Disentangled Cycle Consistency for Highly-realistic Virtual Try-On (https://arxiv.org/abs/2103.09479)
- python 3.6
- pytorch 1.0.0
- torchvision 0.3.0
- cuda 10.0
- opencv
To install requirements:
conda create -n dcton python=3.6
conda activate dcton
conda install pytorch==1.0.10 torchvision==0.3.0 cuda100
pip install tensorboardX
pip install opencv-python
pip install imdb
pip install tqdm
For data preparation, please refer to VITON.
Download trained weights. Put the trained weights in the 'pretrained_model' file.
We here provide some data in 'demo_data' file for demo running.
# Demo data running
bash test.sh
The use of this code is restricted to non-commercial research.
Thanks for pytorch-CycleGAN-and-pix2pix for providing the useful codes.
If you think our work is useful, please feel free to cite.
@inproceedings{ge2021disentangled,
title={Disentangled Cycle Consistency for Highly-realistic Virtual Try-On},
author={Ge, Chongjian and Song, Yibing and Ge, Yuying and Yang, Han and Liu, Wei and Luo, Ping},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={16928--16937},
year={2021}
}