlrzuQ4_00017_83_141_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s5x2LY5TRv8_00027_0_152.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s5x2LY5TRv8_00027_0_152.json new file mode 100644 index 0000000000000000000000000000000000000000..04008b65791cf07c8846783cbb7d3fc2a9be5468 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_s5x2LY5TRv8_00027_0_152.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_s5x2LY5TRv8_00027_0_152", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/s5x2LY5TRv8_00027_0_152_fore", + "duration": 6.341 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/s5x2LY5TRv8_00027_0_152_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/s5x2LY5TRv8_00027_0_152_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sCOmqB73HSE_00007_984_1132.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sCOmqB73HSE_00007_984_1132.json new file mode 100644 index 0000000000000000000000000000000000000000..8e1642462e174c86b565fa796f8684ddcaf6dfb3 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sCOmqB73HSE_00007_984_1132.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sCOmqB73HSE_00007_984_1132", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sCOmqB73HSE_00007_984_1132_fore", + "duration": 5.072 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sCOmqB73HSE_00007_984_1132_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sCOmqB73HSE_00007_984_1132_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 4.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sGr1HBSIic8_00022_242_384.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sGr1HBSIic8_00022_242_384.json new file mode 100644 index 0000000000000000000000000000000000000000..569f280fd937d1b2938c8f0c5af3d93ec971dfb8 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sGr1HBSIic8_00022_242_384.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sGr1HBSIic8_00022_242_384", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sGr1HBSIic8_00022_242_384_fore", + "duration": 4.738 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sGr1HBSIic8_00022_242_384_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sGr1HBSIic8_00022_242_384_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sTpqVab63mY_00025_0_64.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sTpqVab63mY_00025_0_64.json new file mode 100644 index 0000000000000000000000000000000000000000..205c5006443f0dd9dc7570bf29d62c05f2033162 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sTpqVab63mY_00025_0_64.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sTpqVab63mY_00025_0_64", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sTpqVab63mY_00025_0_64_fore", + "duration": 2.56 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sTpqVab63mY_00025_0_64_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sTpqVab63mY_00025_0_64_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sdRHReep4mw_00018_0_70.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sdRHReep4mw_00018_0_70.json new file mode 100644 index 0000000000000000000000000000000000000000..b5b2f7373da93688e2f0b2e1a72957357850b3fe --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sdRHReep4mw_00018_0_70.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sdRHReep4mw_00018_0_70", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sdRHReep4mw_00018_0_70_fore", + "duration": 2.961 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sdRHReep4mw_00018_0_70_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sdRHReep4mw_00018_0_70_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sfOi1AdSkjk_00011_0_127.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sfOi1AdSkjk_00011_0_127.json new file mode 100644 index 0000000000000000000000000000000000000000..6c1170bd3e3a7cb904cac9214dfacf88865a700d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sfOi1AdSkjk_00011_0_127.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sfOi1AdSkjk_00011_0_127", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sfOi1AdSkjk_00011_0_127_fore", + "duration": 5.339 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sfOi1AdSkjk_00011_0_127_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sfOi1AdSkjk_00011_0_127_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 5.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slCwOLXIeU4_00025_56_104.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slCwOLXIeU4_00025_56_104.json new file mode 100644 index 0000000000000000000000000000000000000000..84f02fdea20fb600d08524c1324de56f1b0f09a1 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slCwOLXIeU4_00025_56_104.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_slCwOLXIeU4_00025_56_104", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/slCwOLXIeU4_00025_56_104_fore", + "duration": 2.044 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/slCwOLXIeU4_00025_56_104_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/slCwOLXIeU4_00025_56_104_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slcf7Y3yEdU_00037_0_411.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slcf7Y3yEdU_00037_0_411.json new file mode 100644 index 0000000000000000000000000000000000000000..59d66c77531e8212e6349e8f5d886a6d573bfd34 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_slcf7Y3yEdU_00037_0_411.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_slcf7Y3yEdU_00037_0_411", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/slcf7Y3yEdU_00037_0_411_fore", + "duration": 8.233 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/slcf7Y3yEdU_00037_0_411_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/slcf7Y3yEdU_00037_0_411_fore", + "rendered_video": "none", + "time_window": [ + 5.5, + 8.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_svu306fZ5w8_00015_0_72.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_svu306fZ5w8_00015_0_72.json new file mode 100644 index 0000000000000000000000000000000000000000..36329586aa42b708d2809f02e678b785ee0ae5b8 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_svu306fZ5w8_00015_0_72.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_svu306fZ5w8_00015_0_72", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/svu306fZ5w8_00015_0_72_fore", + "duration": 2.96 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/svu306fZ5w8_00015_0_72_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/svu306fZ5w8_00015_0_72_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sxJh8fI35iw_00033_0_51.