Text-to-Video Generation on VBench Dynamic Degree (Perceptual and Geometric Metrics)
93.09Background ConsistencyOmniVDiff
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| OmniVDiff2026.06 | 93.09 | 92.3 | 98.75 | 61.35 | 53.3 | 19.563 | 0.6157 | 0.4169 | 856.9066 | 9.2397 | 61.33 | |
| VideoWeave-T2V2026.06 | 93.05 | 95.3 | 99.1 | 73.83 | 59.88 | 21.4994 | 0.7082 | 0.2987 | 500.8694 | 7.278 | 70.68 | |
| GeoVideo2026.06 | 91.82 | 88.39 | 96.46 | 63.34 | 53.4 | 18.9986 | 0.6349 | 0.3975 | 1,072.6943 | 33.7536 | 30.33 | |
| Wan-SFT2026.06 | 90.86 | 90 | 97.12 | 66.59 | 56.33 | 19.7444 | 0.5913 | 0.379 | 815.3085 | 19.9983 | 39.35 | |
| VideoGPA2026.06 | 89.72 | 81.55 | 96.47 | 63.14 | 57.4 | 17.0392 | 0.5353 | 0.4061 | 1,865.7591 | 11.143 | 58.66 |