Reconstruction on MMBench2 10 task sets (seen)
39.66PSNR (dB)Ours Post-FT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Ours Post-FTParams=102 M, Latent/frame=40962026.06 | 39.66 | 0.007 | |
| Ours Coverage-awareParams=102 M, Latent/frame=40962026.06 | 38.93 | 0.008 | |
| Ours BaseParams=102 M, Latent/frame=40962026.06 | 38.29 | 0.011 | |
| Wan 2.1 VAEParams=127 M, Latent/frame=40962026.06 | 36.45 | 0.01 | |
| SD-VAE-MSEParams=84 M, Latent/frame=31362026.06 | 33.32 | 0.031 | |
| Cosmos-CV8x8x8 (1.0)Params=106 M, Latent/frame=20482026.06 | 32.8 | 0.05 | |
| DC-AE-f32c32Params=323 M, Latent/frame=20482026.06 | 31.49 | 0.035 |