Shypd.aiThe largest open index of AI toolsRePaint is an open-source PyTorch implementation for image inpainting using Denoising Diffusion Probabilistic Models. It provides code and models for ImageNet, CelebA-HQ, and Places2 datasets.
Trending
About
RePaint is an open-source project offering the official PyTorch code and models for "RePaint: Inpainting using Denoising Diffusion Probabilistic Models," a CVPR 2022 paper. This tool enables users to fill in missing parts of images by leveraging diffusion models, starting from noise and iteratively denoising the image while incorporating known parts. It supports various datasets like ImageNet, CelebA-HQ, and Places2, and can handle diverse mask types, including extreme cases like generating every second line or upscaling. RePaint is an inference scheme, conditioning pre-trained diffusion models rather than training new ones, and has been shown to outperform state-of-the-art methods in user studies.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
