diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0-NOMvTGx3Q_00034_0_80.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0-NOMvTGx3Q_00034_0_80.json new file mode 100644 index 0000000000000000000000000000000000000000..5d0af8c7db84520167570ae25b257d2cb3e5924c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0-NOMvTGx3Q_00034_0_80.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_0-NOMvTGx3Q_00034_0_80", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/0-NOMvTGx3Q_00034_0_80_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/0-NOMvTGx3Q_00034_0_80_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/0-NOMvTGx3Q_00034_0_80_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/0-NOMvTGx3Q_00034_0_80_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 1.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0CyBZEykTu4_00001_59_258.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0CyBZEykTu4_00001_59_258.json new file mode 100644 index 0000000000000000000000000000000000000000..2c88feaff4e03fe2c53ae1baa257273179ab7873 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0CyBZEykTu4_00001_59_258.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_0CyBZEykTu4_00001_59_258", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/0CyBZEykTu4_00001_59_258_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/0CyBZEykTu4_00001_59_258_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/0CyBZEykTu4_00001_59_258_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/0CyBZEykTu4_00001_59_258_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0h7AlrKLslw_00056_0_313.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0h7AlrKLslw_00056_0_313.json new file mode 100644 index 0000000000000000000000000000000000000000..796b8244894b14bd3712cf803d62cb47b1afe18e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_0h7AlrKLslw_00056_0_313.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_0h7AlrKLslw_00056_0_313", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/0h7AlrKLslw_00056_0_313_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/0h7AlrKLslw_00056_0_313_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/0h7AlrKLslw_00056_0_313_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/0h7AlrKLslw_00056_0_313_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_CL4WdaOqcz4_00000_1019_1180.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_CL4WdaOqcz4_00000_1019_1180.json new file mode 100644 index 0000000000000000000000000000000000000000..1ea3ca243d09ce2accb270f28cf2db899e570762 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_CL4WdaOqcz4_00000_1019_1180.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_CL4WdaOqcz4_00000_1019_1180", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/CL4WdaOqcz4_00000_1019_1180_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/CL4WdaOqcz4_00000_1019_1180_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/CL4WdaOqcz4_00000_1019_1180_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/CL4WdaOqcz4_00000_1019_1180_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DKMRRpmgitc_00031_0_398.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DKMRRpmgitc_00031_0_398.json new file mode 100644 index 0000000000000000000000000000000000000000..571d180647c406a614090f781a34f0d890f56d27 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DKMRRpmgitc_00031_0_398.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_DKMRRpmgitc_00031_0_398", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/DKMRRpmgitc_00031_0_398_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/DKMRRpmgitc_00031_0_398_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/DKMRRpmgitc_00031_0_398_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/DKMRRpmgitc_00031_0_398_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DTJjOysP-bI_00000_314_397.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DTJjOysP-bI_00000_314_397.json new file mode 100644 index 0000000000000000000000000000000000000000..8220f2e2ae6f65a76b030948bdd3cbf97c20289e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DTJjOysP-bI_00000_314_397.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_DTJjOysP-bI_00000_314_397", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/DTJjOysP-bI_00000_314_397_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/DTJjOysP-bI_00000_314_397_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/DTJjOysP-bI_00000_314_397_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/DTJjOysP-bI_00000_314_397_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DjYdSs-Uhsw_00032_0_96.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DjYdSs-Uhsw_00032_0_96.json new file mode 100644 index 0000000000000000000000000000000000000000..a54bd8f139867a1ea61940b1ac11a2ceb2b50911 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DjYdSs-Uhsw_00032_0_96.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_DjYdSs-Uhsw_00032_0_96", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/DjYdSs-Uhsw_00032_0_96_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/DjYdSs-Uhsw_00032_0_96_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/DjYdSs-Uhsw_00032_0_96_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/DjYdSs-Uhsw_00032_0_96_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DpVbta6xaUw_00001_647_1125.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DpVbta6xaUw_00001_647_1125.json new file mode 100644 index 0000000000000000000000000000000000000000..63624003f71061ca83fb0c029c833d91a5bde2de --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DpVbta6xaUw_00001_647_1125.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_DpVbta6xaUw_00001_647_1125", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/DpVbta6xaUw_00001_647_1125_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/DpVbta6xaUw_00001_647_1125_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/DpVbta6xaUw_00001_647_1125_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/DpVbta6xaUw_00001_647_1125_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 13.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DsU0Cm6PmzE_00000_0_104.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DsU0Cm6PmzE_00000_0_104.json new file mode 100644 index 0000000000000000000000000000000000000000..679a819857c8713af05660ade25f45b8f0796ba4 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_DsU0Cm6PmzE_00000_0_104.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_DsU0Cm6PmzE_00000_0_104", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/DsU0Cm6PmzE_00000_0_104_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/DsU0Cm6PmzE_00000_0_104_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/DsU0Cm6PmzE_00000_0_104_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/DsU0Cm6PmzE_00000_0_104_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EJuVEQtoi7E_00013_0_77.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EJuVEQtoi7E_00013_0_77.json new file mode 100644 index 0000000000000000000000000000000000000000..a4b8569e85f558e462c7dff9dd3f1ae5f301dbc1 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EJuVEQtoi7E_00013_0_77.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_EJuVEQtoi7E_00013_0_77", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/EJuVEQtoi7E_00013_0_77_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/EJuVEQtoi7E_00013_0_77_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/EJuVEQtoi7E_00013_0_77_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/EJuVEQtoi7E_00013_0_77_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ELV9T7pISm0_00001_0_145.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ELV9T7pISm0_00001_0_145.json new file mode 100644 index 0000000000000000000000000000000000000000..0b00bb295f2f5e7083329ee4372a5489ec4891b4 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ELV9T7pISm0_00001_0_145.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_ELV9T7pISm0_00001_0_145", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/ELV9T7pISm0_00001_0_145_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/ELV9T7pISm0_00001_0_145_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/ELV9T7pISm0_00001_0_145_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/ELV9T7pISm0_00001_0_145_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EiYsUTfJAw0_00058_0_189.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EiYsUTfJAw0_00058_0_189.json new file mode 100644 index 0000000000000000000000000000000000000000..fe7308a1a44ca0c02f15946c485664f3629e6fed --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_EiYsUTfJAw0_00058_0_189.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_EiYsUTfJAw0_00058_0_189", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/EiYsUTfJAw0_00058_0_189_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/EiYsUTfJAw0_00058_0_189_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/EiYsUTfJAw0_00058_0_189_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/EiYsUTfJAw0_00058_0_189_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_Eif-QsyWL7o_00069_0_234.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_Eif-QsyWL7o_00069_0_234.json new file mode 100644 index 0000000000000000000000000000000000000000..3a07163bcb0283bbb74df45c7eef8e4c91a0bea8 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_Eif-QsyWL7o_00069_0_234.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_Eif-QsyWL7o_00069_0_234", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/Eif-QsyWL7o_00069_0_234_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/Eif-QsyWL7o_00069_0_234_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/Eif-QsyWL7o_00069_0_234_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/Eif-QsyWL7o_00069_0_234_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ElIQTOQgWNw_00022_0_113.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ElIQTOQgWNw_00022_0_113.json new file mode 100644 index 0000000000000000000000000000000000000000..bcff1620f474530aedec6e32733d7eba71c30770 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ElIQTOQgWNw_00022_0_113.