Inpainting is a technique used to regenerate parts of an image. To access the inpainting section of the WebUI, click on the img2img tab. Inpainting has two main uses: fixing parts of an image and modifying an existing image with the help of AI. To create an inpainting prompt, describe the specific element of the image that needs to be changed. The settings for inpainting include sampling steps, masked content, inpaint at full resolution, batch count, CFG scale, denoising strength, and height and width. These settings allow for intricate generation of small regions of the image, and the ability to generate multiple images in one go.

Outpainting is a technique used to extend the original image and create large-scale images in any aspect ratio. Outpainting takes into account the image's existing visual elements such as shadows, reflections, and textures to maintain the context of the original image. To use outpainting, set the CFG scale to 6 and the denoising strength to 1. The height and width should be set to the final size of the image after 64 pixels are added in the dimension being expanded. The script for outpainting is called "Poor man's outpainting." The pixels to expand should be set to 64, and the mask blur should be set between 4 to 8. The masked content should be latent noise. The prompt is crucial for the AI to understand what it's looking at for outpainting to work. Finally, expand the image 64 pixels at a time.