Equation-to-Visual Generation on E2V-Bench (test)
0.983Quantity AccuracyDSL-based
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DSL-basedCategory=DSL-based2026.05 | 0.983 | 0.983 | |
| GPT-Image-1Category=Closed-source2026.05 | 0.27 | 0.213 | |
| LMDCategory=Layout-to-image, prompt_refinement=true2026.05 | 0.267 | 0.147 | |
| LMDCategory=Layout-to-image, prompt_refinement=true, regeneration=true2026.05 | 0.267 | 0.153 | |
| LMDCategory=Layout-to-image2026.05 | 0.224 | 0.12 | |
| BagelCategory=Transformer-based, tuning=RSFT2026.05 | 0.167 | 0.147 | |
| Flux.1-devCategory=Diffusion-based2026.05 | 0.087 | 0.063 | |
| Flux.1-devCategory=Diffusion-based, tuning=SFT2026.05 | 0.08 | 0.053 | |
| BagelCategory=Transformer-based2026.05 | 0.08 | 0.057 | |
| BagelCategory=Transformer-based, tuning=SFT2026.05 | 0.08 | 0.063 | |
| SD-3.5-largeCategory=Diffusion-based, full_name=Stable Diffusion-3.5-large2026.05 | 0.067 | 0.033 | |
| Recraft-v3Category=Closed-source2026.05 | 0.06 | 0.047 | |
| Show-o2Category=Transformer-based2026.05 | 0.043 | 0.023 | |
| BlueprintCategory=Layout-to-image2026.05 | 0.027 | 0.013 | |
| PAECategory=Prompt Refinement2026.05 | 0.01 | 0.01 |