StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets

Open paper
Benchmark Model Rank Results
image-generation-on-ffhq-1024-x-1024StyleGAN-XL#2FID: 2.02
image-generation-on-ffhq-256-x-256StyleGAN-XL#4FID: 2.19
image-generation-on-ffhq-256-x-256StyleGAN-XL (DINOv2)#41FD: 240.07Precision: 0.77Recall: 0.43
image-generation-on-ffhq-512-x-512StyleGAN-XL#2FID: 2.41
image-generation-on-imagenet-128x128StyleGAN-XL#3FID: 1.81
image-generation-on-imagenet-256x256StyleGAN-XL#52FID: 2.30
image-generation-on-imagenet-32x32StyleGAN-XL#2FID: 1.10
image-generation-on-imagenet-512x512StyleGAN-XL#30FID: 2.40
image-generation-on-imagenet-64x64StyleGAN-XL#9FID: 1.51NFE: 1
image-generation-on-pokemon-1024x1024StyleGAN-XL#1FID: 25.47
image-generation-on-pokemon-256x256StyleGAN-XL#1FID: 23.97

Rank counts only results with a code link.