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/ZeroI2V ViT-L/14

ZeroI2V ViT-L/14

Reported on 10 benchmarks across 3 tasks · 1 paper

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

Robots4 results

  • Activity RecognitiononHMDB-51
    Average accuracy of 3 splits· uses extra data· 2023-10-02
    83.4
    best: 88.7 (VideoMAE V2-g)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Activity RecognitiononSomething-Something V2
    Top-1 Accuracy· uses extra data· 2023-10-02
    72.2
    best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Activity RecognitiononSomething-Something V2
    Top-5 Accuracy· uses extra data· 2023-10-02
    93
    best: 96.3 (DejaVid)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Activity RecognitiononUCF101
    3-fold Accuracy· uses extra data· 2023-10-02
    98.6
    best: 99.7 (FTP-UniFormerV2-L/14)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324

Time Series4 results

  • Action RecognitiononHMDB-51
    Average accuracy of 3 splits· uses extra data· 2023-10-02
    83.4
    best: 88.7 (VideoMAE V2-g)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Action RecognitiononSomething-Something V2
    Top-1 Accuracy· uses extra data· 2023-10-02
    72.2
    best: 77.3 (MVD (Kinetics400 pretrain, ViT-H, 16 frame))
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Action RecognitiononSomething-Something V2
    Top-5 Accuracy· uses extra data· 2023-10-02
    93
    best: 96.3 (DejaVid)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • Action RecognitiononUCF101
    3-fold Accuracy· uses extra data· 2023-10-02
    98.6
    best: 99.7 (FTP-UniFormerV2-L/14)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324

Computer Vision2 results

  • VideoonKinetics-400
    Acc@1· 2023-10-02
    87.2
    best: 93.6 (OmniVec2)
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324
  • VideoonKinetics-400
    Acc@5· 2023-10-02
    97.6
    best: 98.9 (TubeViT-H (ImageNet-1k))
    ZeroI2V: Zero-Cost Adaptation of Pre-trained Transformers from Image to VideoarXiv:2310.01324