Action Recognition on Charades (test)
0.663mAPTokenLearner
Evaluation Results
| Method | Links | |
|---|---|---|
| TokenLearner2022.12 | 0.663 | |
| Tube-ViT-Lvariant=Interpolated2022.12 | 0.662 | |
| MoViNet-A62022.12 | 0.632 | |
| Tube-ViT-Lvariant=MultiTube2022.12 | 0.618 | |
| AssembleNet++-502022.12 | 0.598 | |
| AssembleNet-1012022.12 | 0.586 | |
| DEEP-HALBackbone=I3D, Features=W+G+ODF+SDF, Resolution=exact2020.01 | 0.5016 | |
| DEEP-HALBackbone=I3D, Features=W+G+ODF+SDF, Resolution=SK10242020.01 | 0.5014 | |
| DEEP-HALBackbone=I3D, Features=W+G+ODF+SDF, Resolution=SK5122020.01 | 0.4906 | |
| DEEP-HALBackbone=I3D, Features=W+G+ODF, Resolution=SK5122020.01 | 0.4722 | |
| DEEP-HALBackbone=I3D, Features=W+G+SDF, Resolution=SK5122020.01 | 0.453 | |
| SlowFast2022.12 | 0.452 | |
| ActionCLIPBackbone=ViT-B/16, Frames=322021.09 | 0.443 | |
| LFB NLbackbone=R101-I3D-NL, modality=RGB2018.12 | 0.434 | |
| X3D-XL (312)Frames=162021.09 | 0.434 | |
| HAF/BOW/FV hal.Backbone=I3D2020.01 | 0.431 | |
| SlowFast 101+NLFrames=16+642021.09 | 0.425 | |
| LFB+NLFrames=322021.09 | 0.425 | |
| 3D ResNet-101 + TCModality=RGB2018.12 | 0.411 | |
| Timeception2021.09 | 0.411 | |
| 3D CNNbackbone=R101-I3D-NL, modality=RGB2018.12 | 0.408 | |
| 3D ResNet-101 + GCNModality=RGB + RP2018.12 | 0.397 | |
| STGR+NL2021.09 | 0.397 | |
| I3D-NL [51]backbone=R101-I3D-NL, modality=RGB2018.12 | 0.395 | |
| SlowFast 50Frames=8+322021.09 | 0.38 | |
| I3D + VideoGraphModality=RGB2019.05 | 0.378 | |
| 3D ResNet-101 + NLModality=RGB2018.12 | 0.375 | |
| 3D ResNet-50 + GCNModality=RGB + RP2018.12 | 0.375 | |
| Nonlocal2021.09 | 0.375 | |
| I3D [5] (from [51])backbone=R101-I3D, modality=RGB2018.12 | 0.372 | |
| I3D + TCModality=RGB2018.12 | 0.372 | |
| I3D + TimeceptionModality=RGB2019.05 | 0.372 | |
| 3D ResNet-101Modality=RGB2018.12 | 0.355 | |
| I3D + ActionVLADModality=RGB2019.05 | 0.354 | |
| STMFrames=162021.09 | 0.353 | |
| I3D [5]backbone=Inception-I3D, modality=RGB2018.12 | 0.344 | |
| I3DModality=RGB2018.12 | 0.329 | |
| I3DModality=RGB2019.05 | 0.329 | |
| ResNet-152 + TCModality=RGB2018.12 | 0.316 | |
| ResNet-152 + TimeceptionModality=RGB2019.05 | 0.316 | |
| Temporal RelationsModality=RGB2018.12 | 0.252 | |
| Temporal RelationsModality=RGB2019.05 | 0.252 | |
| MultiScale TRN2021.09 | 0.252 | |
| ResNet-152Modality=RGB2018.12 | 0.228 | |
| ResNet-152Modality=RGB2019.05 | 0.228 | |
| Temporal FieldsModality=RGB + Flow2018.12 | 0.224 | |
| Temporal FieldsModality=RGB + Flow2019.05 | 0.224 | |
| ActionVLADModality=RGB + iDT2018.12 | 0.21 | |
| ActionVLADModality=RGB + iDT2019.05 | 0.21 | |
| Two-streamModality=RGB + Flow2018.12 | 0.186 | |
| Two-streamModality=RGB + Flow2019.05 | 0.186 | |
| Two-stream + LSTMModality=RGB + Flow2018.12 | 0.178 | |
| Two-stream + LSTMModality=RGB + Flow2019.05 | 0.178 |