Text-to-Image Generation on MSCOCO 30K
6.61FIDRAPHAEL
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| RAPHAELType=Gen. Only, #Params=3B, # Images=5000M+, Zero-shot=true2024.08 | 6.61 | — | — | — | |
| RAPHAELType=Gen. Only, #Params=3B2024.10 | 6.61 | — | — | — | |
| ImagenType=Gen. Only, #Params=3B, # Images=860M, Zero-shot=true2024.08 | 7.27 | — | — | — | |
| ImagenType=Gen. Only, #Params=34B2024.10 | 7.27 | — | — | — | |
| PixArtType=Gen. Only, #Params=0.6B, # Images=25M, Zero-shot=true2024.08 | 7.32 | — | — | — | |
| PixArt-αType=Gen. Only, #Params=0.6B2024.10 | 7.32 | — | — | — | |
| AdaGen-Stable DiffusionParams=1.4B, Steps=32, TFLOPs=22.9, Zero-shot=true2026.03 | 8.14 | — | — | — | |
| Simple DiffusionParams=2B, Steps=512, TFLOPs=151.3, Zero-shot=true2026.03 | 8.3 | — | — | — | |
| JanusType=Und. and Gen., #Params=1.3B2024.10 | 8.53 | — | — | — | |
| Stable DiffusionParams=1.4B, Steps=50, TFLOPs=35.1, Zero-shot=true2026.03 | 8.59 | — | — | — | |
| DreamLLMType=Und. and Gen., #Params=7B, # Images=-, Zero-shot=true2024.08 | 8.76 | — | — | — | |
| Stable DiffusionParams=1.4B, Steps=32, TFLOPs=22.9, Zero-shot=true2026.03 | 9.03 | — | — | — | |
| GigaGANType=Gen. Only, #Params=0.9B, # Images=2700M, Zero-shot=true2024.08 | 9.09 | — | — | — | |
| GigaGANType=Gen. Only, #Params=0.9B2024.10 | 9.09 | — | — | — | |
| Show-oType=Und. and Gen., #Params=1.3B, # Images=35M, Zero-shot=true2024.08 | 9.24 | — | — | — | |
| Show-oType=Und. and Gen., #Params=1.3B2024.10 | 9.24 | — | — | — | |
| SDv1.5Type=Gen. Only, #Params=0.9B, # Images=2000M, Zero-shot=true2024.08 | 9.62 | — | — | — | |
| SDv1.5Type=Gen. Only, #Params=0.9B2024.10 | 9.62 | — | — | — | |
| DALL-E 2Type=Gen. Only, #Params=6.5B, # Images=650M, Zero-shot=true2024.08 | 10.39 | — | — | — | |
| DALL-E 2Type=Gen. Only, #Params=6.5B2024.10 | 10.39 | — | — | — | |
| DALL-E 2Params=6.5B, Steps=250, Zero-shot=true2026.03 | 10.39 | — | — | — | |
| CorgiParams=5.7B, Steps=64, TFLOPs=43.3, Zero-shot=true2026.03 | 10.88 | — | — | — | |
| NExT-GPTType=Und. and Gen., #Params=13B2024.10 | 11.28 | — | — | — | |
| EmuType=Und. and Gen., #Params=13B2024.10 | 11.66 | — | — | — | |
| Make-A-SceneParams=4B, Steps=1536, Zero-shot=true2026.03 | 11.84 | — | — | — | |
| GLIDEType=Gen. Only, #Params=5B, # Images=250M, Zero-shot=true2024.08 | 12.24 | — | — | — | |
| GLIDEType=Gen. Only, #Params=5B2024.10 | 12.24 | — | — | — | |
| GLIDEParams=5B, Steps=250, TFLOPs=77.0, Zero-shot=true2026.03 | 12.24 | — | — | — | |
| LDMType=Gen. Only, #Params=1.4B, # Images=400M, Zero-shot=true2024.08 | 12.64 | — | — | — | |
| LDMType=Gen. Only, #Params=1.4B2024.10 | 12.64 | — | — | — | |
| LWMType=Und. and Gen., #Params=7B, # Images=-, Zero-shot=true2024.08 | 12.68 | — | — | — | |
| LWMType=Und. and Gen., #Params=7B2024.10 | 12.68 | — | — | — | |
| MACE2024.03 | 12.71 | — | — | 29.51 | |
| FMN2024.03 | 13.99 | — | — | 31.31 | |
| SD v1.4description=Base model2024.03 | 14.04 | — | — | 31.34 | |
| AC2024.03 | 14.08 | — | — | 31.29 | |
| SEED-XType=Und. and Gen., #Params=17B, # Images=-, Zero-shot=true2024.08 | 14.99 | — | — | — | |
| SEED-XType=Und. and Gen., #Params=17B2024.10 | 14.99 | — | — | — | |
| ESDvariant=x2024.03 | 15.19 | — | — | 29.52 | |
| ESDvariant=u2024.03 | 17.07 | — | — | 27.76 | |
| DALL-EParams=12B, Steps=256, TFLOPs=309.8, Zero-shot=true2026.03 | 17.89 | — | — | — | |
| SLD-M2024.03 | 17.95 | — | — | 30.87 | |
| CoDIType=Und. and Gen., #Params=-, # Images=400M, Zero-shot=true2024.08 | 22.26 | — | — | — | |
| CogView 2Params=6B, Steps=1088, TFLOPs=117.0, Zero-shot=true2026.03 | 24 | — | — | — | |
| DM-GAN2021.05 | 26.5 | 32.2 | 2.87 | — | |
| DF-GAN2021.05 | 26.5 | 18.7 | 3.09 | — | |
| LAFITEParams=0.2B, Zero-shot=true2026.03 | 26.94 | — | — | — | |
| CogViewselection_strategy=Best of 60 by CapLosses, super-resolution=false, contrast_enhancement=1.52021.05 | 27.1 | 18.2 | 2.43 | — | |
| CogViewParams=4B, Steps=1088, TFLOPs=182.2, Zero-shot=true2026.03 | 27.1 | — | — | — | |
| DALL-EType=Gen. Only, #Params=12B, # Images=250M, Zero-shot=true2024.08 | 27.5 | — | — | — | |
| DALL-Eselection_strategy=Best of 512 by CLIP, super-resolution=false2021.05 | 27.5 | 17.9 | — | — | |
| DALL-EType=Gen. Only, #Params=12B2024.10 | 27.5 | — | — | — | |
| AttnGAN2021.05 | 35.2 | 23.3 | 3.01 | — | |
| UCE2024.03 | 77.72 | — | — | 19.17 |