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_ElIQTOQgWNw_00022_0_113", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/ElIQTOQgWNw_00022_0_113_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/ElIQTOQgWNw_00022_0_113_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/ElIQTOQgWNw_00022_0_113_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/ElIQTOQgWNw_00022_0_113_inter", + "rendered_video": "none", + "time_window": [ + 0.7, + 1.9 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_FSxSuRaBLwQ_00063_0_225.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_FSxSuRaBLwQ_00063_0_225.json new file mode 100644 index 0000000000000000000000000000000000000000..cccc222663ddcc779dd7eee86f96b3dfaccf2349 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_FSxSuRaBLwQ_00063_0_225.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_FSxSuRaBLwQ_00063_0_225", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/FSxSuRaBLwQ_00063_0_225_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/FSxSuRaBLwQ_00063_0_225_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/FSxSuRaBLwQ_00063_0_225_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/FSxSuRaBLwQ_00063_0_225_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_HXBUzJigzrg_00028_0_960.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_HXBUzJigzrg_00028_0_960.json new file mode 100644 index 0000000000000000000000000000000000000000..74869269676024a0a6f7db5a4a3a6af2ee57fe4c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_HXBUzJigzrg_00028_0_960.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_HXBUzJigzrg_00028_0_960", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/HXBUzJigzrg_00028_0_960_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/HXBUzJigzrg_00028_0_960_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/HXBUzJigzrg_00028_0_960_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/HXBUzJigzrg_00028_0_960_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 9.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IHSaR35NJuM_00001_1062_1406.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IHSaR35NJuM_00001_1062_1406.json new file mode 100644 index 0000000000000000000000000000000000000000..0786c10e4f787f32f6de950639ca17ece0a2d525 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IHSaR35NJuM_00001_1062_1406.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_IHSaR35NJuM_00001_1062_1406", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/IHSaR35NJuM_00001_1062_1406_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/IHSaR35NJuM_00001_1062_1406_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/IHSaR35NJuM_00001_1062_1406_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/IHSaR35NJuM_00001_1062_1406_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 9.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IbuZuIWQ16k_00016_0_823.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IbuZuIWQ16k_00016_0_823.json new file mode 100644 index 0000000000000000000000000000000000000000..538f4ed5ace9c2d90f2e180fcd671fc4dddbfb8c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IbuZuIWQ16k_00016_0_823.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_IbuZuIWQ16k_00016_0_823", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/IbuZuIWQ16k_00016_0_823_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/IbuZuIWQ16k_00016_0_823_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/IbuZuIWQ16k_00016_0_823_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/IbuZuIWQ16k_00016_0_823_inter", + "rendered_video": "none", + "time_window": [ + 5.0, + 10.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IvUYGasf4-g_00053_0_113.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IvUYGasf4-g_00053_0_113.json new file mode 100644 index 0000000000000000000000000000000000000000..0910d7f1a043c016bbffea2dc7ceb4a7203c7360 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_IvUYGasf4-g_00053_0_113.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_IvUYGasf4-g_00053_0_113", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/IvUYGasf4-g_00053_0_113_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/IvUYGasf4-g_00053_0_113_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/IvUYGasf4-g_00053_0_113_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/IvUYGasf4-g_00053_0_113_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 2.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J27DCHIJauw_00027_0_82.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J27DCHIJauw_00027_0_82.json new file mode 100644 index 0000000000000000000000000000000000000000..b6ae268129069da156ed50c496645ba219522a77 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J27DCHIJauw_00027_0_82.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_J27DCHIJauw_00027_0_82", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/J27DCHIJauw_00027_0_82_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/J27DCHIJauw_00027_0_82_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/J27DCHIJauw_00027_0_82_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/J27DCHIJauw_00027_0_82_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JY1mxnXCeF8_00003_511_598.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JY1mxnXCeF8_00003_511_598.json new file mode 100644 index 0000000000000000000000000000000000000000..af812270b517e4c26f9c4f4242e44c01b8138873 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JY1mxnXCeF8_00003_511_598.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_JY1mxnXCeF8_00003_511_598", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/JY1mxnXCeF8_00003_511_598_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/JY1mxnXCeF8_00003_511_598_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/JY1mxnXCeF8_00003_511_598_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/JY1mxnXCeF8_00003_511_598_inter", + "rendered_video": "none", + "time_window": [ + 0.7, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JYL3s66-0W4_00056_288_374.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JYL3s66-0W4_00056_288_374.json new file mode 100644 index 0000000000000000000000000000000000000000..b455b1451c62a583c7a32bf64e62e222eb10e0ee --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_JYL3s66-0W4_00056_288_374.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_JYL3s66-0W4_00056_288_374", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/JYL3s66-0W4_00056_288_374_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/JYL3s66-0W4_00056_288_374_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/JYL3s66-0W4_00056_288_374_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/JYL3s66-0W4_00056_288_374_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J_4NWbKHP0I_00012_0_82.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J_4NWbKHP0I_00012_0_82.json new file mode 100644 index 0000000000000000000000000000000000000000..d8f466296afa5ff6e38b88659a6f6c2547e62de5 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_J_4NWbKHP0I_00012_0_82.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_J_4NWbKHP0I_00012_0_82", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/J_4NWbKHP0I_00012_0_82_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/J_4NWbKHP0I_00012_0_82_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/J_4NWbKHP0I_00012_0_82_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/J_4NWbKHP0I_00012_0_82_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_KRheAHqm4SU_00001_423_720.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_KRheAHqm4SU_00001_423_720.json new file mode 100644 index 0000000000000000000000000000000000000000..98a722de62bfc9dd5587b9fbc67fc512d7c650ca --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_KRheAHqm4SU_00001_423_720.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_KRheAHqm4SU_00001_423_720", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/KRheAHqm4SU_00001_423_720_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/KRheAHqm4SU_00001_423_720_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/KRheAHqm4SU_00001_423_720_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/KRheAHqm4SU_00001_423_720_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_L1rESfCgUZo_00010_0_117.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_L1rESfCgUZo_00010_0_117.json new file mode 100644 index 0000000000000000000000000000000000000000..11726a5208cb1219ca770748b8b2ede65077fe3d --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_L1rESfCgUZo_00010_0_117.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_L1rESfCgUZo_00010_0_117", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/L1rESfCgUZo_00010_0_117_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/L1rESfCgUZo_00010_0_117_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/L1rESfCgUZo_00010_0_117_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/L1rESfCgUZo_00010_0_117_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LALTJJyO2js_00000_0_66.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LALTJJyO2js_00000_0_66.json new file mode 100644 index 0000000000000000000000000000000000000000..9139d95e26f76b325b43d0ad11c85f1d0c003909 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LALTJJyO2js_00000_0_66.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_LALTJJyO2js_00000_0_66", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/LALTJJyO2js_00000_0_66_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/LALTJJyO2js_00000_0_66_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/LALTJJyO2js_00000_0_66_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/LALTJJyO2js_00000_0_66_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 1.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LDM9aXVAoOw_00002_0_694.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LDM9aXVAoOw_00002_0_694.json new file mode 100644 index 0000000000000000000000000000000000000000..a37c9c2ec3a13c25b0370da6605e51723dff3f98 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LDM9aXVAoOw_00002_0_694.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_LDM9aXVAoOw_00002_0_694", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/LDM9aXVAoOw_00002_0_694_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/LDM9aXVAoOw_00002_0_694_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/LDM9aXVAoOw_00002_0_694_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/LDM9aXVAoOw_00002_0_694_inter", + "rendered_video": "none", + "time_window": [ + 8.0, + 12.