# ali-vilab/AnyDoor

Official implementations for paper: Anydoor: zero-shot object-level image customization

Repository: https://github.com/ali-vilab/AnyDoor
Canonical: https://ross.abutalabs.com/products/anydoor
Homepage: https://ali-vilab.github.io/AnyDoor-Page/
Language: Python
License: MIT
License Family: permissive
Topics: image-composition, image-customization, image-generation, image-editing
Last push: 2024-04-08T06:31:26+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 81
- inputs: {"age_days": 1143, "days_push": 877, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4238, forks 371 (observed 2026-08-28T04:08:39.582417+00:00)

## What it is
AnyDoor is the official implementation of a diffusion-based model that teleports target objects into new scenes at user-specified locations with zero-shot generalization. It provides training, inference, and demo code along with pretrained checkpoints for object-level image customization.

## Use cases
- move an object into a new background image
- swap objects between two photos
- compose multiple subjects into one scene
- virtual try-on preserving clothing textures
- place a product photo into a marketing scene
- transfer logos or text onto objects

## When to choose
- you need zero-shot object placement in images without per-object fine-tuning
- you want to preserve object texture details while adapting lighting and pose
- you need a baseline for virtual try-on or object moving research

## When to avoid
- you need simple whole-image style transfer or text-to-image generation without object control
- you lack a GPU or cannot download large checkpoints
- you need a production-ready polished application rather than research code

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning
- domain: image-processing, artificial-intelligence, computer-vision
- platform: python, cross-platform
- tags: diffusion-models, image-generation, image-editing, zero-shot, object-composition, virtual-try-on, research-code, gpu

## Member repositories
- ali-vilab/AnyDoor (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:39.582417+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 2026-08-29T18:22:16.599415+00:00, confidence not recorded.
  - readme: https://github.com/ali-vilab/AnyDoor (fetched 2026-08-28T04:08:39.582417+00:00, sha 32ee0f5e7f3b)
  - homepage: https://ali-vilab.github.io/AnyDoor-Page/ (fetched 2026-08-29T09:12:10.964940+00:00, sha c6e889442e38)
- Data as of 2026-08-30T08:39:29.467469+00:00.