Music-Image-to-Video (MI2V) Dance Generation on CIPE-Dance
67.11Image Quality (IQ)Ground Truth
Evaluation Results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Ground Truth2026.06 | 67.11 | 54.16 | 91.51 | 93.62 | 97.23 | 95.36 | — | — | 8.24 | 4.71 | 28.8 | 12.86 | |
| OmniDanceconditioning=MI2V2026.06 | 65.61 | 54.96 | 92.63 | 93.8 | 98.09 | 96.65 | 17.55 | 1.64 | 10.55 | 5.6 | 29.3 | 12.4 | |
| WAN-S2V2026.06 | 65.33 | 55.81 | 89.97 | 91.92 | 97.34 | 95.47 | 21.47 | 2.83 | 10.25 | 5.39 | 28.3 | 12.1 | |
| Hallo22026.06 | 64.09 | 52.37 | 90.77 | 92.94 | 97.05 | 95.72 | 16.49 | 1.81 | 10.3 | 5.45 | 24.4 | 11.85 | |
| Echomimic-V32026.06 | 62.45 | 53.38 | 89.92 | 92.26 | 97.66 | 95.71 | 21.13 | 1.99 | 10.38 | 5.09 | 27.4 | 12.02 |