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sxJh8fI35iw_00033_0_51.json new file mode 100644 index 0000000000000000000000000000000000000000..3d53f40badddb9b0dde5d5ddd064a006147ce1cd --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_sxJh8fI35iw_00033_0_51.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_sxJh8fI35iw_00033_0_51", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/sxJh8fI35iw_00033_0_51_fore", + "duration": 2.169 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/sxJh8fI35iw_00033_0_51_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/sxJh8fI35iw_00033_0_51_fore", + "rendered_video": "none", + "time_window": [ + 0.7, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3JoimGNERA_00010_0_88.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3JoimGNERA_00010_0_88.json new file mode 100644 index 0000000000000000000000000000000000000000..5c9e7bef78b33f71dccaebe62e9baa6e36fb2fbe --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3JoimGNERA_00010_0_88.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_t3JoimGNERA_00010_0_88", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/t3JoimGNERA_00010_0_88_fore", + "duration": 3.003 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/t3JoimGNERA_00010_0_88_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/t3JoimGNERA_00010_0_88_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3ybHGZxjjQ_00014_0_208.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3ybHGZxjjQ_00014_0_208.json new file mode 100644 index 0000000000000000000000000000000000000000..f7165c3fc286e09b66352cb003fcf8681eac3e34 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t3ybHGZxjjQ_00014_0_208.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_t3ybHGZxjjQ_00014_0_208", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/t3ybHGZxjjQ_00014_0_208_fore", + "duration": 8.676 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/t3ybHGZxjjQ_00014_0_208_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/t3ybHGZxjjQ_00014_0_208_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 8.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t5wg4c0ScAc_00035_453_724.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t5wg4c0ScAc_00035_453_724.json new file mode 100644 index 0000000000000000000000000000000000000000..c5457ec91cc476ee00bed92d0f5a83320f5e91cd --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_t5wg4c0ScAc_00035_453_724.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_t5wg4c0ScAc_00035_453_724", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/t5wg4c0ScAc_00035_453_724_fore", + "duration": 5.44 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/t5wg4c0ScAc_00035_453_724_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/t5wg4c0ScAc_00035_453_724_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tFc5jBcitPY_00021_54_214.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tFc5jBcitPY_00021_54_214.json new file mode 100644 index 0000000000000000000000000000000000000000..d66bc628d02dc7ea975eb77865fd44016031f1b4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tFc5jBcitPY_00021_54_214.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tFc5jBcitPY_00021_54_214", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tFc5jBcitPY_00021_54_214_fore", + "duration": 6.4 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tFc5jBcitPY_00021_54_214_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tFc5jBcitPY_00021_54_214_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tHokJbz14nA_00024_0_240.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tHokJbz14nA_00024_0_240.json new file mode 100644 index 0000000000000000000000000000000000000000..a5a552a762766cebc250a28d2030410dd7677b59 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tHokJbz14nA_00024_0_240.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tHokJbz14nA_00024_0_240", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tHokJbz14nA_00024_0_240_fore", + "duration": 4.038 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tHokJbz14nA_00024_0_240_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tHokJbz14nA_00024_0_240_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tLnTZliGbAc_00072_0_147.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tLnTZliGbAc_00072_0_147.json new file mode 100644 index 0000000000000000000000000000000000000000..9f331278670be8d865dc6714e6b6714427274a58 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tLnTZliGbAc_00072_0_147.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tLnTZliGbAc_00072_0_147", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tLnTZliGbAc_00072_0_147_fore", + "duration": 4.939 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tLnTZliGbAc_00072_0_147_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tLnTZliGbAc_00072_0_147_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tMSHEtgk4xA_00009_74_168.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tMSHEtgk4xA_00009_74_168.json new file mode 100644 index 0000000000000000000000000000000000000000..6625a5588afc2cfcbbcb112e6d687c97d3b42891 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tMSHEtgk4xA_00009_74_168.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tMSHEtgk4xA_00009_74_168", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tMSHEtgk4xA_00009_74_168_fore", + "duration": 3.17 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tMSHEtgk4xA_00009_74_168_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tMSHEtgk4xA_00009_74_168_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tUjWaZMz_Yo_00014_0_63.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tUjWaZMz_Yo_00014_0_63.json new file mode 100644 index 0000000000000000000000000000000000000000..ce6978d08d5017972d3095dcee988364e05af61f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tUjWaZMz_Yo_00014_0_63.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tUjWaZMz_Yo_00014_0_63", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tUjWaZMz_Yo_00014_0_63_fore", + "duration": 2.64 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tUjWaZMz_Yo_00014_0_63_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tUjWaZMz_Yo_00014_0_63_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYi23Gsg6Is_00017_157_465.