Multi-reference Image Generation on OmniContext SCENE 1.0 (test)
8.9Character Fidelity Score
GPT-4o
Updated 4d ago
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
2025.12 | 8.9 | 8.44 | 8.6 | 8.75 | |
2025.12 | 8.62 | 8 | 8.08 | 8.09 | |
2025.12 | 7.55 | 8.67 | 7.99 | 8.26 | |
2025.12 | 7.07 | 8.16 | 8.2 | 7.89 | |
2025.12 | 6.38 | 6.71 | 7.04 | 6.97 | |
2025.12 | 5.05 | 7.06 | 7.08 | 6.85 | |
2025.12 | 4.07 | 5.71 | 5.47 | 5.55 | |
2025.12 | 3.59 | 4.32 | 5.12 | 4.8 | |
2025.12 | 3.02 | 3.89 | 2.92 | 3.12 | |
2025.12 | 2.06 | 4.33 | 4.37 | 4.03 |