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LNvgFIUv9hI_00023_0_59.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LNvgFIUv9hI_00023_0_59.json new file mode 100644 index 0000000000000000000000000000000000000000..5fc0788fed87585a2a43a347e82696c0f71008c6 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LNvgFIUv9hI_00023_0_59.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_LNvgFIUv9hI_00023_0_59", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/LNvgFIUv9hI_00023_0_59_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/LNvgFIUv9hI_00023_0_59_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/LNvgFIUv9hI_00023_0_59_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/LNvgFIUv9hI_00023_0_59_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LOWyPvH0210_00002_0_103.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LOWyPvH0210_00002_0_103.json new file mode 100644 index 0000000000000000000000000000000000000000..443678779a8f7efbcff0b207861410a19022d782 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LOWyPvH0210_00002_0_103.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_LOWyPvH0210_00002_0_103", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/LOWyPvH0210_00002_0_103_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/LOWyPvH0210_00002_0_103_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/LOWyPvH0210_00002_0_103_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/LOWyPvH0210_00002_0_103_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LX5G4VkRH3o_00035_0_74.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LX5G4VkRH3o_00035_0_74.json new file mode 100644 index 0000000000000000000000000000000000000000..e6efec996b90967908f74fb97ed4cc3f9e475f5d --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_LX5G4VkRH3o_00035_0_74.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_LX5G4VkRH3o_00035_0_74", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/LX5G4VkRH3o_00035_0_74_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/LX5G4VkRH3o_00035_0_74_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/LX5G4VkRH3o_00035_0_74_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/LX5G4VkRH3o_00035_0_74_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_M3oVDmDIHek_00095_0_134.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_M3oVDmDIHek_00095_0_134.json new file mode 100644 index 0000000000000000000000000000000000000000..fc22fab26b71cf6e782ee6c596e839575540e1f5 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_M3oVDmDIHek_00095_0_134.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_M3oVDmDIHek_00095_0_134", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/M3oVDmDIHek_00095_0_134_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/M3oVDmDIHek_00095_0_134_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/M3oVDmDIHek_00095_0_134_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/M3oVDmDIHek_00095_0_134_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__1tzdCdC4UY_00005_77_203.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__1tzdCdC4UY_00005_77_203.json new file mode 100644 index 0000000000000000000000000000000000000000..af0004e5d8d1104ccda675d60ebffa473d1f7388 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__1tzdCdC4UY_00005_77_203.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL__1tzdCdC4UY_00005_77_203", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/_1tzdCdC4UY_00005_77_203_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/_1tzdCdC4UY_00005_77_203_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/_1tzdCdC4UY_00005_77_203_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/_1tzdCdC4UY_00005_77_203_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__MXhQXP5hEI_00002_0_662.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__MXhQXP5hEI_00002_0_662.json new file mode 100644 index 0000000000000000000000000000000000000000..9910307c4cca66d2627e990d4beb3d0329925172 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__MXhQXP5hEI_00002_0_662.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL__MXhQXP5hEI_00002_0_662", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/_MXhQXP5hEI_00002_0_662_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/_MXhQXP5hEI_00002_0_662_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/_MXhQXP5hEI_00002_0_662_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/_MXhQXP5hEI_00002_0_662_inter", + "rendered_video": "none", + "time_window": [ + 3.5, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__eyx3-8oIr0_00009_0_234.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__eyx3-8oIr0_00009_0_234.json new file mode 100644 index 0000000000000000000000000000000000000000..34f95d20db08ff4ced0e9d23685f4e02d139f0ed --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__eyx3-8oIr0_00009_0_234.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL__eyx3-8oIr0_00009_0_234", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/_eyx3-8oIr0_00009_0_234_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/_eyx3-8oIr0_00009_0_234_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/_eyx3-8oIr0_00009_0_234_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/_eyx3-8oIr0_00009_0_234_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__fIznC6ZIX0_00032_0_67.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__fIznC6ZIX0_00032_0_67.json new file mode 100644 index 0000000000000000000000000000000000000000..4a7ca3c9f341a2ee0cbeee25a23e4851aa6e0b45 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__fIznC6ZIX0_00032_0_67.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL__fIznC6ZIX0_00032_0_67", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/_fIznC6ZIX0_00032_0_67_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/_fIznC6ZIX0_00032_0_67_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/_fIznC6ZIX0_00032_0_67_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/_fIznC6ZIX0_00032_0_67_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.4 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__oYIT7I3avY_00018_947_1296.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__oYIT7I3avY_00018_947_1296.json new file mode 100644 index 0000000000000000000000000000000000000000..7acff3bd2cb6655e4a62aff3f69264e0e91ade62 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL__oYIT7I3avY_00018_947_1296.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL__oYIT7I3avY_00018_947_1296", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/_oYIT7I3avY_00018_947_1296_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/_oYIT7I3avY_00018_947_1296_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/_oYIT7I3avY_00018_947_1296_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/_oYIT7I3avY_00018_947_1296_inter", + "rendered_video": "none", + "time_window": [ + 3.5, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aDuB1arf8aE_00030_0_111.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aDuB1arf8aE_00030_0_111.json new file mode 100644 index 0000000000000000000000000000000000000000..caa88228313cf09128be25e9983ec0acf01f51af --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aDuB1arf8aE_00030_0_111.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_aDuB1arf8aE_00030_0_111", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/aDuB1arf8aE_00030_0_111_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/aDuB1arf8aE_00030_0_111_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/aDuB1arf8aE_00030_0_111_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/aDuB1arf8aE_00030_0_111_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aEFB3NGhx9E_00034_0_146.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aEFB3NGhx9E_00034_0_146.json new file mode 100644 index 0000000000000000000000000000000000000000..5794e1284f30b1cae701ef65f0b0c6e310dc90e1 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_aEFB3NGhx9E_00034_0_146.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_aEFB3NGhx9E_00034_0_146", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/aEFB3NGhx9E_00034_0_146_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/aEFB3NGhx9E_00034_0_146_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/aEFB3NGhx9E_00034_0_146_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/aEFB3NGhx9E_00034_0_146_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ae4jA0iRvtg_00030_0_120.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ae4jA0iRvtg_00030_0_120.json new file mode 100644 index 0000000000000000000000000000000000000000..1d4adda8b9c242b9a80329df43c3c9d5a44c9d34 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ae4jA0iRvtg_00030_0_120.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_ae4jA0iRvtg_00030_0_120", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/ae4jA0iRvtg_00030_0_120_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/ae4jA0iRvtg_00030_0_120_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/ae4jA0iRvtg_00030_0_120_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/ae4jA0iRvtg_00030_0_120_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_b9-gJuDgKnU_00012_0_86.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_b9-gJuDgKnU_00012_0_86.json new file mode 100644 index 0000000000000000000000000000000000000000..7390658d6f5476b457961f6e015ed043f13e94be --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_b9-gJuDgKnU_00012_0_86.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_b9-gJuDgKnU_00012_0_86", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/b9-gJuDgKnU_00012_0_86_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/b9-gJuDgKnU_00012_0_86_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/b9-gJuDgKnU_00012_0_86_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/b9-gJuDgKnU_00012_0_86_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bFX5Jd2bHS8_00001_0_135.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bFX5Jd2bHS8_00001_0_135.json new file mode 100644 index 0000000000000000000000000000000000000000..0a1aaff41216ae0f5dfb5df6a786d815f0388f05 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bFX5Jd2bHS8_00001_0_135.