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYi23Gsg6Is_00017_157_465.json new file mode 100644 index 0000000000000000000000000000000000000000..67c27a0dc4fa9b2bf5be8c82cf37885656436847 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYi23Gsg6Is_00017_157_465.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tYi23Gsg6Is_00017_157_465", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tYi23Gsg6Is_00017_157_465_fore", + "duration": 10.378 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tYi23Gsg6Is_00017_157_465_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tYi23Gsg6Is_00017_157_465_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 10.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYnPjJD8Y0I_00072_231_303.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYnPjJD8Y0I_00072_231_303.json new file mode 100644 index 0000000000000000000000000000000000000000..0efa8c8d8a8ac19410992006aec089cfcfd6a07f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tYnPjJD8Y0I_00072_231_303.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tYnPjJD8Y0I_00072_231_303", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tYnPjJD8Y0I_00072_231_303_fore", + "duration": 2.96 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tYnPjJD8Y0I_00072_231_303_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tYnPjJD8Y0I_00072_231_303_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tcUJC5EVwRk_00035_0_63.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tcUJC5EVwRk_00035_0_63.json new file mode 100644 index 0000000000000000000000000000000000000000..59cf62702bdcec9b54a398c242004d491c7da771 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tcUJC5EVwRk_00035_0_63.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tcUJC5EVwRk_00035_0_63", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tcUJC5EVwRk_00035_0_63_fore", + "duration": 2.236 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tcUJC5EVwRk_00035_0_63_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tcUJC5EVwRk_00035_0_63_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tevz58dPHXw_00000_0_91.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tevz58dPHXw_00000_0_91.json new file mode 100644 index 0000000000000000000000000000000000000000..59337c0087f63dc74cb285ad4a1b210130418fad --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tevz58dPHXw_00000_0_91.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tevz58dPHXw_00000_0_91", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tevz58dPHXw_00000_0_91_fore", + "duration": 3.204 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tevz58dPHXw_00000_0_91_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tevz58dPHXw_00000_0_91_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tfXFKIQ10ro_00023_0_315.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tfXFKIQ10ro_00023_0_315.json new file mode 100644 index 0000000000000000000000000000000000000000..2362dd7e4fcd33a3c62496e0fc8e9dc425ffa0e7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tfXFKIQ10ro_00023_0_315.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tfXFKIQ10ro_00023_0_315", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tfXFKIQ10ro_00023_0_315_fore", + "duration": 5.281 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tfXFKIQ10ro_00023_0_315_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tfXFKIQ10ro_00023_0_315_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 5.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tkc8RnGP86w_00006_484_1006.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tkc8RnGP86w_00006_484_1006.json new file mode 100644 index 0000000000000000000000000000000000000000..57d70c5a59d44a25ebe98be7c5227edcd95990ee --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_tkc8RnGP86w_00006_484_1006.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_tkc8RnGP86w_00006_484_1006", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/tkc8RnGP86w_00006_484_1006_fore", + "duration": 17.434 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/tkc8RnGP86w_00006_484_1006_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/tkc8RnGP86w_00006_484_1006_fore", + "rendered_video": "none", + "time_window": [ + 7.0, + 17.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_todl3gWNdag_00004_0_118.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_todl3gWNdag_00004_0_118.json new file mode 100644 index 0000000000000000000000000000000000000000..5193ded189c9e1267c85c09100318aee9a98af53 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_todl3gWNdag_00004_0_118.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_todl3gWNdag_00004_0_118", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/todl3gWNdag_00004_0_118_fore", + "duration": 4.88 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/todl3gWNdag_00004_0_118_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/todl3gWNdag_00004_0_118_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u-27fhZvFrU_00038_0_210.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u-27fhZvFrU_00038_0_210.json new file mode 100644 index 0000000000000000000000000000000000000000..82be3217fb058a9c38e96e28fe8492a5df1c2b26 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u-27fhZvFrU_00038_0_210.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_u-27fhZvFrU_00038_0_210", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/u-27fhZvFrU_00038_0_210_fore", + "duration": 3.522 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/u-27fhZvFrU_00038_0_210_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/u-27fhZvFrU_00038_0_210_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u1PVDPE-hgE_00007_0_86.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u1PVDPE-hgE_00007_0_86.json new file mode 100644 index 0000000000000000000000000000000000000000..38014758df7f1e7f578648e579d88d7744be60ec --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_u1PVDPE-hgE_00007_0_86.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_u1PVDPE-hgE_00007_0_86", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/u1PVDPE-hgE_00007_0_86_fore", + "duration": 3.629 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/u1PVDPE-hgE_00007_0_86_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/u1PVDPE-hgE_00007_0_86_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uQpxQUE5XiA_00000_634_1022.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uQpxQUE5XiA_00000_634_1022.json new file mode 100644 index 0000000000000000000000000000000000000000..4b0672aeec2a18943684bcabc6c492241246fab7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uQpxQUE5XiA_00000_634_1022.