cmate-virtual-tryon

KO | EN

gitlite — search

#python #react #ai #golang #docker #macos #java #windows #awesome-list #vue #pytorch #security #ios #open-source

cmate-virtual-tryon

(★ 39)

Virtual clothing try on tool powered by deep learning.

cmate-virtual-tryon 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

<img src="https://github.com/xanjay/cmate-virtual-tryon/blob/master/src/flask_app/static/images/logo.png" height="200" /> 
<img src="https://img.shields.io/docker/cloud/build/xanjay/cmate" /> 
## About CMate
CMate is a virtual clothing try on tool. Given a profile image and a source image containing cloth, it applies the cloth to your profile image and let you see how does it look on you as if you were actually wearing it. 
It is a flask web application powered by deep learning. Demo video: https://youtu.be/M4NjXy27Yrs. 
<img src="https://github.com/xanjay/cmate-virtual-tryon/blob/master/docs/cmate-demo.gif" height="300" /> 
### Features
- Supports all kinds of upper wear clothes for both men and women.
- Accurate cloth extraction and fitting.
- Simple, elegent and easy to use web app.

### How does it work ?
It works by extractng cloth from source image and fitting into your profile image. 
Under the hood, it uses two separate deep learning models: 
**Cloth segmentation model**: Custom Deeplab model to extract cloth from source image. 
**Pose estimator model**: Pretrained Openpose body_25 model used to locate shoulder points. 
Extracted cloth is blended into profile image based on shoulder location.

## Dataset
- **Deepfashion2 Dataset**: used to train cloth segmentation model.

## Installation
Use [docker image](https://hub.docker.com/r/xanjay/cmate) OR see [installation.md](https://github.com/xanjay/cmate-virtual-tryon/blob/master/docs/installation.md).

## References
- Deeplab: https://github.com/tensorflow/models/tree/master/research/deeplab
- Deepfashion2 Dataset: https://github.com/switchablenorms/DeepFashion2
- Openpose: https://github.com/CMU-Perceptual-Computing-Lab/openpose

## Contributing
If you have any issue fixes or improvement changes. Fork this repo, make changes and submit pull request.

## License
CMate is freely available for free non-commercial use under [Apache License 2.0](https://github.com/xanjay/cmate-virtual-tryon/blob/master/LICENSE).

cmate-virtual-tryon

cmate-virtual-tryon 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

cmate-virtual-tryon

cmate-virtual-tryon 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README