/sd-webui-yandere-inpaint-masked-content

Adds yandere inpainting, based on ESRGAN, into stable diffusion webui

Primary LanguagePython

Yandere Inpainting as masked content

This extenstion for AUTOMATIC1111/stable-diffusion-webui adds new value of "Masked content" field in img2img -> inpaint tab. It uses ESRGAN 1x upscaler 1x-NMKD-YandereInpaint

Installation:

  • Install this extension: go into Extensions/Available, and find "yandere inpainting" there. Or install by url
  • Download this model here (find YandereInpaint dir, and .pt file there) and place it in models/ESRGAN folder
  • Restart webui

Usage:

  • in inpainting tab select masked content - yandere inpainting
  • set denoising to low. If you use 0, set sampler to Euler, or it will be crash. 0 means almost original model output
  • be sure your resolution is good
  • use small brush size, in other way the model will generate white areas
  • If you need to remove white artifacts, reduce resolution in Settings -> Extra Inpaint or in extras tab if lama installed

It works good for removing small objects/watermarks in anime arts. But for big areas and for photos lama cleaner is better

If you install lama cleaner, "Yandere Inpaint" will appear as model in extras tab