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_uQpxQUE5XiA_00000_634_1022", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/uQpxQUE5XiA_00000_634_1022_fore", + "duration": 12.98 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/uQpxQUE5XiA_00000_634_1022_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/uQpxQUE5XiA_00000_634_1022_fore", + "rendered_video": "none", + "time_window": [ + 6.0, + 13.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uYBMmOeixow_00040_185_304.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uYBMmOeixow_00040_185_304.json new file mode 100644 index 0000000000000000000000000000000000000000..6dd644cb2f637e30cb7858d53a9c07506fb3cfb7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uYBMmOeixow_00040_185_304.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_uYBMmOeixow_00040_185_304", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/uYBMmOeixow_00040_185_304_fore", + "duration": 3.95 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/uYBMmOeixow_00040_185_304_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/uYBMmOeixow_00040_185_304_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uao897AZC08_00024_64_161.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uao897AZC08_00024_64_161.json new file mode 100644 index 0000000000000000000000000000000000000000..bb1607a65bf570cabf9c0e7991b7c8f92232aa8a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uao897AZC08_00024_64_161.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_uao897AZC08_00024_64_161", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/uao897AZC08_00024_64_161_fore", + "duration": 3.27 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/uao897AZC08_00024_64_161_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/uao897AZC08_00024_64_161_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 3.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uhuO0DFe0ZY_00033_0_114.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uhuO0DFe0ZY_00033_0_114.json new file mode 100644 index 0000000000000000000000000000000000000000..89ff6131937c6f0e746f548e87ea3d937e2f840f --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uhuO0DFe0ZY_00033_0_114.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_uhuO0DFe0ZY_00033_0_114", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/uhuO0DFe0ZY_00033_0_114_fore", + "duration": 4.6 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/uhuO0DFe0ZY_00033_0_114_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/uhuO0DFe0ZY_00033_0_114_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uieRXHzxCRQ_00041_144_324.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uieRXHzxCRQ_00041_144_324.json new file mode 100644 index 0000000000000000000000000000000000000000..b921b388dbfef87660d239069620162a6d4667f9 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_uieRXHzxCRQ_00041_144_324.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_uieRXHzxCRQ_00041_144_324", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/uieRXHzxCRQ_00041_144_324_fore", + "duration": 6.04 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/uieRXHzxCRQ_00041_144_324_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/uieRXHzxCRQ_00041_144_324_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ux467aEJkjo_00035_0_182.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ux467aEJkjo_00035_0_182.json new file mode 100644 index 0000000000000000000000000000000000000000..53142c883f869cbf43684c32c9a4a68970b682c4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_ux467aEJkjo_00035_0_182.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_ux467aEJkjo_00035_0_182", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/ux467aEJkjo_00035_0_182_fore", + "duration": 3.054 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/ux467aEJkjo_00035_0_182_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/ux467aEJkjo_00035_0_182_fore", + "rendered_video": "none", + "time_window": [ + 2.7, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v3J14lxJmHo_00021_0_178.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v3J14lxJmHo_00021_0_178.json new file mode 100644 index 0000000000000000000000000000000000000000..7babbfdb9f8c9ec5cc019c273c912e8486a24d52 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v3J14lxJmHo_00021_0_178.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_v3J14lxJmHo_00021_0_178", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/v3J14lxJmHo_00021_0_178_fore", + "duration": 2.997 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/v3J14lxJmHo_00021_0_178_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/v3J14lxJmHo_00021_0_178_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v5dFCq3ZZlo_00014_482_620.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v5dFCq3ZZlo_00014_482_620.json new file mode 100644 index 0000000000000000000000000000000000000000..5e2ef1d65b06801d73fbc59718b642aecb127831 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v5dFCq3ZZlo_00014_482_620.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_v5dFCq3ZZlo_00014_482_620", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/v5dFCq3ZZlo_00014_482_620_fore", + "duration": 2.321 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/v5dFCq3ZZlo_00014_482_620_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/v5dFCq3ZZlo_00014_482_620_fore", + "rendered_video": "none", + "time_window": [ + 1.3, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v81_5bBEakI_00085_0_247.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v81_5bBEakI_00085_0_247.json new file mode 100644 index 0000000000000000000000000000000000000000..e37d8d54460cdc0f90164584d3fad6782636579e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v81_5bBEakI_00085_0_247.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_v81_5bBEakI_00085_0_247", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/v81_5bBEakI_00085_0_247_fore", + "duration": 4.98 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/v81_5bBEakI_00085_0_247_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/v81_5bBEakI_00085_0_247_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 4.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v9m0cWjtP0A_00036_28_148.