Box Inpainting on ImageNet 256x256 (val)
0.138LPIPSVML-MAP
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VML-MAP2025.12 | 0.138 | 75.8 | — | — | — | |
| MAPGA2025.12 | 0.15 | 91.14 | — | — | — | |
| DAPSiterations=4K2025.12 | 0.187 | 102.8 | — | — | — | |
| DAPSiterations=1K2025.12 | 0.201 | 112.6 | — | — | — | |
| ΠGDM2025.12 | 0.203 | 123.9 | — | — | — | |
| Trust Sampling2024.11 | 0.21 | 99.87 | — | — | — | |
| DDRM2025.12 | 0.211 | 133.1 | — | — | — | |
| DPS+DSG2024.11 | 0.247 | 115.9 | — | — | — | |
| LGD-MCn=102024.11 | 0.28 | 142.77 | — | — | — | |
| DPS2024.11 | 0.282 | 142.03 | — | — | — | |
| LGD-MCn=1002024.11 | 0.282 | 131.23 | — | — | — | |
| ReSampleSpace=Latent2026.05 | 0.309 | — | 18.087 | 71.3 | 265.139 | |
| CLAMPSpace=Latent2026.05 | 0.342 | — | 20.836 | 73.1 | 37.393 | |
| LatentDAPSSpace=Latent2026.05 | 0.543 | — | 22.818 | 56.1 | 88.725 | |
| PSLDSpace=Latent2026.05 | 0.549 | — | 18.762 | 42.4 | 99.832 | |
| EquiCon-PSLDResolution=256 x 256, Noise level (sigma_y)=0.05, Model Type=Latent Diffusion2025.05 | — | 137.6 | 19.25 | — | — | |
| PSLDResolution=256 x 256, Noise level (sigma_y)=0.05, Model Type=Latent Diffusion2025.05 | — | 151.4 | 16.28 | — | — |