Video Generation Alignment on Novel set of scenes Video
7.66VLM ScoreATATA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ATATABase Model=WAN 2.1 rectified flow model2026.01 | 7.66 | 0.98 | 0.89 | |
| MatchDiffusionType=Joint generation method2026.01 | 7.64 | 0.96 | 1.12 | |
| VACEConditioning=depth-conditioned ControlNet, Type=Editing-based2026.01 | 6.26 | 0.98 | 2.6 | |
| Lucy-editType=Editing-based2026.01 | 4.77 | 0.97 | 0.84 |