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/LST

LST

Reported on 64 benchmarks across 10 tasks · 2 papers · 26 SOTA

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

Computer Vision29 results

  • VideoonNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • VideoonNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • VideoonN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Image ClassificationonMini-Imagenet 5-way (1-shot)
    Accuracy· 2019-06-03
    70.1
    best: 97.95 (SgVA-CLIP)
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Image ClassificationonTiered ImageNet 5-way (1-shot)
    Accuracy· 2019-06-03
    77.7
    best: 96.8 (CAML [Laion-2b])
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Image ClassificationonTiered ImageNet 5-way (5-shot)
    Accuracy· 2019-06-03
    85.2
    best: 98.8 (CAML [Laion-2b])
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Few-Shot Image ClassificationonMini-Imagenet 5-way (1-shot)
    Accuracy· 2019-06-03
    70.1
    best: 97.95 (SgVA-CLIP)
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Few-Shot Image ClassificationonTiered ImageNet 5-way (1-shot)
    Accuracy· 2019-06-03
    77.7
    best: 96.8 (CAML [Laion-2b])
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Few-Shot Image ClassificationonTiered ImageNet 5-way (5-shot)
    Accuracy· 2019-06-03
    85.2
    best: 98.8 (CAML [Laion-2b])
    SOTA
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • VideoonNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • VideoonNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • VideoonNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • VideoonNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Temporal Action LocalizationonNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action LocalizationonNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Image ClassificationonMini-Imagenet 5-way (5-shot)
    Accuracy· 2019-06-03
    78.7
    best: 98.72 (SgVA-CLIP)
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562
  • Few-Shot Image ClassificationonMini-Imagenet 5-way (5-shot)
    Accuracy· 2019-06-03
    78.7
    best: 98.72 (SgVA-CLIP)
    Learning to Self-Train for Semi-Supervised Few-Shot ClassificationarXiv:1906.00562

Time Series14 results

  • Action DetectiononNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action DetectiononNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 96.7 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 95.6 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 97.4 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 99.6 (PoseC3D (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Action RecognitiononNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318

Methodology7 results

  • Zero-Shot LearningonNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Zero-Shot LearningonNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318

Robots7 results

  • Activity RecognitiononN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 96.7 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 95.6 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 97.4 (DSCNet (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 99.6 (PoseC3D (RGB + Pose))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • Activity RecognitiononNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318

Natural Language Processing7 results

  • 3D Action RecognitiononNTU RGB+D 120
    Accuracy (Cross-Setup)· uses extra data· 2022-08-10
    91.1
    best: 92.2 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononNTU RGB+D 120
    Accuracy (Cross-Subject)· uses extra data· 2022-08-10
    89.9
    best: 90.9 (ProtoGCN)
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononN-UCLA
    Accuracy· 2022-08-10
    97.2
    best: 99.1 (DSCNet (RGB + Pose))
    SOTA
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononNTU RGB+D 120
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononNTU RGB+D
    Accuracy (CS)· uses extra data· 2022-08-10
    92.9
    best: 94.3 (Hulk(Finetune, ViT-L))
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononNTU RGB+D
    Accuracy (CV)· uses extra data· 2022-08-10
    97
    best: 98.3 (ST-GCN [PYSKL, 2D Skeleton])
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318
  • 3D Action RecognitiononNTU RGB+D
    Ensembled Modalities· uses extra data· 2022-08-10
    4
    best: 6 (ProtoGCN)
    Generative Action Description Prompts for Skeleton-based Action RecognitionarXiv:2208.05318