https://github.com/honlan/dmt

Disentangled Makeup Transfer with Generative Adversarial Network
https://github.com/honlan/dmt

computer-vision deep-learning disentangled-representations generative-adversarial-network makeup-transfer tensorflow

Last synced: about 2 months ago
JSON representation

Disentangled Makeup Transfer with Generative Adversarial Network

• Host: GitHub
• URL: https://github.com/honlan/dmt
• Owner: Honlan
• Created: 2019-06-28T07:14:11.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2019-07-15T06:31:25.000Z (over 5 years ago)
• Last Synced: 2023-10-20T18:30:42.317Z (about 1 year ago)
• Topics: computer-vision, deep-learning, disentangled-representations, generative-adversarial-network, makeup-transfer, tensorflow
• Language: Python
• Size: 45.2 MB
• Stars: 116
• Watchers: 7
• Forks: 18
• Open Issues: 10
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# DMT

TensorFlow implementation of [Disentangled Makeup Transfer with Generative Adversarial Network](https://arxiv.org/abs/1907.01144)

The facial images are disentangled into identity codes and makeup codes to achieve diverse scenarios of makeup transfer

## Results

![pairwise makeup transfer](output/pairwise.jpg)

![interpolated makeup transfer](output/interpolated.jpg)

![hybrid makeup transfer](output/hybrid.jpg)

![multimodal makeup transfer](output/multimodal.jpg)

## Files

- `main.py`: the main code

- `dmt.pb`: the pre-trained model

- `faces`: images of makeup and non-makeup faces

- `output`: the generated images

## Usage

```

python main.py

```

If you want to use other non-makeup or makeup images, set the paths to the target images

```

no_makeup = os.path.join('faces', 'no_makeup', 'xfsy_0055.png')

makeup_a = os.path.join('faces', 'makeup', 'XMY-074.png')

makeup_b = os.path.join('faces', 'makeup', 'vFG112.png')

```