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.

Models/UniHCP (finetune)

UniHCP (finetune)

Reported on 29 benchmarks across 13 tasks · 1 paper · 13 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision12 results

  • Pose EstimationonMS-COCO
    AP· 2023-03-06
    76.5
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D Pose EstimationonHuman3.6M
    EPE· 2023-03-06
    6.6
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Pedestrian Attribute RecognitiononPA-100K
    Accuracy· uses extra data· 2023-03-06
    86.18
    best: 90.8 (PATH (Partial FT))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Pedestrian Attribute RecognitiononRAPv2
    Accuracy· 2023-03-06
    82.34
    best: 85.86 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Human Part SegmentationonHuman3.6M
    mIoU· 2023-03-06
    65.95
    best: 69.89 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Human Part SegmentationonCIHP
    Mean IoU· 2023-03-06
    69.8
    best: 72.68 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Person Re-IdentificationonMSMT17
    mAP· 2023-03-06
    67.3
    best: 86.7 (CLIP-ReID (with re-ranking))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Person Re-IdentificationonMarket-1501
    mAP· 2023-03-06
    90.3
    best: 96.21 (Unsupervised Pre-training (ResNet101+RK))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Person Re-IdentificationonCUHK03
    MAP· 2023-03-06
    83.1
    best: 94.3 (Proposed SGGNN)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Pose EstimationonAIC
    AP· 2023-03-06
    33.6
    best: 37.1 (Hulk(Finetune, ViT-L))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Object DetectiononCrowdHuman (full body)
    AP· 2023-03-06
    92.5
    best: 97.2 (InternImage-H)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Object DetectiononCrowdHuman (full body)
    mMR· 2023-03-06
    41.6
    best: 50.49 (Faster RCNN (ResNet50))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936

Methodology11 results

  • 3DonMS-COCO
    AP· 2023-03-06
    76.5
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D ClassificationonHuman3.6M
    EPE· 2023-03-06
    6.6
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 3DonCrowdHuman (full body)
    AP· 2023-03-06
    92.5
    best: 97.2 (InternImage-H)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 3DonCrowdHuman (full body)
    mMR· 2023-03-06
    41.6
    best: 50.49 (Faster RCNN (ResNet50))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 3DonAIC
    AP· 2023-03-06
    33.6
    best: 37.1 (Hulk(Finetune, ViT-L))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D ClassificationonCrowdHuman (full body)
    AP· 2023-03-06
    92.5
    best: 97.2 (InternImage-H)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D ClassificationonCrowdHuman (full body)
    mMR· 2023-03-06
    41.6
    best: 50.49 (Faster RCNN (ResNet50))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D Object DetectiononCrowdHuman (full body)
    AP· 2023-03-06
    92.5
    best: 97.2 (InternImage-H)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D Object DetectiononCrowdHuman (full body)
    mMR· 2023-03-06
    41.6
    best: 50.49 (Faster RCNN (ResNet50))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 16konCrowdHuman (full body)
    AP· 2023-03-06
    92.5
    best: 97.2 (InternImage-H)
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 16konCrowdHuman (full body)
    mMR· 2023-03-06
    41.6
    best: 50.49 (Faster RCNN (ResNet50))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936

Audio4 results

  • 2D Semantic SegmentationonHuman3.6M
    mIoU· 2023-03-06
    65.95
    best: 69.89 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 2D Semantic SegmentationonCIHP
    Mean IoU· 2023-03-06
    69.8
    best: 72.68 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 1 Image, 2*2 StitchionMS-COCO
    AP· 2023-03-06
    76.5
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • 1 Image, 2*2 StitchionAIC
    AP· 2023-03-06
    33.6
    best: 37.1 (Hulk(Finetune, ViT-L))
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936

Robots2 results

  • Autonomous VehiclesonPA-100K
    Accuracy· uses extra data· 2023-03-06
    86.18
    best: 90.8 (PATH (Partial FT))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936
  • Autonomous VehiclesonRAPv2
    Accuracy· 2023-03-06
    82.34
    best: 85.86 (Hulk(Finetune, ViT-L))
    SOTA
    UniHCP: A Unified Model for Human-Centric PerceptionsarXiv:2303.02936