Consistent Image-Set Generation on Consistent Image-Set Generation Evaluation Set 1.0 (test)
92User CountMatch-and-Fuse
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Match-and-FuseComparison Baseline=FLUX, Evaluation Protocol=2AFC (2-Alternative Forced Choice) preference vs baseline2025.11 | 92 | 0.94 | |
| Match-and-FuseComparison Baseline=IC-LoRA, Evaluation Protocol=2AFC (2-Alternative Forced Choice) preference vs baseline2025.11 | 90 | 0.92 | |
| Match-and-FuseComparison Baseline=Kontext, Evaluation Protocol=2AFC (2-Alternative Forced Choice) preference vs baseline2025.11 | 88 | 0.82 | |
| Match-and-FuseComparison Baseline=Edicho, Evaluation Protocol=2AFC (2-Alternative Forced Choice) preference vs baseline2025.11 | 83 | 0.78 |