TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Datasets/Human3.6M

Human3.6M

ImagesVideosCustom (research-only, attribution)Introduced 2014-01-01

The Human3.6M dataset is one of the largest motion capture datasets, which consists of 3.6 million human poses and corresponding images captured by a high-speed motion capture system. There are 4 high-resolution progressive scan cameras to acquire video data at 50 Hz. The dataset contains activities by 11 professional actors in 17 scenarios: discussion, smoking, taking photo, talking on the phone, etc., as well as provides accurate 3D joint positions and high-resolution videos.

Source: Space-Time Representation of People Based on 3D Skeletal Data: A Review

Image Source: Yu et al

Benchmarks

1 Image, 2*2 Stitchi/Average MPJPE (mm)1 Image, 2*2 Stitchi/Using 2D ground-truth joints1 Image, 2*2 Stitchi/Multi-View or Monocular1 Image, 2*2 Stitchi/PA-MPJPE1 Image, 2*2 Stitchi/Acceleration Error1 Image, 2*2 Stitchi/Angular Error1 Image, 2*2 Stitchi/MPVE (mm)1 Image, 2*2 Stitchi/Use Video Sequence1 Image, 2*2 Stitchi/Frames Needed1 Image, 2*2 Stitchi/Need Ground Truth 2D Pose1 Image, 2*2 Stitchi/2D detector1 Image, 2*2 Stitchi/Number of Views1 Image, 2*2 Stitchi/Number of Frames Per View1 Image, 2*2 Stitchi/3D Annotations1 Image, 2*2 Stitchi/Average PMPJPE (mm)1 Image, 2*2 Stitchi/Average MPJPE (mm) for occluded Joints1 Image, 2*2 Stitchi/Expected Calibration Error1 Image, 2*2 Stitchi/MRPE1 Image, 2*2 Stitchi/Average MPJPE (mm) @ 400ms1 Image, 2*2 Stitchi/Average MPJPE (mm) @ 1000 ms1 Image, 2*2 Stitchi/MAR, walking, 400ms1 Image, 2*2 Stitchi/MAR, walking, 1,000ms1 Image, 2*2 Stitchi/APD1 Image, 2*2 Stitchi/ADE1 Image, 2*2 Stitchi/FDE1 Image, 2*2 Stitchi/MMADE1 Image, 2*2 Stitchi/MMFDE1 Image, 2*2 Stitchi/CMD1 Image, 2*2 Stitchi/FID1 Image, 2*2 Stitchi/Average NDMS at 4s 2D Classification/EPE2D Pose Estimation/EPE2D Semantic Segmentation/mIoU3D/Average MPJPE (mm)3D/Using 2D ground-truth joints3D/Multi-View or Monocular3D/PA-MPJPE3D/Acceleration Error3D/Angular Error3D/MPVE (mm)3D/Use Video Sequence3D/Frames Needed3D/Need Ground Truth 2D Pose3D/2D detector3D/Number of Views3D/Number of Frames Per View3D/3D Annotations3D/Average PMPJPE (mm)3D/Average MPJPE (mm) for occluded Joints3D/Expected Calibration Error3D/MRPE3D/Average MPJPE (mm) @ 400ms3D/Average MPJPE (mm) @ 1000 ms3D/MAR, walking, 400ms3D/MAR, walking, 1,000ms3D/APD3D/ADE3D/FDE3D/MMADE3D/MMFDE3D/CMD3D/FID3D/Average NDMS at 4s 3D/MPJPE3D/N-MPJPE3D/P-MPJPE3D Absolute Human Pose Estimation/MRPE3D Absolute Human Pose Estimation/Average MPJPE (mm)3D Absolute Human Pose Estimation/PA-MPJPE3D Human Pose Estimation/Average MPJPE (mm)3D Human Pose Estimation/Using 2D ground-truth joints3D Human Pose Estimation/Multi-View or Monocular3D Human Pose Estimation/PA-MPJPE3D Human Pose Estimation/Acceleration Error3D Human Pose Estimation/Angular Error3D Human Pose Estimation/MPVE (mm)3D Human Pose Estimation/Use Video Sequence3D Human Pose Estimation/Frames Needed3D Human Pose Estimation/Need Ground Truth 2D Pose3D Human Pose Estimation/2D detector3D Human Pose Estimation/Number of Views3D Human Pose Estimation/Number of Frames Per View3D Human Pose Estimation/3D Annotations3D Human Pose Estimation/Average PMPJPE (mm)3D Human Pose Estimation/Average MPJPE (mm) for occluded Joints3D Human Pose Estimation/Expected Calibration Error3D Human Pose Estimation/MRPE3D Pose Estimation/Average MPJPE (mm)3D Reconstruction/MPJPE3D Reconstruction/PA-MPJPE3D Reconstruction/N-MPJPE3D Reconstruction/P-MPJPEActivity Recognition/MMDaActivity Recognition/MMDsHuman Part Segmentation/mIoUHuman action generation/MMDaHuman action generation/MMDsPose Estimation/Average MPJPE (mm)Pose Estimation/Using 2D ground-truth jointsPose Estimation/Multi-View or MonocularPose Estimation/PA-MPJPEPose Estimation/Acceleration ErrorPose Estimation/Angular ErrorPose Estimation/MPVE (mm)Pose Estimation/Use Video SequencePose Estimation/Frames NeededPose Estimation/Need Ground Truth 2D PosePose Estimation/2D detectorPose Estimation/Number of ViewsPose Estimation/Number of Frames Per ViewPose Estimation/3D AnnotationsPose Estimation/Average PMPJPE (mm)Pose Estimation/Average MPJPE (mm) for occluded JointsPose Estimation/Expected Calibration ErrorPose Estimation/MRPEPose Estimation/Average MPJPE (mm) @ 400msPose Estimation/Average MPJPE (mm) @ 1000 msPose Estimation/MAR, walking, 400msPose Estimation/MAR, walking, 1,000msPose Estimation/APDPose Estimation/ADEPose Estimation/FDEPose Estimation/MMADEPose Estimation/MMFDEPose Estimation/CMDPose Estimation/FIDPose Estimation/Average NDMS at 4s Pose Retrieval/Hit@1Pose Retrieval/Hit@10Video/SSIMVideo/MSEVideo/MAEVideo Prediction/SSIMVideo Prediction/MSEVideo Prediction/MAE

Statistics

Papers
783
Benchmarks
135

Links

Homepage

Tasks

1 Image, 2*2 Stitchi2D Classification2D Pose Estimation2D Semantic Segmentation3D3D Absolute Human Pose Estimation3D Human Pose Estimation3D Human Pose Estimation in Limited Data3D Pose Estimation3D ReconstructionActivity RecognitionHuman Part SegmentationHuman Pose ForecastingHuman action generationMonocular 3D Human Pose EstimationMulti-Hypotheses 3D Human Pose EstimationPose EstimationPose RetrievalRoot Joint LocalizationUnsupervised 3D Human Pose EstimationUnsupervised Human Pose EstimationVideoVideo PredictionWeakly-supervised 3D Human Pose Estimation