Cross-Task Visual In-Context Learning on Second-tier group Shadow Removal to Deraining
19.49PSNRGemini 2.5 Flash
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Gemini 2.5 FlashPrompting Strategy=Fixed2025.11 | 19.49 | 0.555 | 7.54 | |
| T2T-VICL (Gemini 2.5 Flash)Model=Gemini 2.5 Flash, Prompting Strategy=Ours2025.11 | 18.98 | 0.538 | 8.2 | |
| Seedream 4.0Prompting Strategy=Fixed2025.11 | 17.37 | 0.487 | 7.51 | |
| T2T-VICL (Seedream 4.0)Model=Seedream 4.0, Prompting Strategy=Ours2025.11 | 15.25 | 0.431 | 7.68 |