Complete video soundtrack generation on V2ST-Bench
0.3CLAP ScoreGT
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| GT2026.06 | 0.3 | 0.36 | 8.03 | 0.14 | 4.33 | 4.37 | 4.42 | |
| Foley-Omni2026.06 | 0.27 | 0.26 | 7.59 | 0.16 | 3.92 | 4.13 | 4.14 | |
| MMAudio + CosyVoice 3 + AudioX2026.06 | 0.26 | 0.25 | 10.57 | 0.85 | 2.99 | 3.01 | 2.37 | |
| MMAudio + LipVoicer + AudioX2026.06 | 0.22 | 0.16 | 37.84 | 0.26 | 2.09 | 2.31 | 2.51 |