Generative inpainting
AI intelligently fills in missing parts of an image, making it look complete.
Generative inpainting uses artificial intelligence to fill in blank or missing areas in a digital image. Think of it like a smart digital artist who can guess what should be there and paint it in seamlessly.
For example, if you have a photo with an unwanted object or a part that got cut off, generative inpainting can remove the object or extend the image by creating new pixels that look like they were always part of the original picture.
Generative inpainting is a technique where AI models, typically diffusion models or GANs, are used to synthesize plausible image content for masked or corrupted regions. The model learns the underlying distribution of image data to generate realistic textures, patterns, and structures that blend harmoniously with the unmasked areas.
Key challenges include maintaining global image coherence, avoiding repetitive artifacts, and ensuring semantic consistency with the surrounding context. Tradeoffs often involve computational cost versus generation quality and resolution.
Generative inpainting is a core capability powering advanced image editing features in consumer products. Recent advancements in generative AI models have made these tools more accessible and effective, allowing for sophisticated manipulations like re-framing photos.
- It's all about the angle: Your photos, re-composedGoogle · Google Research Blog ·
- Google Gemini
- Google Photos uses AI, likely involving generative inpainting techniques, to allow users to re-frame photos from new angles after they are taken.
- Outpainting
- Outpainting extends an image beyond its original borders, creating
- Image Restoration
- Image restoration aims to fix degraded images (e.g., noise, b
- Image Segmentation
- Image segmentation identifies and separates different object
The concept of inpainting originated in traditional image processing to reconstruct damaged or missing parts of images. Generative inpainting specifically leverages deep learning models to perform this task with high fidelity.
Open Google Photos and select a picture. Look for editing tools that allow you to remove objects or magically fill in parts of the photo. Try removing a small, unwanted object from the background of a photo to see the AI fill in the space.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is generative inpainting?
- Generative inpainting is an AI technique that intelligently fills in missing or masked areas of an image with realistic content.
- How does it differ from simply copying pixels?
- Instead of copying existing pixels, generative inpainting creates entirely new pixels based on its understanding of the image's context and style.
First seen on Subvolts Apr 22, 2026. Updated Sep 15, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.