bycloudai / pf-afn-windows

Forked official code for "Parser-Free Virtual Try-on via Distilling Appearance Flows", CVPR 2021. https://arxiv.org/abs/2103.04559

Parser-Free Virtual Try-on via Distilling Appearance Flows, CVPR 2021

Official code for CVPR 2021 paper 'Parser-Free Virtual Try-on via Distilling Appearance Flows'

[Paper] [Supplementary Material]

[Checkpoints]

Our Test Environment

anaconda3

pytorch 1.1.0

torchvision 0.3.0

cuda 9.0

cupy 6.0.0

opencv-python 4.5.1

1 GTX1080 GPU

python 3.6

Installation

conda create -n tryon python=3.6

conda activate tryon

conda install pytorch=1.1.0 torchvision=0.3.0 cudatoolkit=9.0 -c pytorch

pip install cupy==6.0.0

pip install opencv-python

git clone https://github.com/geyuying/PF-AFN.git

cd PF-AFN

Run the demo

1. First, you need to download the checkpoints from google drive and put the folder "PFAFN" under the folder "checkpoints". The folder "checkpoints/PFAFN" shold contain "warp_model_final.pth" and "gen_model_final.pth".
2. The "dataset" folder contains the demo images for test, where the "test_img" folder contains the person images, the "test_clothes" folder contains the clothes images, and the "test_edge" folder contains edges extracted from the clothes images with the built-in function in python (We saved the extracted edges from the clothes images for convenience). 'demo.txt' records the test pairs.
3. During test, a person image, a clothes image and its extracted edge are fed into the network to generate the try-on image. No human parsing results or human pose estimation results are needed for test.
4. To test with the saved model, run in anaconda with the command python test.py --name demo --resize_or_crop None --batchSize 1 --gpu_ids 0 and the results will be saved in the folder "results".
5. To reproduce our results from the saved model, your test environment should be the same as our test environment, especifically for the version of cupy.

Dataset

1. VITON contains a training set of 14,221 image pairs and a test set of 2,032 image pairs, each of which has a front-view woman photo and a top clothing image with the resolution 256 x 192. Our saved model is trained on the VITON training set and tested on the VITON test set.
2. To test our saved model on the complete VITON test set, you can download VITON_test.

License

The use of this code is RESTRICTED to non-commercial research and educational purposes.

Citation

@article{ge2021parser,
  title={Parser-Free Virtual Try-on via Distilling Appearance Flows},
  author={Ge, Yuying and Song, Yibing and Zhang, Ruimao and Ge, Chongjian and Liu, Wei and Luo, Ping},
  journal={arXiv preprint arXiv:2103.04559},
  year={2021}
}