advimman/lama

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

10217 stars · 1080 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

LaMa is a PyTorch-based image inpainting model that fills large missing regions in images using fast Fourier convolutions, generalizing well to resolutions far above its training resolution. It is the reference implementation of the WACV 2022 paper 'Resolution-robust Large Mask Inpainting with Fourier Convolutions'.

Use cases

• remove unwanted objects from photos
• fill in large missing areas of an image
• inpaint high-resolution images up to 2k
• complete periodic structures like fences or textures
• erase watermarks or text from images
• build an object removal web tool

When to choose

• you need to fill large masks or holes in images
• you need inpainting at resolutions higher than the model was trained on
• you want a proven research model with pretrained weights and many third-party integrations

When to avoid

• you need text-guided generative inpainting rather than mask-based completion
• you need an actively developed project with frequent updates
• you need a lightweight CPU-only solution

Facets

library · maturity maintenance

image-processing machine-learning deep-learning computer-vision image-processing deep-learning machine-learning python cross-platform image-inpainting gan fourier-convolutions pytorch object-removal high-resolution research-model wacv-2022 gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="advimman/lama")