Video-to-Audio Generation on VisualSound (test)
1.76KLDV-AURA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| V-AURAType=Autoregressive (AR)2024.09 | 1.76 | 3.27 | 29.5 | 138 | |
| FrierenType=Rectified Flow Matching (RFM)2024.09 | 2.45 | 2.03 | 23.39 | 285 | |
| Diff-FoleyType=Diffusion (DM)2024.09 | 2.84 | 7.24 | 19.79 | 338 | |
| SpecVQGANType=Autoregressive (AR)2024.09 | 3.41 | 9.02 | 10.87 | 423 |