Text-to-Image Generation on Aesthetic-4K v1.0 (test)
1.51IRSEGA
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SEGAResolution=4096 × 4096, Backbone=Qwen2026.05 | 1.51 | 23.84 | 0.33 | 30.12 | 45.03 | 0.74 | 148.26 | 65.72 | |
| SEGAResolution=2048 × 4096, Backbone=Qwen2026.05 | 1.5 | 23.63 | 0.3 | 29.75 | 50.52 | 0.72 | 149.49 | 53.07 | |
| SEGAResolution=3072 × 3072, Backbone=Qwen2026.05 | 1.49 | 23.67 | 0.32 | 29.97 | 47.95 | 0.72 | 150.31 | 46.22 | |
| SEGAResolution=4096 × 2048, Backbone=Qwen2026.05 | 1.27 | 23.09 | 0.29 | 29.94 | 51.23 | 0.74 | 147.51 | 51.1 | |
| DyPEResolution=3072 × 3072, Backbone=Qwen2026.05 | 1.13 | 22.38 | 0.28 | 29.03 | 47.12 | 0.7 | 151.71 | 66.44 | |
| DyPEResolution=2048 × 4096, Backbone=Qwen2026.05 | 1.07 | 22.2 | 0.27 | 28.6 | 47.44 | 0.65 | 157.86 | 75.97 | |
| UltraImageResolution=3072 × 3072, Backbone=Qwen2026.05 | 1.04 | 22.08 | 0.26 | 29 | 47.92 | 0.66 | 153.5 | 35.8 | |
| DyPEResolution=4096 × 4096, Backbone=Qwen2026.05 | 0.97 | 22.04 | 0.27 | 29.08 | 37.39 | 0.63 | 159.78 | 66 | |
| UltraImageResolution=2048 × 4096, Backbone=Qwen2026.05 | 0.9 | 21.61 | 0.24 | 28.66 | 47.21 | 0.63 | 158.97 | 41.77 | |
| UltraImageResolution=4096 × 4096, Backbone=Qwen2026.05 | 0.81 | 21.53 | 0.27 | 28.55 | 34.74 | 0.59 | 167.04 | 65.9 | |
| BaseResolution=2048 × 4096, Backbone=Qwen2026.05 | 0.6 | 21.77 | 0.25 | 28.17 | 45.05 | 0.64 | 156.53 | 60.33 | |
| UltraImageResolution=4096 × 2048, Backbone=Qwen2026.05 | 0.46 | 21.04 | 0.23 | 28.21 | 46.11 | 0.63 | 164.06 | 40.16 | |
| DyPEResolution=4096 × 2048, Backbone=Qwen2026.05 | 0.44 | 21.39 | 0.26 | 28.73 | 48.06 | 0.65 | 159.38 | 81.79 | |
| BaseResolution=3072 × 3072, Backbone=Qwen2026.05 | 0.37 | 23.17 | 0.25 | 28.03 | 43.07 | 0.65 | 162.39 | 58.61 | |
| BaseResolution=4096 × 4096, Backbone=Qwen2026.05 | -0.1 | 20.95 | 0.21 | 27.82 | 28.48 | 0.54 | 174 | 71.25 | |
| BaseResolution=4096 × 2048, Backbone=Qwen2026.05 | -0.4 | 20.84 | 0.23 | 27.7 | 42.93 | 0.63 | 191.89 | 70.43 |