List your company

Video-to-Audio Generation on VGG-Sound

0Fréchet Distance (FD)Ground truth
-1.25787.2323515.722524.21265Apr 2, 2026

Evaluation results

MethodLinks
Ground truthParams=-, TC=✓2026.040033.16319.8
Woosh-VFlowParams=413M, TC=Recap2026.0419.4421.338276216.1
Woosh-VFlowParams=413M, TC=×2026.0420.8372.11828.657.713.6
Woosh-DVFlowParams=413M, TC=Recap2026.0422.3431.47225.967.515
Woosh-DVFlowParams=413M, TC=×2026.0424.3362.19226.665.512.8
Woosh-DVFlowParams=413M, TC=✓2026.0425.4691.72227.867.323.5
Woosh-VFlowParams=413M, TC=✓2026.0425.7111.69128.263.327.8
MMAudio-MParams=621M, TC=×2026.0430.5281.8463243.616.2
MMAudio-MParams=621M, TC=✓2026.0430.5391.41332.944.922.4
MMAudio-MParams=621M, TC=Recap2026.0431.4451.5430.944.816.6