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_bFX5Jd2bHS8_00001_0_135", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/bFX5Jd2bHS8_00001_0_135_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/bFX5Jd2bHS8_00001_0_135_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/bFX5Jd2bHS8_00001_0_135_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/bFX5Jd2bHS8_00001_0_135_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bN-o8SRUdC4_00010_67_186.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bN-o8SRUdC4_00010_67_186.json new file mode 100644 index 0000000000000000000000000000000000000000..70741881611f275e5a27179447fbd1fa04caa451 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bN-o8SRUdC4_00010_67_186.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_bN-o8SRUdC4_00010_67_186", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/bN-o8SRUdC4_00010_67_186_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/bN-o8SRUdC4_00010_67_186_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/bN-o8SRUdC4_00010_67_186_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/bN-o8SRUdC4_00010_67_186_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bS5ve8rG-Ik_00052_0_93.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bS5ve8rG-Ik_00052_0_93.json new file mode 100644 index 0000000000000000000000000000000000000000..b6c5bc47647b76487176eddcf69dd06260920b01 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bS5ve8rG-Ik_00052_0_93.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_bS5ve8rG-Ik_00052_0_93", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/bS5ve8rG-Ik_00052_0_93_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/bS5ve8rG-Ik_00052_0_93_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/bS5ve8rG-Ik_00052_0_93_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/bS5ve8rG-Ik_00052_0_93_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bwr6x5xT9nE_00001_120_711.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bwr6x5xT9nE_00001_120_711.json new file mode 100644 index 0000000000000000000000000000000000000000..89911bdb006dd3b2cd9c3cc26b761e4151c9bc02 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_bwr6x5xT9nE_00001_120_711.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_bwr6x5xT9nE_00001_120_711", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/bwr6x5xT9nE_00001_120_711_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/bwr6x5xT9nE_00001_120_711_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/bwr6x5xT9nE_00001_120_711_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/bwr6x5xT9nE_00001_120_711_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_c03xSP9PzyA_00044_0_101.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_c03xSP9PzyA_00044_0_101.json new file mode 100644 index 0000000000000000000000000000000000000000..4186c94307bc6c28df6584a34a2ee214deabac86 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_c03xSP9PzyA_00044_0_101.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_c03xSP9PzyA_00044_0_101", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/c03xSP9PzyA_00044_0_101_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/c03xSP9PzyA_00044_0_101_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/c03xSP9PzyA_00044_0_101_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/c03xSP9PzyA_00044_0_101_inter", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cAEYHc5dzfs_00003_297_392.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cAEYHc5dzfs_00003_297_392.json new file mode 100644 index 0000000000000000000000000000000000000000..2eb9787a3c13a69ad7434f034ccdcdbe48ce84fe --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cAEYHc5dzfs_00003_297_392.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_cAEYHc5dzfs_00003_297_392", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/cAEYHc5dzfs_00003_297_392_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/cAEYHc5dzfs_00003_297_392_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/cAEYHc5dzfs_00003_297_392_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/cAEYHc5dzfs_00003_297_392_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cEdo6STidC0_00023_361_458.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cEdo6STidC0_00023_361_458.json new file mode 100644 index 0000000000000000000000000000000000000000..c7f943638cd8d462c238fb96b69637c8489c1415 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cEdo6STidC0_00023_361_458.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_cEdo6STidC0_00023_361_458", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/cEdo6STidC0_00023_361_458_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/cEdo6STidC0_00023_361_458_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/cEdo6STidC0_00023_361_458_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/cEdo6STidC0_00023_361_458_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cUUQtoBnAJU_00000_2062_2222.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cUUQtoBnAJU_00000_2062_2222.json new file mode 100644 index 0000000000000000000000000000000000000000..958e990bef74944c11ed9c65dafce869922058d5 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_cUUQtoBnAJU_00000_2062_2222.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_cUUQtoBnAJU_00000_2062_2222", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/cUUQtoBnAJU_00000_2062_2222_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/cUUQtoBnAJU_00000_2062_2222_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/cUUQtoBnAJU_00000_2062_2222_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/cUUQtoBnAJU_00000_2062_2222_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_d3uu7C26flM_00001_0_83.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_d3uu7C26flM_00001_0_83.json new file mode 100644 index 0000000000000000000000000000000000000000..43097e1728d4f65c60a31c3b0bd4cd5d23aba639 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_d3uu7C26flM_00001_0_83.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_d3uu7C26flM_00001_0_83", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/d3uu7C26flM_00001_0_83_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/d3uu7C26flM_00001_0_83_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/d3uu7C26flM_00001_0_83_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/d3uu7C26flM_00001_0_83_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00045_0_209.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00045_0_209.json new file mode 100644 index 0000000000000000000000000000000000000000..c52838c1e7bfae23764470a2d4f3eda8936a8605 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00045_0_209.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_htSmYNfT8QA_00045_0_209", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/htSmYNfT8QA_00045_0_209_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/htSmYNfT8QA_00045_0_209_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/htSmYNfT8QA_00045_0_209_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/htSmYNfT8QA_00045_0_209_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00054_0_195.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00054_0_195.json new file mode 100644 index 0000000000000000000000000000000000000000..130923eab13d408aa5d66e0f2fe9e9ce07c56a3c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_htSmYNfT8QA_00054_0_195.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_htSmYNfT8QA_00054_0_195", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/htSmYNfT8QA_00054_0_195_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/htSmYNfT8QA_00054_0_195_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/htSmYNfT8QA_00054_0_195_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/htSmYNfT8QA_00054_0_195_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_httbW7TujP0_00057_0_170.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_httbW7TujP0_00057_0_170.json new file mode 100644 index 0000000000000000000000000000000000000000..4f68553bb54f172c31cee252ab3a0ea97ebb699b --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_httbW7TujP0_00057_0_170.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_httbW7TujP0_00057_0_170", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/httbW7TujP0_00057_0_170_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/httbW7TujP0_00057_0_170_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/httbW7TujP0_00057_0_170_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/httbW7TujP0_00057_0_170_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_iIVJN0Yz1Y0_00027_0_84.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_iIVJN0Yz1Y0_00027_0_84.json new file mode 100644 index 0000000000000000000000000000000000000000..e104537f087e632feadeaf09d35adc0621de36bc --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_iIVJN0Yz1Y0_00027_0_84.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_iIVJN0Yz1Y0_00027_0_84", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/iIVJN0Yz1Y0_00027_0_84_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/iIVJN0Yz1Y0_00027_0_84_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/iIVJN0Yz1Y0_00027_0_84_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/iIVJN0Yz1Y0_00027_0_84_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ish8GggzGsA_00006_0_109.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ish8GggzGsA_00006_0_109.json new file mode 100644 index 0000000000000000000000000000000000000000..ee38b000b27923f3ce8f844b0c3d9b4c04b6360e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ish8GggzGsA_00006_0_109.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_ish8GggzGsA_00006_0_109", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/ish8GggzGsA_00006_0_109_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/ish8GggzGsA_00006_0_109_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/ish8GggzGsA_00006_0_109_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/ish8GggzGsA_00006_0_109_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 2.1 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jD8dLyj_7-U_00011_0_95.