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v9m0cWjtP0A_00036_28_148.json new file mode 100644 index 0000000000000000000000000000000000000000..d26dfb685c38b57a7d446daae1cf5bee52c0725d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_v9m0cWjtP0A_00036_28_148.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_v9m0cWjtP0A_00036_28_148", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/v9m0cWjtP0A_00036_28_148_fore", + "duration": 4.007 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/v9m0cWjtP0A_00036_28_148_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/v9m0cWjtP0A_00036_28_148_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vKREmiZfeCs_00043_0_126.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vKREmiZfeCs_00043_0_126.json new file mode 100644 index 0000000000000000000000000000000000000000..821ced57b6cc8c73a303cf1e642632dca89c5255 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vKREmiZfeCs_00043_0_126.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vKREmiZfeCs_00043_0_126", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vKREmiZfeCs_00043_0_126_fore", + "duration": 5.16 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vKREmiZfeCs_00043_0_126_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vKREmiZfeCs_00043_0_126_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOBy-BKZIJc_00008_35_109.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOBy-BKZIJc_00008_35_109.json new file mode 100644 index 0000000000000000000000000000000000000000..92ab5915679442424dc4a9ce37c5e6a37b5109f1 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOBy-BKZIJc_00008_35_109.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vOBy-BKZIJc_00008_35_109", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vOBy-BKZIJc_00008_35_109_fore", + "duration": 2.503 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vOBy-BKZIJc_00008_35_109_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vOBy-BKZIJc_00008_35_109_fore", + "rendered_video": "none", + "time_window": [ + 1.6, + 2.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOzStTFIJ1g_00047_286_460.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOzStTFIJ1g_00047_286_460.json new file mode 100644 index 0000000000000000000000000000000000000000..915905ed2e690ae454294eb33a195e66504ca967 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vOzStTFIJ1g_00047_286_460.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vOzStTFIJ1g_00047_286_460", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vOzStTFIJ1g_00047_286_460_fore", + "duration": 2.92 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vOzStTFIJ1g_00047_286_460_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vOzStTFIJ1g_00047_286_460_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vXgqnnkqNok_00008_0_106.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vXgqnnkqNok_00008_0_106.json new file mode 100644 index 0000000000000000000000000000000000000000..500de9e5af8b9d581ddd07ff7e7369f677ae9832 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vXgqnnkqNok_00008_0_106.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vXgqnnkqNok_00008_0_106", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vXgqnnkqNok_00008_0_106_fore", + "duration": 4.463 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vXgqnnkqNok_00008_0_106_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vXgqnnkqNok_00008_0_106_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vbf2cZMFasU_00009_0_66.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vbf2cZMFasU_00009_0_66.json new file mode 100644 index 0000000000000000000000000000000000000000..5284376d24a7ceec3de30d39c558fef2b223e37d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vbf2cZMFasU_00009_0_66.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vbf2cZMFasU_00009_0_66", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vbf2cZMFasU_00009_0_66_fore", + "duration": 2.303 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vbf2cZMFasU_00009_0_66_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vbf2cZMFasU_00009_0_66_fore", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_voMV3Mhta1U_00016_0_106.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_voMV3Mhta1U_00016_0_106.json new file mode 100644 index 0000000000000000000000000000000000000000..58dc3a7fa27840395bdd8e95ccc9d96da31bdd50 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_voMV3Mhta1U_00016_0_106.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_voMV3Mhta1U_00016_0_106", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/voMV3Mhta1U_00016_0_106_fore", + "duration": 4.463 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/voMV3Mhta1U_00016_0_106_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/voMV3Mhta1U_00016_0_106_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 4.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vosAVclG2TI_00000_313_612.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vosAVclG2TI_00000_313_612.json new file mode 100644 index 0000000000000000000000000000000000000000..700d32e902f0c5a363b06a12b4f8b1aaed42c26b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_vosAVclG2TI_00000_313_612.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_vosAVclG2TI_00000_313_612", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/vosAVclG2TI_00000_313_612_fore", + "duration": 10.044 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/vosAVclG2TI_00000_313_612_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/vosAVclG2TI_00000_313_612_fore", + "rendered_video": "none", + "time_window": [ + 6.5, + 9.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wKH4hW8VQa8_00089_0_221.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wKH4hW8VQa8_00089_0_221.json new file mode 100644 index 0000000000000000000000000000000000000000..63ecb0770682c5bf26b8cc6c864a412a4c93b0b4 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wKH4hW8VQa8_00089_0_221.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wKH4hW8VQa8_00089_0_221", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wKH4hW8VQa8_00089_0_221_fore", + "duration": 7.408 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wKH4hW8VQa8_00089_0_221_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wKH4hW8VQa8_00089_0_221_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 7.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wadfKQg4Yt4_00024_0_140.