RePaint: Inpainting using Denoising Diffusion Probabilistic Models
This addresses the challenge of generating semantically meaningful content for image inpainting across various mask types, which is incremental as it builds on existing DDPM techniques.
The paper tackles the problem of free-form image inpainting, where existing methods struggle with generalization to unseen mask types and produce simple textural extensions, by proposing RePaint, a method using a pretrained unconditional DDPM that samples unmasked regions during reverse diffusion, resulting in outperforming state-of-the-art approaches for at least five out of six mask distributions.
Free-form inpainting is the task of adding new content to an image in the regions specified by an arbitrary binary mask. Most existing approaches train for a certain distribution of masks, which limits their generalization capabilities to unseen mask types. Furthermore, training with pixel-wise and perceptual losses often leads to simple textural extensions towards the missing areas instead of semantically meaningful generation. In this work, we propose RePaint: A Denoising Diffusion Probabilistic Model (DDPM) based inpainting approach that is applicable to even extreme masks. We employ a pretrained unconditional DDPM as the generative prior. To condition the generation process, we only alter the reverse diffusion iterations by sampling the unmasked regions using the given image information. Since this technique does not modify or condition the original DDPM network itself, the model produces high-quality and diverse output images for any inpainting form. We validate our method for both faces and general-purpose image inpainting using standard and extreme masks. RePaint outperforms state-of-the-art Autoregressive, and GAN approaches for at least five out of six mask distributions. Github Repository: git.io/RePaint