Audio-Visual Generative Editing on SpongeBob-Bench
285.93FVDSpongeBob
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| SpongeBob2026.05 | 285.93 | 99 | 0.36 | 95.1 | 6.45 | 23.8 | 4.5 | 8.73 | 24 | 81 | 7.6 | |
| AVI-Edit2026.05 | 318.56 | 98.5 | 0.35 | 93.2 | 6.12 | 22.5 | 3.45 | 10.28 | 21 | 72 | 6.2 | |
| VACE-Foley2026.05 | 372.15 | 98.2 | 0.32 | 91.8 | 5.85 | 20.8 | 1.85 | 11.42 | 19 | 62 | 5.3 | |
| VACE+Coh.variant=Coherent2026.05 | 372.15 | 98.2 | 0.32 | 91.8 | 5.62 | 19.8 | 1.72 | 11.65 | 18 | 68 | 5.1 | |
| AvED2026.05 | 548.37 | 95.2 | 0.18 | 86.2 | 4.85 | 21.5 | 1.15 | 12.85 | 15 | 52 | 3.6 |