Imagenet-1M-trained image results

Display Order: Input Context, Ground Truth, Generated Inpainting

Imagenet-1M-trained image results