Sequential Image-to-Image Editing on Targeted Task Pairs Light Enhancement -> Deraining
0.364LPIPST2T-VICL
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| T2T-VICLBase Model=Seedream 4.02025.11 | 0.364 | — | 0.228 | 104.4 | — | |
| Seedream 4.0Variant=Fixed2025.11 | 0.398 | — | 0.255 | 131.6 | — | |
| T2T-VICLBase Model=Qwen-Image-Edit2025.11 | 0.715 | — | 0.364 | 219.7 | — | |
| Qwen-Image-EditVariant=Fixed2025.11 | 0.757 | — | 0.387 | 235.4 | — | |
| T2T-VICLBase Model=FireRed-Image-Edit2025.11 | 0.781 | — | 0.392 | 250.4 | — | |
| FireRed-Image-EditVariant=Fixed2025.11 | 0.801 | — | 0.403 | 268 | — |