Video-to-music generation on OES-Pub
4.95FAD*V2M-Zero
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| V2M-ZeroV-M Pairs=No, Training data size=0h2026.03 | 4.95 | 0.23 | 0.61 | |
| MTCV2MV-M Pairs=Yes, Training data size=147h2026.03 | 5.44 | 0.2 | 0.37 | |
| GVMGenV-M Pairs=Yes, Training data size=147h2026.03 | 6.25 | 0.18 | 0.35 | |
| M2UGenV-M Pairs=Yes, Training data size=36.7h2026.03 | 6.67 | 0.07 | 0.35 | |
| SONIQUEV-M Pairs=No, Training data size=0h2026.03 | 6.8 | 0.09 | 0.27 | |
| AudioXV-M Pairs=Yes, Training data size=15793h2026.03 | 7.46 | 0.19 | 0.33 | |
| VidMuseV-M Pairs=Yes, Training data size=18000h2026.03 | 10.4 | 0.16 | 0.4 |