introducing CycleGAN-Turbo and pix2pix-turbo

[junyanz]

Hi friends,

We have released img2img-turbo repo, which includes both pix2pix-turbo and CycleGAN-Turbo. Our new one-step image-to-image translation methods can support both paired and unpaired training and produce better results by leveraging the pre-trained StableDiffusion-Turbo model. The inference time for a 512x512 image is 0.29 sec on A6000 and 0.11 sec on A100.

This new repo will be actively maintained. More models will be added very soon. Here are some results.

Best,
Jun-Yan

Cat Sketching

Fish Sketching

More Results