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jD8dLyj_7-U_00011_0_95.json new file mode 100644 index 0000000000000000000000000000000000000000..3750742a7dbaa7e86e88c90fd2fad4c52b1eee1a --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jD8dLyj_7-U_00011_0_95.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_jD8dLyj_7-U_00011_0_95", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/jD8dLyj_7-U_00011_0_95_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/jD8dLyj_7-U_00011_0_95_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/jD8dLyj_7-U_00011_0_95_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/jD8dLyj_7-U_00011_0_95_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_je-1dG6dBbM_00008_0_114.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_je-1dG6dBbM_00008_0_114.json new file mode 100644 index 0000000000000000000000000000000000000000..2b30ce823c5e54f80fe0da3b3ebfa1849666b3ee --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_je-1dG6dBbM_00008_0_114.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_je-1dG6dBbM_00008_0_114", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/je-1dG6dBbM_00008_0_114_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/je-1dG6dBbM_00008_0_114_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/je-1dG6dBbM_00008_0_114_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/je-1dG6dBbM_00008_0_114_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jfxw2B0uBX0_00031_293_436.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jfxw2B0uBX0_00031_293_436.json new file mode 100644 index 0000000000000000000000000000000000000000..4755d6fc086eb1966559f3653ff86863e4d1354a --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jfxw2B0uBX0_00031_293_436.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_jfxw2B0uBX0_00031_293_436", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/jfxw2B0uBX0_00031_293_436_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/jfxw2B0uBX0_00031_293_436_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/jfxw2B0uBX0_00031_293_436_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/jfxw2B0uBX0_00031_293_436_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jz8L6htl0yA_00038_311_415.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jz8L6htl0yA_00038_311_415.json new file mode 100644 index 0000000000000000000000000000000000000000..86796eefc228825964d0c0c7d745d05dc1498d90 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_jz8L6htl0yA_00038_311_415.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_jz8L6htl0yA_00038_311_415", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/jz8L6htl0yA_00038_311_415_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/jz8L6htl0yA_00038_311_415_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/jz8L6htl0yA_00038_311_415_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/jz8L6htl0yA_00038_311_415_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_k0X8Mz1ksEg_00013_0_154.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_k0X8Mz1ksEg_00013_0_154.json new file mode 100644 index 0000000000000000000000000000000000000000..ad9b994c5bfa878d5e4f0be063cbce07ec983b6b --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_k0X8Mz1ksEg_00013_0_154.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_k0X8Mz1ksEg_00013_0_154", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/k0X8Mz1ksEg_00013_0_154_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/k0X8Mz1ksEg_00013_0_154_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/k0X8Mz1ksEg_00013_0_154_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/k0X8Mz1ksEg_00013_0_154_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHQ8ib3EQb8_00014_0_556.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHQ8ib3EQb8_00014_0_556.json new file mode 100644 index 0000000000000000000000000000000000000000..0b4b0838d724c7c1a4c04d7104772c562f78349f --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHQ8ib3EQb8_00014_0_556.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_kHQ8ib3EQb8_00014_0_556", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/kHQ8ib3EQb8_00014_0_556_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/kHQ8ib3EQb8_00014_0_556_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/kHQ8ib3EQb8_00014_0_556_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/kHQ8ib3EQb8_00014_0_556_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 10.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHU2bS89gVI_00010_178_246.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHU2bS89gVI_00010_178_246.json new file mode 100644 index 0000000000000000000000000000000000000000..2d23355715d1b5af81c63c10ce53626359f8c45a --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kHU2bS89gVI_00010_178_246.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_kHU2bS89gVI_00010_178_246", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/kHU2bS89gVI_00010_178_246_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/kHU2bS89gVI_00010_178_246_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/kHU2bS89gVI_00010_178_246_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/kHU2bS89gVI_00010_178_246_inter", + "rendered_video": "none", + "time_window": [ + 0.9, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kvfNBJXHugg_00048_549_824.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kvfNBJXHugg_00048_549_824.json new file mode 100644 index 0000000000000000000000000000000000000000..08438215427feaeba9e37d848cee3615aa6e53e5 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_kvfNBJXHugg_00048_549_824.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_kvfNBJXHugg_00048_549_824", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/kvfNBJXHugg_00048_549_824_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/kvfNBJXHugg_00048_549_824_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/kvfNBJXHugg_00048_549_824_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/kvfNBJXHugg_00048_549_824_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_qqANXkZiVoU_00027_0_254.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_qqANXkZiVoU_00027_0_254.json new file mode 100644 index 0000000000000000000000000000000000000000..ad6e46e422ae5006fa5160864f2f24455992565c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_qqANXkZiVoU_00027_0_254.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_qqANXkZiVoU_00027_0_254", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/qqANXkZiVoU_00027_0_254_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/qqANXkZiVoU_00027_0_254_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/qqANXkZiVoU_00027_0_254_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/qqANXkZiVoU_00027_0_254_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rEtmuByf-0E_00006_1922_2189.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rEtmuByf-0E_00006_1922_2189.json new file mode 100644 index 0000000000000000000000000000000000000000..d7ebda105e80a5e4052cf3070b14e32bbdbd08af --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rEtmuByf-0E_00006_1922_2189.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_rEtmuByf-0E_00006_1922_2189", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/rEtmuByf-0E_00006_1922_2189_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/rEtmuByf-0E_00006_1922_2189_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/rEtmuByf-0E_00006_1922_2189_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/rEtmuByf-0E_00006_1922_2189_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rGsF1notX2A_00020_0_88.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rGsF1notX2A_00020_0_88.json new file mode 100644 index 0000000000000000000000000000000000000000..a4d00ef23e83e7acf773ddecdb7f31679b2a3781 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rGsF1notX2A_00020_0_88.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_rGsF1notX2A_00020_0_88", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/rGsF1notX2A_00020_0_88_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/rGsF1notX2A_00020_0_88_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/rGsF1notX2A_00020_0_88_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/rGsF1notX2A_00020_0_88_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rLoHXD9dxUU_00042_97_166.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rLoHXD9dxUU_00042_97_166.json new file mode 100644 index 0000000000000000000000000000000000000000..b6cfed99f5ee199d09b1b5b513ac70c2a9504d9b --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rLoHXD9dxUU_00042_97_166.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_rLoHXD9dxUU_00042_97_166", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/rLoHXD9dxUU_00042_97_166_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/rLoHXD9dxUU_00042_97_166_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/rLoHXD9dxUU_00042_97_166_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/rLoHXD9dxUU_00042_97_166_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rguBtTP7mw4_00022_0_406.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rguBtTP7mw4_00022_0_406.json new file mode 100644 index 0000000000000000000000000000000000000000..55c1ee52051e1f1e2d89eb2a65e099ccd4119af2 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rguBtTP7mw4_00022_0_406.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_rguBtTP7mw4_00022_0_406", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/rguBtTP7mw4_00022_0_406_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/rguBtTP7mw4_00022_0_406_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/rguBtTP7mw4_00022_0_406_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/rguBtTP7mw4_00022_0_406_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rz1r2f68qx8_00002_220_296.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rz1r2f68qx8_00002_220_296.json new file mode 100644 index 0000000000000000000000000000000000000000..6b9dc60a5705cdb0844f46035f57ba710899e6aa --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_rz1r2f68qx8_00002_220_296.