SCGAN

This is the code for the ICME 2019 Paper: Semi-supervised Compatibility Learning Across Categories for Clothing Matching.

Usage

Paper data and code

Here are two datasets we used in our paper. After downloaded the datasets, you can put them in the folder data/:

• Taobao: https://tianchi.aliyun.com/dataset/dataDetail?dataId=52 the released dataset is been changed a litte by Alibaba. If you want the totally same dataset as the paper used, you can click here.

• Amazon: http://jmcauley.ucsd.edu/data/amazon/links.html

Quick Start

You need to run the file data/data_preprocess.py first to preprocess the data.

cd data; python data_preprocess.py

Then use vgg-16 to generate the image feature of each items from their images. Click here if you don't want to generate image feature again, we will upload our extracted feature on Google Drive.

python convert_image.py

Then you can run the file ./clothing_matching.py to train the model.

For example: cd pytorch_code; python clothing_matching.py

You can also change other parameters according to the usage in the file './config.py'

Requirements

• Python 2.7
• Tensorflow 1.5.0

Citation

Please cite our paper if you use the code:

@inproceedings{li2019semi,
  title={Semi-Supervised Compatibility Learning Across Categories for Clothing Matching},
  author={Li, Zekun and Cui, Zeyu and Wu, Shu and Zhang, Xiaoyu and Wang, Liang},
  booktitle={2019 IEEE International Conference on Multimedia and Expo (ICME)},
  pages={484--489},
  year={2019},
  organization={IEEE}
}