DrawBench
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Text-to-Image Generation | DrawBench | Pick Score23.26 | 40 | |
| Text-to-image synthesis evaluation (Human Correlation - Overall) | DrawBench | Kendall's Tau0.223 | 19 | |
| Text-to-image generation | DrawBench | Latency (s)5.5 | 18 | |
| Text-to-Image Generation | DrawBench | VQAScore0.901 | 18 | |
| Text-to-Image Generation | DrawBench FLUX.1 (dev) | IR0.9655 | 13 | |
| Text-to-Image Generation | DrawBench | VS8.847 | 12 | |
| Text-to-Image Generation | DrawBench | PickScore24.05 | 12 | |
| Text-to-image generation | DrawBench 512 x 512 resolution | ImageReward1.0092 | 10 | |
| Text-to-Image Generation | DrawBench | PickScore23.8 | 9 | |
| Text-to-image synthesis evaluation (Human Correlation - Error Counting) | DrawBench | Kendall's Tau0.2125 | 9 | |
| Grounding Accuracy | DrawBench | Spatial60 | 8 | |
| Text-to-Image Generation | DrawBench | Spatial Fidelity (Human)93.13 | 8 | |
| Text-to-Image Reranking | DrawBench (test) | Quality Score87.1 | 8 | |
| Text-to-Image Generation | DrawBench | PickScore17.597 | 7 | |
| Text-to-Image Generation | DrawBench | UnifiedReward3.06 | 7 | |
| Text-to-Image Generation | DrawBench | Aes.5.897 | 5 | |
| Image Quality | DrawBench | DeQA4.42 | 5 | |
| Human Preference | DrawBench | PickScore23.53 | 5 | |
| Text-to-Image Generation | DrawBench (test) | Accuracy53 | 5 | |
| Text-to-image generation | DrawBench | Aesthetic Score6.046 | 4 | |
| Human Preference Alignment | DrawBench Task-specific (test) | PickScore (Task Metric)24.64 | 4 | |
| Visual Text Rendering | DrawBench Task-specific Prompts (test) | OCR Accuracy95 | 4 | |
| Compositional Image Generation | DrawBench and Task-specific (test) | GenEval0.97 | 4 | |
| Text-to-Image Alignment | DrawBench (test) | CLIP Score0.291 | 3 | |
| Text-to-Video Generation | DrawBench 200 prompts | Quality0.7688 | 2 |
Showing 25 of 25 rows