Audio-Visual Generation on LandScape
628.07FVDBridgeDiT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| BridgeDiTProcess=T → VA2025.10 | 628.07 | 4.78 | 0.258 | |
| JavisDiTProcess=T → VA2025.10 | 668.87 | 9.22 | 0.185 | |
| Wan + SDAProcess=T → V ∥ T → A2025.10 | 700.18 | 5.8 | 0.177 | |
| Wan + MMAudioProcess=T → V + V → A2025.10 | 700.18 | 5.31 | 0.218 | |
| Wan + SeeingHearingProcess=T → V + V → A2025.10 | 700.18 | 7.59 | 0.166 | |
| JointDiTProcess=T → I + I → VA2025.10 | 937.09 | 6.35 | 0.075 | |
| SSVGProcess=T → VA2025.10 | 1,186.39 | 8.97 | 0.143 | |
| CoDiProcess=T → VA2025.10 | 1,220.31 | 13.75 | 0.082 | |
| SDA + TPosProcess=T → A + A → V2025.10 | 1,837.42 | 5.8 | 0.143 | |
| SDA + TempoTokenProcess=T → A + A → V2025.10 | 2,089.05 | 5.8 | 0.206 |