# muzishen/IMAGDressing

[AAAI 2025]👔IMAGDressing👔: Interactive Modular Apparel Generation for Virtual Dressing. It enables customizable human image generation with flexible garment, pose, and scene control, ensuring high fidelity and garment consistency for virtual dressing.

Repository: https://github.com/muzishen/IMAGDressing
Canonical: https://ross.abutalabs.com/products/imagdressing
Homepage: https://imagdressing.github.io/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: datasets, diffusion-models, try-on, text-to-image-generation
Last push: 2025-09-30T17:26:49+00:00

## Health v2 (maintenance only)
Score: 44/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 44, release rhythm 35, longevity 58
- inputs: {"age_days": 823, "days_push": 337, "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 1343, forks 117 (observed 2026-08-28T04:04:26.360981+00:00)

## What it is
IMAGDressing-v1 is a diffusion-based framework for customizable virtual dressing that generates human images with fixed garments and controllable pose, face, and scene via a garment UNet with hybrid attention. It includes inference and training code, a Gradio WebUI, the CAMI evaluation metric, and the 300k-pair IGPair dataset.

## Use cases
- generate model photos wearing a specific garment for online stores
- virtual try-on of clothes onto a person image
- change the outfit of a person in a photo via inpainting
- control pose and scene when generating dressed human images
- train a garment-consistent diffusion model on the IGPair dataset
- evaluate garment consistency of generated images with CAMI
- create cartoon-style virtual dressing images

## When to choose
- you need garment-faithful human image generation rather than localized inpainting try-on
- you want flexible control over pose, face, and background with Stable Diffusion plugins like ControlNet and IP-Adapter
- you need a large garment-person paired dataset for training virtual dressing models

## When to avoid
- you only need simple 2D warping-based try-on without generative synthesis
- you lack a GPU or cannot run Stable Diffusion locally
- you need production e-commerce integration out of the box rather than research code

## Facets
- artifact type: library
- maturity: active
- function: image-processing, machine-learning, deep-learning, llm-inference
- domain: computer-vision, image-processing, artificial-intelligence, e-commerce
- platform: python
- tags: virtual-try-on, diffusion-models, stable-diffusion, garment-generation, text-to-image, controlnet, ip-adapter, dataset, gradio, aaai-2025, gpu, linux

## Member repositories
- muzishen/IMAGDressing (main) score 44

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:26.360981+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 2026-08-30T04:43:01.020934+00:00, confidence not recorded.
  - readme: https://github.com/muzishen/IMAGDressing (fetched 2026-08-28T04:04:26.360981+00:00, sha 404d7ad06627)
  - homepage: https://imagdressing.github.io/ (fetched 2026-08-29T12:02:17.804699+00:00, sha 7c9870d89bc5)
- Data as of 2026-08-30T08:39:29.467469+00:00.