Audio-driven video generation on EMTD (test)
15.66FIDWan-S2V-14B
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Wan-S2V-14Bcheckpoint=official2025.12 | 15.66 | 129.57 | 6.41 | 0.435 | 0.142 | 0.677 | |
| Hunyuan-Avatar2025.12 | 18.07 | 145.77 | 4.71 | 0.379 | 0.145 | 0.583 | |
| TalkVerse-5B2025.12 | 18.66 | 147.21 | 5.47 | 0.412 | 0.151 | 0.658 | |
| FantasyTalking2025.12 | 22.6 | 178.12 | 3 | 0.281 | 0.087 | 0.626 | |
| EMO22025.12 | 27.28 | 129.41 | 4.58 | 0.553 | 0.198 | 0.65 | |
| EchoMimicV22025.12 | 33.42 | 217.71 | 4.44 | 0.425 | 0.15 | 0.519 |