Video Classification on Kinetics (test)
80.8Top-1 AccuracyMBT
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MBTModality=Audio-Visual (AV)2021.06 | 80.8 | 94.6 | |
| ViViT-BaseModality=Visual2021.06 | 80 | 94 | |
| SlowFastBackbone=ResNet-101-NL, Modality=Visual2021.06 | 79.8 | 93.9 | |
| LGD-3DBackbone=ResNet-101, Modality=Visual2021.06 | 79.4 | 94.4 | |
| MBTModality=Visual-only2021.06 | 79.4 | 94 | |
| AVSlowFastBackbone=ResNet-101, Modality=Audio-Visual2021.06 | 78.8 | 93.6 | |
| 3-stream SATTModality=Visual2021.06 | 77.7 | 93.2 | |
| TS S3D-GModality=Visual2021.06 | 77.2 | 93 | |
| TEAModality=Visual2021.06 | 76.1 | 92.5 | |
| STMModality=Visual2021.06 | 73.7 | 91.6 | |
| blVNetModality=Visual2021.06 | 73.5 | 91.2 | |
| MBTModality=Audio-only2021.06 | 25 | 43.9 |