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wadfKQg4Yt4_00024_0_140.json new file mode 100644 index 0000000000000000000000000000000000000000..4ac6d5cc411b591bfcce477ed400036306df7e20 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wadfKQg4Yt4_00024_0_140.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wadfKQg4Yt4_00024_0_140", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wadfKQg4Yt4_00024_0_140_fore", + "duration": 5.881 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wadfKQg4Yt4_00024_0_140_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wadfKQg4Yt4_00024_0_140_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 5.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wo1yMzZCMko_00002_59_146.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wo1yMzZCMko_00002_59_146.json new file mode 100644 index 0000000000000000000000000000000000000000..9d2b1504cb2d087a87a06574f939881dc64b1fda --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wo1yMzZCMko_00002_59_146.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wo1yMzZCMko_00002_59_146", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wo1yMzZCMko_00002_59_146_fore", + "duration": 3.003 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wo1yMzZCMko_00002_59_146_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wo1yMzZCMko_00002_59_146_fore", + "rendered_video": "none", + "time_window": [ + 1.7, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wpstnTGr73M_00001_269_432.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wpstnTGr73M_00001_269_432.json new file mode 100644 index 0000000000000000000000000000000000000000..2c4457875f4cab1338e3db72e1113b1b7ff09f6d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wpstnTGr73M_00001_269_432.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wpstnTGr73M_00001_269_432", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wpstnTGr73M_00001_269_432_fore", + "duration": 5.506 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wpstnTGr73M_00001_269_432_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wpstnTGr73M_00001_269_432_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 5.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wqrpka69v1w_00022_331_483.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wqrpka69v1w_00022_331_483.json new file mode 100644 index 0000000000000000000000000000000000000000..0772cbe092716fbe88ca6296950bd1467e2b0926 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wqrpka69v1w_00022_331_483.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wqrpka69v1w_00022_331_483", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wqrpka69v1w_00022_331_483_fore", + "duration": 5.172 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wqrpka69v1w_00022_331_483_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wqrpka69v1w_00022_331_483_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wrb-EABXqCY_00029_0_147.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wrb-EABXqCY_00029_0_147.json new file mode 100644 index 0000000000000000000000000000000000000000..e54d721624536c7af6a3c8570327b361667b318a --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wrb-EABXqCY_00029_0_147.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wrb-EABXqCY_00029_0_147", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wrb-EABXqCY_00029_0_147_fore", + "duration": 2.486 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wrb-EABXqCY_00029_0_147_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wrb-EABXqCY_00029_0_147_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wt3UQzkXcEA_00031_1197_1444.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wt3UQzkXcEA_00031_1197_1444.json new file mode 100644 index 0000000000000000000000000000000000000000..99aa81dbfe5eb22d13d794f3dc46a2b63e9bc77d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wt3UQzkXcEA_00031_1197_1444.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wt3UQzkXcEA_00031_1197_1444", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wt3UQzkXcEA_00031_1197_1444_fore", + "duration": 4.143 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wt3UQzkXcEA_00031_1197_1444_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wt3UQzkXcEA_00031_1197_1444_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wzX4jdw5u1k_00025_0_54.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wzX4jdw5u1k_00025_0_54.json new file mode 100644 index 0000000000000000000000000000000000000000..9e085eea1778e7f9f475cf2defce527dcac14dd6 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_wzX4jdw5u1k_00025_0_54.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_wzX4jdw5u1k_00025_0_54", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/wzX4jdw5u1k_00025_0_54_fore", + "duration": 2.294 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/wzX4jdw5u1k_00025_0_54_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/wzX4jdw5u1k_00025_0_54_fore", + "rendered_video": "none", + "time_window": [ + 1.8, + 2.3 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x3jLXMIdPXo_00010_142_271.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x3jLXMIdPXo_00010_142_271.json new file mode 100644 index 0000000000000000000000000000000000000000..dabd420f38975a8d0133e31fffdee76cf21f1aef --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x3jLXMIdPXo_00010_142_271.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_x3jLXMIdPXo_00010_142_271", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/x3jLXMIdPXo_00010_142_271_fore", + "duration": 4.483 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/x3jLXMIdPXo_00010_142_271_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/x3jLXMIdPXo_00010_142_271_fore", + "rendered_video": "none", + "time_window": [ + 3.5, + 4.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x7yA3xtWNCU_00020_0_105.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x7yA3xtWNCU_00020_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..0e24a7fc2bfe7e13ccc664b36664f84027203259 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x7yA3xtWNCU_00020_0_105.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_x7yA3xtWNCU_00020_0_105", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/x7yA3xtWNCU_00020_0_105_fore", + "duration": 4.422 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/x7yA3xtWNCU_00020_0_105_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/x7yA3xtWNCU_00020_0_105_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 4.