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_rz1r2f68qx8_00002_220_296", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/rz1r2f68qx8_00002_220_296_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/rz1r2f68qx8_00002_220_296_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/rz1r2f68qx8_00002_220_296_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/rz1r2f68qx8_00002_220_296_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3L1T34Zrdo_00047_0_236.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3L1T34Zrdo_00047_0_236.json new file mode 100644 index 0000000000000000000000000000000000000000..ff4977c3e748ac23b250d38af80ec689e19cef84 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3L1T34Zrdo_00047_0_236.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_s3L1T34Zrdo_00047_0_236", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/s3L1T34Zrdo_00047_0_236_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/s3L1T34Zrdo_00047_0_236_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/s3L1T34Zrdo_00047_0_236_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/s3L1T34Zrdo_00047_0_236_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3PvzlrzuQ4_00017_83_141.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3PvzlrzuQ4_00017_83_141.json new file mode 100644 index 0000000000000000000000000000000000000000..64a68b890af99ba381632b8bdec79822bcdd147c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s3PvzlrzuQ4_00017_83_141.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_s3PvzlrzuQ4_00017_83_141", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/s3PvzlrzuQ4_00017_83_141_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/s3PvzlrzuQ4_00017_83_141_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/s3PvzlrzuQ4_00017_83_141_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/s3PvzlrzuQ4_00017_83_141_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s5x2LY5TRv8_00027_0_152.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s5x2LY5TRv8_00027_0_152.json new file mode 100644 index 0000000000000000000000000000000000000000..2d78b8f1b81b32bf472c009184b5b9c24ddfa521 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_s5x2LY5TRv8_00027_0_152.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_s5x2LY5TRv8_00027_0_152", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/s5x2LY5TRv8_00027_0_152_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/s5x2LY5TRv8_00027_0_152_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/s5x2LY5TRv8_00027_0_152_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/s5x2LY5TRv8_00027_0_152_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slCwOLXIeU4_00025_56_104.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slCwOLXIeU4_00025_56_104.json new file mode 100644 index 0000000000000000000000000000000000000000..9317f973879e714cfaf055ec3979732e75e73821 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slCwOLXIeU4_00025_56_104.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_slCwOLXIeU4_00025_56_104", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/slCwOLXIeU4_00025_56_104_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/slCwOLXIeU4_00025_56_104_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/slCwOLXIeU4_00025_56_104_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/slCwOLXIeU4_00025_56_104_inter", + "rendered_video": "none", + "time_window": [ + 0.4, + 0.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slcf7Y3yEdU_00037_0_411.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slcf7Y3yEdU_00037_0_411.json new file mode 100644 index 0000000000000000000000000000000000000000..73e129210b2cf9e76dc6d5e5cb5deeb0cfef3f4e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_slcf7Y3yEdU_00037_0_411.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_slcf7Y3yEdU_00037_0_411", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/slcf7Y3yEdU_00037_0_411_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/slcf7Y3yEdU_00037_0_411_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/slcf7Y3yEdU_00037_0_411_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/slcf7Y3yEdU_00037_0_411_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 5.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3JoimGNERA_00010_0_88.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3JoimGNERA_00010_0_88.json new file mode 100644 index 0000000000000000000000000000000000000000..7bc5cc4c65966a12abab4ba34ac601695ad2eb4c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3JoimGNERA_00010_0_88.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_t3JoimGNERA_00010_0_88", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/t3JoimGNERA_00010_0_88_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/t3JoimGNERA_00010_0_88_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/t3JoimGNERA_00010_0_88_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/t3JoimGNERA_00010_0_88_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3ybHGZxjjQ_00014_0_208.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3ybHGZxjjQ_00014_0_208.json new file mode 100644 index 0000000000000000000000000000000000000000..10b64eb4f0e8e8a3bfd9483a43205b720ac36939 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t3ybHGZxjjQ_00014_0_208.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_t3ybHGZxjjQ_00014_0_208", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/t3ybHGZxjjQ_00014_0_208_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/t3ybHGZxjjQ_00014_0_208_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/t3ybHGZxjjQ_00014_0_208_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/t3ybHGZxjjQ_00014_0_208_inter", + "rendered_video": "none", + "time_window": [ + 2.5, + 5.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t5wg4c0ScAc_00035_453_724.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t5wg4c0ScAc_00035_453_724.json new file mode 100644 index 0000000000000000000000000000000000000000..1f736e054288ab257eede3c8dd421b7d47b27905 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_t5wg4c0ScAc_00035_453_724.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_t5wg4c0ScAc_00035_453_724", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/t5wg4c0ScAc_00035_453_724_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/t5wg4c0ScAc_00035_453_724_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/t5wg4c0ScAc_00035_453_724_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/t5wg4c0ScAc_00035_453_724_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tFc5jBcitPY_00021_54_214.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tFc5jBcitPY_00021_54_214.json new file mode 100644 index 0000000000000000000000000000000000000000..3eee69bceab62634888b02ea535eaa2299c0bde7 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tFc5jBcitPY_00021_54_214.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tFc5jBcitPY_00021_54_214", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tFc5jBcitPY_00021_54_214_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tFc5jBcitPY_00021_54_214_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tFc5jBcitPY_00021_54_214_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tFc5jBcitPY_00021_54_214_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 3.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tHokJbz14nA_00024_0_240.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tHokJbz14nA_00024_0_240.json new file mode 100644 index 0000000000000000000000000000000000000000..5bc7dbdaa77bb0d0783c86ef94dca0d256a5d34d --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tHokJbz14nA_00024_0_240.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tHokJbz14nA_00024_0_240", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tHokJbz14nA_00024_0_240_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tHokJbz14nA_00024_0_240_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tHokJbz14nA_00024_0_240_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tHokJbz14nA_00024_0_240_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYi23Gsg6Is_00017_157_465.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYi23Gsg6Is_00017_157_465.json new file mode 100644 index 0000000000000000000000000000000000000000..97d319cae92b0451a98a0e184f373c27b264618d --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYi23Gsg6Is_00017_157_465.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tYi23Gsg6Is_00017_157_465", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tYi23Gsg6Is_00017_157_465_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tYi23Gsg6Is_00017_157_465_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tYi23Gsg6Is_00017_157_465_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tYi23Gsg6Is_00017_157_465_inter", + "rendered_video": "none", + "time_window": [ + 2.5, + 8.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYnPjJD8Y0I_00072_231_303.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYnPjJD8Y0I_00072_231_303.json new file mode 100644 index 0000000000000000000000000000000000000000..e31277296a15731f9acde5f80363eb4dcad9af76 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tYnPjJD8Y0I_00072_231_303.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tYnPjJD8Y0I_00072_231_303", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tYnPjJD8Y0I_00072_231_303_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tYnPjJD8Y0I_00072_231_303_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tYnPjJD8Y0I_00072_231_303_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tYnPjJD8Y0I_00072_231_303_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tcUJC5EVwRk_00035_0_63.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tcUJC5EVwRk_00035_0_63.json new file mode 100644 index 0000000000000000000000000000000000000000..e58046ef468408786bc4752dc32afe44785c6a9e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tcUJC5EVwRk_00035_0_63.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tcUJC5EVwRk_00035_0_63", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tcUJC5EVwRk_00035_0_63_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tcUJC5EVwRk_00035_0_63_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tcUJC5EVwRk_00035_0_63_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tcUJC5EVwRk_00035_0_63_inter", + "rendered_video": "none", + "time_window": [ + 0.6, + 1.2 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tfXFKIQ10ro_00023_0_315.