Multi-label Video Classification on Charades
50.4mAPBIKE VIT-L
Evaluation Results
| Method | Links | |
|---|---|---|
| BIKE VIT-LFrames=162022.12 | 50.4 | |
| ViT-LFrames=162023.07 | 48.5 | |
| ActionCLIPFrames=322022.12 | 44.3 | |
| ActionCLIPFrames=322023.07 | 44.3 | |
| X3D-XL (312↑)Frames=162022.12 | 43.4 | |
| X3D-XL (312↑)Frames=162023.07 | 43.4 | |
| SlowFast R101Frames=16+642022.12 | 42.5 | |
| SlowFast R101Frames=16+642023.07 | 42.5 | |
| LFB+NLFrames=322023.07 | 42.5 | |
| SlowFast R50Frames=8+322023.07 | 38 | |
| Nonlocal2023.07 | 37.5 | |
| STMFrames=162022.12 | 35.3 | |
| STMFrames=162023.07 | 35.3 | |
| MultiScale TRNFrames=-2022.12 | 25.2 | |
| MultiScale TRN2023.07 | 25.2 |