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00007_0_194.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00007_0_194.json new file mode 100644 index 0000000000000000000000000000000000000000..4dd6c9f5db91a0dc36194674efd4df04dd1c44ec --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00007_0_194.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xD1ywMiX99s_00007_0_194", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xD1ywMiX99s_00007_0_194_fore", + "duration": 6.474 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xD1ywMiX99s_00007_0_194_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xD1ywMiX99s_00007_0_194_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00010_333_755.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00010_333_755.json new file mode 100644 index 0000000000000000000000000000000000000000..30b5be0c1dbec685d862a16378649063e99d5ae3 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xD1ywMiX99s_00010_333_755.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xD1ywMiX99s_00010_333_755", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xD1ywMiX99s_00010_333_755_fore", + "duration": 14.127 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xD1ywMiX99s_00010_333_755_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xD1ywMiX99s_00010_333_755_fore", + "rendered_video": "none", + "time_window": [ + 8.0, + 13.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xJojrk13FG4_00030_0_148.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xJojrk13FG4_00030_0_148.json new file mode 100644 index 0000000000000000000000000000000000000000..008373486fb104df63e64ce511148517868e7a89 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xJojrk13FG4_00030_0_148.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xJojrk13FG4_00030_0_148", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xJojrk13FG4_00030_0_148_fore", + "duration": 5.039 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xJojrk13FG4_00030_0_148_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xJojrk13FG4_00030_0_148_fore", + "rendered_video": "none", + "time_window": [ + 3.8, + 4.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xOfH2zAPyhw_00040_503_585.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xOfH2zAPyhw_00040_503_585.json new file mode 100644 index 0000000000000000000000000000000000000000..657a6cf6901cf28360c7844aeaa358981ac8b4ec --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xOfH2zAPyhw_00040_503_585.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xOfH2zAPyhw_00040_503_585", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xOfH2zAPyhw_00040_503_585_fore", + "duration": 3.423 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xOfH2zAPyhw_00040_503_585_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xOfH2zAPyhw_00040_503_585_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPYSM_X_gnQ_00007_0_110.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPYSM_X_gnQ_00007_0_110.json new file mode 100644 index 0000000000000000000000000000000000000000..1c34a0767b36c9fdf5820c6d63355f6d6ecfcf3d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPYSM_X_gnQ_00007_0_110.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xPYSM_X_gnQ_00007_0_110", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xPYSM_X_gnQ_00007_0_110_fore", + "duration": 3.704 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xPYSM_X_gnQ_00007_0_110_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xPYSM_X_gnQ_00007_0_110_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPtkAYDLzBQ_00005_310_395.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPtkAYDLzBQ_00005_310_395.json new file mode 100644 index 0000000000000000000000000000000000000000..3f6a17be3c0f3b05a9b31fdc3ed1ca9ec820b09d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xPtkAYDLzBQ_00005_310_395.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xPtkAYDLzBQ_00005_310_395", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xPtkAYDLzBQ_00005_310_395_fore", + "duration": 2.87 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xPtkAYDLzBQ_00005_310_395_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xPtkAYDLzBQ_00005_310_395_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xVH3jwjG8ok_00015_0_94.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xVH3jwjG8ok_00015_0_94.json new file mode 100644 index 0000000000000000000000000000000000000000..02aa3993620539a5fafa95d0a401a5711437db70 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xVH3jwjG8ok_00015_0_94.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xVH3jwjG8ok_00015_0_94", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xVH3jwjG8ok_00015_0_94_fore", + "duration": 3.17 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xVH3jwjG8ok_00015_0_94_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xVH3jwjG8ok_00015_0_94_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x_prt6sDXUQ_00059_174_390.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x_prt6sDXUQ_00059_174_390.json new file mode 100644 index 0000000000000000000000000000000000000000..144b46c4bc6e7af13cb54fb4a8d464c2cb737eb7 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_x_prt6sDXUQ_00059_174_390.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_x_prt6sDXUQ_00059_174_390", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/x_prt6sDXUQ_00059_174_390_fore", + "duration": 8.64 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/x_prt6sDXUQ_00059_174_390_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/x_prt6sDXUQ_00059_174_390_fore", + "rendered_video": "none", + "time_window": [ + 4.0, + 8.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xfhzzAmk9GI_00021_559_843.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xfhzzAmk9GI_00021_559_843.json new file mode 100644 index 0000000000000000000000000000000000000000..1d78df1801e8a69854036c9c74e79a6a1f450665 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_xfhzzAmk9GI_00021_559_843.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_xfhzzAmk9GI_00021_559_843", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/xfhzzAmk9GI_00021_559_843_fore", + "duration": 9.484 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/xfhzzAmk9GI_00021_559_843_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/xfhzzAmk9GI_00021_559_843_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 9.