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tfXFKIQ10ro_00023_0_315.json new file mode 100644 index 0000000000000000000000000000000000000000..924e67fa528ca672d124377aa8f20b5d67417284 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tfXFKIQ10ro_00023_0_315.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tfXFKIQ10ro_00023_0_315", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tfXFKIQ10ro_00023_0_315_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tfXFKIQ10ro_00023_0_315_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tfXFKIQ10ro_00023_0_315_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tfXFKIQ10ro_00023_0_315_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tkc8RnGP86w_00006_484_1006.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tkc8RnGP86w_00006_484_1006.json new file mode 100644 index 0000000000000000000000000000000000000000..b7c461626d09e9fd6c728d3755a6033ef0817be1 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_tkc8RnGP86w_00006_484_1006.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_tkc8RnGP86w_00006_484_1006", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/tkc8RnGP86w_00006_484_1006_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/tkc8RnGP86w_00006_484_1006_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/tkc8RnGP86w_00006_484_1006_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/tkc8RnGP86w_00006_484_1006_inter", + "rendered_video": "none", + "time_window": [ + 3.5, + 10.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_u-27fhZvFrU_00038_0_210.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_u-27fhZvFrU_00038_0_210.json new file mode 100644 index 0000000000000000000000000000000000000000..eecf32def9b9a3b90237f98e1c62955e678d67f6 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_u-27fhZvFrU_00038_0_210.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_u-27fhZvFrU_00038_0_210", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/u-27fhZvFrU_00038_0_210_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/u-27fhZvFrU_00038_0_210_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/u-27fhZvFrU_00038_0_210_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/u-27fhZvFrU_00038_0_210_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_uQpxQUE5XiA_00000_634_1022.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_uQpxQUE5XiA_00000_634_1022.json new file mode 100644 index 0000000000000000000000000000000000000000..c9f2caad25d7d36215da43694c327800a6c67af2 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_uQpxQUE5XiA_00000_634_1022.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_uQpxQUE5XiA_00000_634_1022", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/uQpxQUE5XiA_00000_634_1022_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/uQpxQUE5XiA_00000_634_1022_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/uQpxQUE5XiA_00000_634_1022_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/uQpxQUE5XiA_00000_634_1022_inter", + "rendered_video": "none", + "time_window": [ + 2.5, + 10.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ux467aEJkjo_00035_0_182.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ux467aEJkjo_00035_0_182.json new file mode 100644 index 0000000000000000000000000000000000000000..f379e3ea484fecb5ecefdb62a39fecef902eff2c --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_ux467aEJkjo_00035_0_182.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_ux467aEJkjo_00035_0_182", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/ux467aEJkjo_00035_0_182_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/ux467aEJkjo_00035_0_182_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/ux467aEJkjo_00035_0_182_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/ux467aEJkjo_00035_0_182_inter", + "rendered_video": "none", + "time_window": [ + 1.2, + 2.7 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v3J14lxJmHo_00021_0_178.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v3J14lxJmHo_00021_0_178.json new file mode 100644 index 0000000000000000000000000000000000000000..eb3792ca89659fd5f83799d5487515ba23b08540 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v3J14lxJmHo_00021_0_178.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_v3J14lxJmHo_00021_0_178", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/v3J14lxJmHo_00021_0_178_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/v3J14lxJmHo_00021_0_178_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/v3J14lxJmHo_00021_0_178_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/v3J14lxJmHo_00021_0_178_inter", + "rendered_video": "none", + "time_window": [ + 0.4, + 1.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v81_5bBEakI_00085_0_247.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v81_5bBEakI_00085_0_247.json new file mode 100644 index 0000000000000000000000000000000000000000..f54cfe0fc4d9e08586d31311722c7d8427419d3b --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_v81_5bBEakI_00085_0_247.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_v81_5bBEakI_00085_0_247", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/v81_5bBEakI_00085_0_247_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/v81_5bBEakI_00085_0_247_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/v81_5bBEakI_00085_0_247_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/v81_5bBEakI_00085_0_247_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOBy-BKZIJc_00008_35_109.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOBy-BKZIJc_00008_35_109.json new file mode 100644 index 0000000000000000000000000000000000000000..a4998c219771a9415c4d092fde90d72eec0da447 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOBy-BKZIJc_00008_35_109.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_vOBy-BKZIJc_00008_35_109", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/vOBy-BKZIJc_00008_35_109_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/vOBy-BKZIJc_00008_35_109_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/vOBy-BKZIJc_00008_35_109_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/vOBy-BKZIJc_00008_35_109_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 1.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOzStTFIJ1g_00047_286_460.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOzStTFIJ1g_00047_286_460.json new file mode 100644 index 0000000000000000000000000000000000000000..fae09bb526aea1d6f8f71a58a0e7e4d801862013 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vOzStTFIJ1g_00047_286_460.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_vOzStTFIJ1g_00047_286_460", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/vOzStTFIJ1g_00047_286_460_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/vOzStTFIJ1g_00047_286_460_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/vOzStTFIJ1g_00047_286_460_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/vOzStTFIJ1g_00047_286_460_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vXgqnnkqNok_00008_0_106.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vXgqnnkqNok_00008_0_106.json new file mode 100644 index 0000000000000000000000000000000000000000..357a3144cab1c18dbd6579cbdc75462dc2c73553 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vXgqnnkqNok_00008_0_106.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_vXgqnnkqNok_00008_0_106", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/vXgqnnkqNok_00008_0_106_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/vXgqnnkqNok_00008_0_106_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/vXgqnnkqNok_00008_0_106_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/vXgqnnkqNok_00008_0_106_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vbf2cZMFasU_00009_0_66.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vbf2cZMFasU_00009_0_66.json new file mode 100644 index 0000000000000000000000000000000000000000..990eb766d008f2dbc3acb8a1ef71eb83c8f10b22 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vbf2cZMFasU_00009_0_66.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_vbf2cZMFasU_00009_0_66", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/vbf2cZMFasU_00009_0_66_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/vbf2cZMFasU_00009_0_66_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/vbf2cZMFasU_00009_0_66_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/vbf2cZMFasU_00009_0_66_inter", + "rendered_video": "none", + "time_window": [ + 0.5, + 1.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_voMV3Mhta1U_00016_0_106.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_voMV3Mhta1U_00016_0_106.json new file mode 100644 index 0000000000000000000000000000000000000000..d8448e1b1ab90bf93ebeaecdea85877b3afd61b7 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_voMV3Mhta1U_00016_0_106.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_voMV3Mhta1U_00016_0_106", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/voMV3Mhta1U_00016_0_106_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/voMV3Mhta1U_00016_0_106_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/voMV3Mhta1U_00016_0_106_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/voMV3Mhta1U_00016_0_106_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vosAVclG2TI_00000_313_612.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vosAVclG2TI_00000_313_612.json new file mode 100644 index 0000000000000000000000000000000000000000..35a3fe7093eb3b99852e76a6c374404d39f2393f --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_vosAVclG2TI_00000_313_612.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_vosAVclG2TI_00000_313_612", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/vosAVclG2TI_00000_313_612_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/vosAVclG2TI_00000_313_612_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/vosAVclG2TI_00000_313_612_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/vosAVclG2TI_00000_313_612_inter", + "rendered_video": "none", + "time_window": [ + 3.5, + 6.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wqrpka69v1w_00022_331_483.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wqrpka69v1w_00022_331_483.json new file mode 100644 index 0000000000000000000000000000000000000000..bba7495073484500ceecd23fa5a791571c187309 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wqrpka69v1w_00022_331_483.