T2I-CompBench
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Text-to-Image Generation | T2I-CompBench | Shape Fidelity73.66 | 94 | |
| Text-to-Image Generation | T2I-CompBench (test) | Color Accuracy81 | 67 | |
| Text-to-Image Generation | T2I-CompBench++ | Non-Spatial31.97 | 31 | |
| Text-to-Image Generation | T2I-CompBench | T2I-CompBench Score0.4625 | 27 | |
| Text-to-Image Generation | T2I-CompBench | B-VQA Score75 | 16 | |
| Text-to-Image Generation | T2I-CompBench | Color Fidelity0.8743 | 16 | |
| Text-to-Image Generation | T2I-CompBench 1.0 (test) | CLIP Score0.344 | 14 | |
| Text-to-Image Generation | T2I-CompBench | Evaluation Time (min)5 | 12 | |
| Text-to-Image | T2I-CompBench | Color Fidelity75.68 | 9 | |
| Text-to-Image Alignment | T2I-Compbench | T2I-Compbench Alignment0.5064 | 9 | |
| Text-to-Image Generation | T2I-CompBench | DINO Score0.799 | 9 | |
| Text-to-Image Generation | T2I-CompBench color set | 2 Objects Exist76.3 | 9 | |
| Text-to-Image Generation | T2I-CompBench out-of-domain | Semantic Consistency51.37 | 7 | |
| Text-to-image Generation | T2I-Compbench Count | Human Accuracy48 | 7 | |
| Text-to-image generation | T2I-CompBench | B-VQA75 | 6 | |
| Text-to-image generation | T2I-CompBench | ImageReward0.9377 | 6 | |
| Text-to-Image Generation | T2I-CompBench | 2D Spatial Score48 | 4 | |
| Text-to-Image Generation | T2I-CompBench spatial set | 2 Objects Exist47.2 | 2 |
Showing 18 of 18 rows