Video Generation on 6M video dataset
47.81FVDOurs
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OursSteps=600K, External representations=None2026.03 | 47.81 | 8.92 | |
| REPA w/ DINOv2Steps=600K, External representations=DINOv22026.03 | 49.59 | 9.39 | |
| SRASteps=600K, External representations=None2026.03 | 49.75 | 9.02 | |
| Vanilla FlowSteps=600K, External representations=None2026.03 | 50.95 | 9.28 | |
| REPA w/ Depth Anything 3Steps=600K, External representations=Depth Anything 32026.03 | 51.52 | 9.85 | |
| REPA w/ V-JEPA2Steps=600K, External representations=V-JEPA22026.03 | 53.55 | 9.91 |