Action Recognition on EPIC-SOUNDS
78.2Top-1 AccuracyMirasol3B
Evaluation Results
| Method | Links | |
|---|---|---|
| Mirasol3BLearnable Param.=3B2025.03 | 78.2 | |
| CA2STLearnable Param.=62M2025.03 | 61 | |
| CA2STModality=A+V, Param.=320M, GFLOPs=4542025.03 | 61 | |
| CAVALearnable Param.=44M2025.03 | 60.3 | |
| CAVAModality=A+V, Param.=216M, GFLOPs=2362025.03 | 60.3 | |
| TIM(A+V)2025.03 | 58.3 | |
| TIMModality=A+V, Param.=435M, GFLOPs=460172025.03 | 58.3 | |
| DiffSED2025.03 | 56.9 | |
| MC32025.03 | 56 | |
| TIMModality=A, Param.=64M, GFLOPs=1162025.03 | 55.7 | |
| ASTLearnable Param.=87M2025.03 | 54.9 | |
| Auditory SlowFastLearnable Param.=27M2025.03 | 53.8 | |
| Auditory SlowFastModality=A, Param.=26M, GFLOPs=42025.03 | 53.8 | |
| SSASTLearnable Param.=87M2025.03 | 53.5 | |
| SSASTModality=A2025.03 | 53.5 | |
| CASTLearnable Param.=45M2025.03 | 47.8 | |
| CASTModality=V, Param.=217M, GFLOPs=3912025.03 | 47.8 |