# geekyutao/Inpaint-Anything

Inpaint anything using Segment Anything and inpainting models.

Repository: https://github.com/geekyutao/Inpaint-Anything
Canonical: https://ross.abutalabs.com/products/inpaint-anything
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2026-08-22T04:47:40+00:00

## Health v2 (maintenance only)
Score: 65/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 8, longevity 88
- inputs: {"age_days": 1242, "days_push": 11, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7703, forks 659 (observed 2026-08-28T04:10:02.503775+00:00)

## What it is
Inpaint Anything combines Segment Anything (SAM) with inpainting models like LaMa and Stable Diffusion to remove, fill, or replace objects in images, videos, and 3D scenes via a simple click. A newer beta branch modernizes the stack with SAM 3, SDXL, and ProPainter, adding text-prompted object selection and hand-removal tooling for robotics data pipelines.

## Use cases
- remove unwanted objects from photos by clicking on them
- fill a masked region with content described by a text prompt
- replace the background of an object in an image
- inpaint missing regions across video frames
- erase human hands from egocentric video for robot learning datasets
- inpaint textures in 3D scenes

## When to choose
- you need click-based or text-based object removal/filling in images or videos
- you want to combine SAM segmentation with LaMa or Stable Diffusion inpainting
- you are preparing egocentric robot-learning data and need hand removal with masks

## When to avoid
- you need a polished GUI product rather than Python scripts and notebooks
- you lack a GPU or cannot meet CUDA/PyTorch requirements of the newer models
- you need the NeRF-based 3D pipeline verified end-to-end (still unverified in the beta branch)

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, computer-vision, video-processing
- domain: image-processing, computer-vision, artificial-intelligence, robotics
- platform: python, cross-platform
- tags: inpainting, segment-anything, stable-diffusion, lama, object-removal, image-editing, video-inpainting, 3d-scenes, sdxl, egocentric-video, gpu

## Member repositories
- geekyutao/Inpaint-Anything (main) score 65

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:02.503775+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:35:27.192514+00:00, confidence not recorded.
  - readme: https://github.com/geekyutao/Inpaint-Anything (fetched 2026-08-28T04:10:02.503775+00:00, sha 74194e59ab8b)
- Data as of 2026-08-30T08:39:29.467469+00:00.