Semantic Image Synthesis on Cityscapes (val)
76.3mIoUDP-SIMS
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DP-SIMSBackbone=ConvNeXt-L2023.12 | 76.3 | — | 38.2 | 78.5 | |
| DP-GAN2023.12 | 73.6 | — | — | — | |
| OASISNumber of parameters=94M, VGG Perceptual Loss=false2020.12 | 69.3 | — | 47.7 | — | |
| OASIS2023.12 | 69.3 | — | 47.7 | 72 | |
| SDM2023.12 | 69.1 | — | 42.1 | 72.8 | |
| LGGANNumber of parameters=n/a, VGG Perceptual Loss=true2020.12 | 68.4 | — | 57.7 | — | |
| CC-FPSENumber of parameters=131M, VGG Perceptual Loss=true2020.12 | 65.5 | — | 54.3 | — | |
| SPADE+Number of parameters=102M, VGG Perceptual Loss=true2020.12 | 64 | — | 47.8 | — | |
| SPADENumber of parameters=102M, VGG Perceptual Loss=true2020.12 | 62.3 | — | 71.8 | — | |
| SPADE2023.12 | 62.3 | — | 71.8 | — | |
| Pix2pixHDNumber of parameters=183M, VGG Perceptual Loss=true2020.12 | 58.3 | — | 95 | — | |
| CRNNumber of parameters=84M, VGG Perceptual Loss=true2020.12 | 52.4 | — | 104.7 | — | |
| Pix2pixHD2023.12 | 52.4 | — | 104.7 | — | |
| SPADE+Number of parameters=102M, VGG Perceptual Loss=false2020.12 | 47.6 | — | 61.4 | — | |
| SIMSNumber of parameters=56M2020.12 | 47.2 | — | 49.7 | — | |
| Cascaded Refinement NetworkBaseline compared against=Image-space loss2017.07 | — | 99.7 | — | — | |
| Cascaded Refinement NetworkBaseline compared against=GAN+SemSeg2017.07 | — | 98.5 | — | — | |
| Cascaded Refinement NetworkBaseline compared against=Isola et al. [16]2017.07 | — | 96.9 | — | — | |
| Cascaded Refinement NetworkBaseline compared against=Encoder-decoder2017.07 | — | 78.3 | — | — | |
| Cascaded Refinement NetworkBaseline compared against=Full-resolution network2017.07 | — | 67.7 | — | — |