Unlocking Pre-trained Image Backbones for Semantic Image Synthesis

Benchmark Model Rank Results
image-to-image-translation-on-ade20k-labels-to-photosDP-SIMS (ConvNext-L)–mIoU: 54.3FID: 22.7
image-to-image-translation-on-cityscapes-labels-to-photoDP-SIMS (ConvNext-L)–mIoU: 76.3FID: 38.2
image-to-image-translation-on-coco-stuff-labels-to-photosDP-SIMS (ConvNext-L)–FID: 13.6
image-to-image-translation-on-coco-stuff-labels-to-photosDP-SIMS (ConvNext-XL)–FID: 13.3