Instruction-guided Image Editing on HumanEdit Relation
6.153Overall Average ScoreMLLM-as-a-Judge
Evaluation Results
| Method | Links | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MLLM-as-a-JudgeEvaluator=Our Judge2026.02 | 6.153 | 6.167 | 6.084 | 6.278 | — | — | — | — | — | — | — | — | — | — | — | — | |
| HumanEvaluator=Human2026.02 | 5.557 | 5.383 | 5.702 | 5.44 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Human Avg2026.02 | 0.794 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| Background Consistency2026.02 | — | — | — | — | — | 0.504 | — | — | — | — | — | — | — | — | — | — | |
| Clip Image Norm2026.02 | — | — | — | — | — | — | 0.922 | — | — | — | — | — | — | — | — | — | |
| Clip Text Norm2026.02 | — | — | — | — | — | — | — | 0.63 | — | — | — | — | — | — | — | — | |
| Dino Image Norm2026.02 | — | — | — | — | — | — | — | — | 0.805 | — | — | — | — | — | — | — | |
| L1 Error2026.02 | — | — | — | — | — | — | — | — | — | 0.38 | — | — | — | — | — | — | |
| L2 Error2026.02 | — | — | — | — | — | — | — | — | — | — | 0.332 | — | — | — | — | — | |
| Lpips2026.02 | — | — | — | — | — | — | — | — | — | — | — | 0.475 | — | — | — | — | |
| Mask Lpips2026.02 | — | — | — | — | — | — | — | — | — | — | — | — | 0.475 | — | — | — | |
| Mask Ssim2026.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.504 | — | — | |
| MLLM-as-a-Judgeprompt=Main2026.02 | — | — | — | — | 0.804 | — | — | — | — | — | — | — | — | — | — | — | |
| Psnr2026.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.432 | — | |
| Ssim2026.02 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.504 |