Image Reconstruction on ImageNet-1K
0.28rFIDDINO-Tok-AE
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| DINO-Tok-AEType=AE, Latent=16 × 162025.11 | 0.28 | — | — | — | — | — | |
| TransVQResolution=256x2562026.02 | 0.432 | — | — | 0.067 | — | — | |
| RAEType=AE, Latent=16 × 162025.11 | 0.49 | — | — | — | — | — | |
| VA-VAE-d32Resolution=2562026.05 | 0.52 | 26.2 | 0.77 | — | — | 4.56 | |
| SimVQResolution=256x2562026.02 | 0.551 | — | — | 0.088 | — | — | |
| NS-VQResolution=256x2562026.02 | 0.561 | — | — | 0.087 | — | — | |
| VQGAN-FCResolution=256x2562026.02 | 0.6 | — | — | 0.087 | — | — | |
| MUSEResolution=2562026.05 | 0.62 | 24.9 | 0.78 | — | — | 3.08 | |
| SVGType=AE, Latent=16 × 162025.11 | 0.65 | — | — | — | — | — | |
| VTP-L-d64Resolution=2562026.05 | 0.75 | 24.7 | 0.73 | — | — | 3.01 | |
| UniTokResolution=2562026.05 | 0.76 | 24.1 | 0.71 | — | — | 6.45 | |
| UniLIPResolution=2562026.05 | 0.79 | 23 | 0.75 | — | — | 5.73 | |
| SD-VAEType=AE, Latent=16 × 162025.11 | 0.87 | — | — | — | — | — | |
| VQGANResolution=256x2562026.02 | 1.06 | — | — | 0.113 | — | — | |
| VQGAN-EMAResolution=256x2562026.02 | 1.27 | — | — | 0.124 | — | — | |
| TokenFlowResolution=2562026.05 | 1.37 | 21.6 | 0.68 | — | — | 7.66 | |
| SD-VAEResolution=2562026.05 | 1.85 | 23.6 | 0.68 | — | — | 7.45 | |
| FLUX-VAETokenizer Config=f8d162026.02 | — | 31.01 | 0.914 | 0.017 | 2.88 | — | |
| LV-RAETokenizer Config=f16d7682026.02 | — | 32.5 | 0.938 | 0.017 | 5.51 | — | |
| LV-RAE + Noise AugmentationTokenizer Config=f16d768, Noise Augmentation=true2026.02 | — | 31.11 | 0.927 | 0.019 | 2.49 | — | |
| SD-VAETokenizer Config=f8d42026.02 | — | 25.65 | 0.748 | 0.068 | 13.78 | — | |
| SVGTokenizer Config=f16d3922026.02 | — | 21.87 | 0.632 | 0.114 | 24.92 | — | |
| VA-VAETokenizer Config=f16d322026.02 | — | 26.26 | 0.786 | 0.046 | 4.18 | — |