Video Classification on Kinetics-Sounds (test)
85Top-1 AccuracyAVSlowFast
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AVSlowFastBackbone=ResNet-101, Modality=Audio-Visual2021.06 | 85 | — | |
| MBTModality=Audio-Visual (AV)2021.06 | 85 | 96.8 | |
| MBTModality=Visual-only2021.06 | 80.7 | 94.9 | |
| MBTModality=Audio-only2021.06 | 52.6 | 71.5 |