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zaJV54MKG-A_00012_0_149.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zaJV54MKG-A_00012_0_149.json new file mode 100644 index 0000000000000000000000000000000000000000..a70e4d0800a11dbbe2bb5d07227afc26b1e55615 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zaJV54MKG-A_00012_0_149.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zaJV54MKG-A_00012_0_149", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zaJV54MKG-A_00012_0_149_fore", + "duration": 6.0 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zaJV54MKG-A_00012_0_149_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zaJV54MKG-A_00012_0_149_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zhvbtWYA1v4_00015_0_471.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zhvbtWYA1v4_00015_0_471.json new file mode 100644 index 0000000000000000000000000000000000000000..4288c2298afc610940453640255c1686ede4a95e --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zhvbtWYA1v4_00015_0_471.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zhvbtWYA1v4_00015_0_471", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zhvbtWYA1v4_00015_0_471_fore", + "duration": 7.908 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zhvbtWYA1v4_00015_0_471_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zhvbtWYA1v4_00015_0_471_fore", + "rendered_video": "none", + "time_window": [ + 2.5, + 7.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00003_184_338.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00003_184_338.json new file mode 100644 index 0000000000000000000000000000000000000000..7b987f7dcf5dd87bab24554b534a0e250e1e6b4d --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00003_184_338.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00003_184_338", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zkPa9-Y_M7k_00003_184_338_fore", + "duration": 3.1 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zkPa9-Y_M7k_00003_184_338_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zkPa9-Y_M7k_00003_184_338_fore", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00064_0_146.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00064_0_146.json new file mode 100644 index 0000000000000000000000000000000000000000..c47ec82a9836ecae2f2c464b80e0b6bc0f95594b --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00064_0_146.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zkPa9-Y_M7k_00064_0_146", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zkPa9-Y_M7k_00064_0_146_fore", + "duration": 2.94 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zkPa9-Y_M7k_00064_0_146_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zkPa9-Y_M7k_00064_0_146_fore", + "rendered_video": "none", + "time_window": [ + 0.8, + 2.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoJXcEfo-Tg_00032_0_160.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoJXcEfo-Tg_00032_0_160.json new file mode 100644 index 0000000000000000000000000000000000000000..0e4c3e5c1fdaf4078f1e6320b82b3920ebce2e84 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoJXcEfo-Tg_00032_0_160.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zoJXcEfo-Tg_00032_0_160", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zoJXcEfo-Tg_00032_0_160_fore", + "duration": 5.341 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zoJXcEfo-Tg_00032_0_160_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zoJXcEfo-Tg_00032_0_160_fore", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoplKrZlnQQ_00012_0_90.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoplKrZlnQQ_00012_0_90.json new file mode 100644 index 0000000000000000000000000000000000000000..da4b7b2c19db48ef38fd3c8016cb855700564d92 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zoplKrZlnQQ_00012_0_90.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zoplKrZlnQQ_00012_0_90", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zoplKrZlnQQ_00012_0_90_fore", + "duration": 3.64 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zoplKrZlnQQ_00012_0_90_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zoplKrZlnQQ_00012_0_90_fore", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zpjWgmz7_MQ_00026_0_119.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zpjWgmz7_MQ_00026_0_119.json new file mode 100644 index 0000000000000000000000000000000000000000..a20f5a3cbfa4d918ebce8139cb50911e6d0efeed --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zpjWgmz7_MQ_00026_0_119.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zpjWgmz7_MQ_00026_0_119", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zpjWgmz7_MQ_00026_0_119_fore", + "duration": 5.005 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zpjWgmz7_MQ_00026_0_119_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zpjWgmz7_MQ_00026_0_119_fore", + "rendered_video": "none", + "time_window": [ + 1.2, + 4.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zta4suL4pxA_00001_454_1076.json b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zta4suL4pxA_00001_454_1076.json new file mode 100644 index 0000000000000000000000000000000000000000..67e464d599562d9965a509b1a4d22bfcffc3a5e3 --- /dev/null +++ b/Tasks/Video/VID_FORE_HUMAN/L1_VID_FORE_HUMAN_zta4suL4pxA_00001_454_1076.json @@ -0,0 +1,33 @@ +{ + "metadata": { + "task_id": "L1_VID_FORE_HUMAN_zta4suL4pxA_00001_454_1076", + "task_level": "Level1", + "task_type": "VID_FORE_HUMAN", + "original_video": "none", + "original_motion": "Data_Processed/motion_fore/zta4suL4pxA_00001_454_1076_fore", + "duration": 10.402 + }, + "input": { + "prompts": { + "general_instruction": "You are an advanced 3D motion generation model designed to synthesize physically plausible human movements. I will provide you with a video of a real human that ONLY shows the first half of a continuous action sequence. Your objective is to predict and generate the unseen second half of the motion sequence. You must ensure the generated future motion naturally continues from the action in the video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the predicted future motion.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_FORE_HUMAN/zta4suL4pxA_00001_454_1076_fore_input.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_fore/zta4suL4pxA_00001_454_1076_fore", + "rendered_video": "none", + "time_window": [ + 5.0, + 10.0 + ] + } +} \ No newline at end of file