Event Localization (V → A) on VGGSound AVEL 90K
69.7Segment-level AccuracyCoDAAR
Evaluation Results
| Method | Links | |
|---|---|---|
| CoDAARSetting=AVT2026.05 | 69.7 | |
| DCIDSetting=AVT2026.05 | 64 | |
| CoDAARModality Setting=AV2026.05 | 60 | |
| DCIDModality Setting=AV2026.05 | 59.7 | |
| MICUSetting=AVT2026.05 | 45.6 | |
| MICUModality Setting=AV2026.05 | 42.9 | |
| CMCMModality Setting=AV2026.05 | 38.4 | |
| CODISModality Setting=AV2026.05 | 33.8 | |
| TURNModality Setting=AV2026.05 | 32.5 | |
| BaselineSetting=AVT2026.05 | 8.6 | |
| BaselineModality Setting=AV2026.05 | 6.2 |