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_wqrpka69v1w_00022_331_483", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/wqrpka69v1w_00022_331_483_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/wqrpka69v1w_00022_331_483_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/wqrpka69v1w_00022_331_483_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/wqrpka69v1w_00022_331_483_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wzX4jdw5u1k_00025_0_54.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wzX4jdw5u1k_00025_0_54.json new file mode 100644 index 0000000000000000000000000000000000000000..6267d475c43af51817b58eef483280b1c7435edf --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_wzX4jdw5u1k_00025_0_54.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_wzX4jdw5u1k_00025_0_54", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/wzX4jdw5u1k_00025_0_54_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/wzX4jdw5u1k_00025_0_54_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/wzX4jdw5u1k_00025_0_54_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/wzX4jdw5u1k_00025_0_54_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 1.8 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x7yA3xtWNCU_00020_0_105.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x7yA3xtWNCU_00020_0_105.json new file mode 100644 index 0000000000000000000000000000000000000000..6c13d9b8b6ddd697299ae339b9a418f49a6ca8d4 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x7yA3xtWNCU_00020_0_105.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_x7yA3xtWNCU_00020_0_105", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/x7yA3xtWNCU_00020_0_105_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/x7yA3xtWNCU_00020_0_105_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/x7yA3xtWNCU_00020_0_105_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/x7yA3xtWNCU_00020_0_105_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xD1ywMiX99s_00007_0_194.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xD1ywMiX99s_00007_0_194.json new file mode 100644 index 0000000000000000000000000000000000000000..b1c41504d9b0590806b819e13ddf2edb5ab4598e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xD1ywMiX99s_00007_0_194.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_xD1ywMiX99s_00007_0_194", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/xD1ywMiX99s_00007_0_194_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/xD1ywMiX99s_00007_0_194_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/xD1ywMiX99s_00007_0_194_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/xD1ywMiX99s_00007_0_194_inter", + "rendered_video": "none", + "time_window": [ + 2.0, + 4.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPYSM_X_gnQ_00007_0_110.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPYSM_X_gnQ_00007_0_110.json new file mode 100644 index 0000000000000000000000000000000000000000..3767e3a70722b256ed046581b928161692b4fc1e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPYSM_X_gnQ_00007_0_110.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_xPYSM_X_gnQ_00007_0_110", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/xPYSM_X_gnQ_00007_0_110_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/xPYSM_X_gnQ_00007_0_110_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/xPYSM_X_gnQ_00007_0_110_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/xPYSM_X_gnQ_00007_0_110_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 2.5 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPtkAYDLzBQ_00005_310_395.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPtkAYDLzBQ_00005_310_395.json new file mode 100644 index 0000000000000000000000000000000000000000..73066ff7126401850f6879d2b20e836dce064fef --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xPtkAYDLzBQ_00005_310_395.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_xPtkAYDLzBQ_00005_310_395", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/xPtkAYDLzBQ_00005_310_395_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/xPtkAYDLzBQ_00005_310_395_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/xPtkAYDLzBQ_00005_310_395_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/xPtkAYDLzBQ_00005_310_395_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xVH3jwjG8ok_00015_0_94.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xVH3jwjG8ok_00015_0_94.json new file mode 100644 index 0000000000000000000000000000000000000000..aeda62cef51d3fc8fbf9e16ffbe7f357fc9e336f --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xVH3jwjG8ok_00015_0_94.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_xVH3jwjG8ok_00015_0_94", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/xVH3jwjG8ok_00015_0_94_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/xVH3jwjG8ok_00015_0_94_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/xVH3jwjG8ok_00015_0_94_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/xVH3jwjG8ok_00015_0_94_inter", + "rendered_video": "none", + "time_window": [ + 1.0, + 2.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x_prt6sDXUQ_00059_174_390.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x_prt6sDXUQ_00059_174_390.json new file mode 100644 index 0000000000000000000000000000000000000000..5a58469a832237d5973b058ca658c1df5fed6961 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_x_prt6sDXUQ_00059_174_390.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_x_prt6sDXUQ_00059_174_390", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/x_prt6sDXUQ_00059_174_390_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/x_prt6sDXUQ_00059_174_390_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/x_prt6sDXUQ_00059_174_390_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/x_prt6sDXUQ_00059_174_390_inter", + "rendered_video": "none", + "time_window": [ + 2.5, + 6.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xfhzzAmk9GI_00021_559_843.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xfhzzAmk9GI_00021_559_843.json new file mode 100644 index 0000000000000000000000000000000000000000..a0c1cd818c7eead948dfc6ebf4a96196645d1668 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_xfhzzAmk9GI_00021_559_843.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_xfhzzAmk9GI_00021_559_843", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/xfhzzAmk9GI_00021_559_843_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/xfhzzAmk9GI_00021_559_843_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/xfhzzAmk9GI_00021_559_843_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/xfhzzAmk9GI_00021_559_843_inter", + "rendered_video": "none", + "time_window": [ + 3.0, + 7.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zaJV54MKG-A_00012_0_149.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zaJV54MKG-A_00012_0_149.json new file mode 100644 index 0000000000000000000000000000000000000000..221df2488d1ecfc50f5be36651b21d6105fef05d --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zaJV54MKG-A_00012_0_149.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_zaJV54MKG-A_00012_0_149", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/zaJV54MKG-A_00012_0_149_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/zaJV54MKG-A_00012_0_149_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/zaJV54MKG-A_00012_0_149_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/zaJV54MKG-A_00012_0_149_inter", + "rendered_video": "none", + "time_window": [ + 1.5, + 3.0 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zpjWgmz7_MQ_00026_0_119.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zpjWgmz7_MQ_00026_0_119.json new file mode 100644 index 0000000000000000000000000000000000000000..25f58e401dbe84ebbc4250cc5253bb5834fad818 --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zpjWgmz7_MQ_00026_0_119.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_zpjWgmz7_MQ_00026_0_119", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/zpjWgmz7_MQ_00026_0_119_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/zpjWgmz7_MQ_00026_0_119_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/zpjWgmz7_MQ_00026_0_119_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/zpjWgmz7_MQ_00026_0_119_inter", + "rendered_video": "none", + "time_window": [ + 0.8, + 2.6 + ] + } +} \ No newline at end of file diff --git a/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zta4suL4pxA_00001_454_1076.json b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zta4suL4pxA_00001_454_1076.json new file mode 100644 index 0000000000000000000000000000000000000000..a397a49b96df42868c535967b02737468c99c54e --- /dev/null +++ b/Tasks/Video/VID_INTER_SKEL/L1_VID_INTER_SKEL_zta4suL4pxA_00001_454_1076.json @@ -0,0 +1,34 @@ +{ + "metadata": { + "task_id": "L1_VID_INTER_SKEL_zta4suL4pxA_00001_454_1076", + "task_level": "Level1", + "task_type": "VID_INTER_SKEL", + "original_video": "none", + "original_motion": "Data_Processed/motion_inter/zta4suL4pxA_00001_454_1076_inter", + "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 two video segments of a rendered skeleton showing the starting portion and the ending portion of an action, with a temporal gap in between. Your objective is to predict and generate the missing intermediate movements to bridge this temporal gap. You must ensure the generated motion logically connects the action in the starting video to the action in the ending video and maintains physical balance. Your output MUST be a continuous sequence of ACTION TOKENS representing the missing middle sequence.", + "modifier_instruction": "" + }, + "modalities": { + "text": "", + "audios": [], + "videos_processed": [ + "Data_Processed/Video/VID_INTER_SKEL/zta4suL4pxA_00001_454_1076_start_skel.mp4", + "Data_Processed/Video/VID_INTER_SKEL/zta4suL4pxA_00001_454_1076_end_skel.mp4" + ], + "spatial_coordinates": [], + "input_images": [] + } + }, + "ground_truth": { + "motion_parameters": "Data_Processed/motion_inter/zta4suL4pxA_00001_454_1076_inter", + "rendered_video": "none", + "time_window": [ + 2.5, + 8.0 + ] + } +} \ No newline at end of file