Semi-Siamese-Training

Introduction

This is an implementation of Semi-Siamese Training (SST) by Pytorch. SST is a novel training method to address challenges in Shallow Face Learning (each ID with limited samples), one can refer to our paper "Semi-Siamese Training for Shallow Face Learning" (Paper) in European Conference on Computer Vision (ECCV) 2020.

• [Jan. 2021] One can find the complete implementation at FaceX-Zoo (https://github.com/JDAI-CV/FaceX-Zoo).

Acknowledgement

This work is done with the support of prior works by Tianyu Fu, Shuo Wang, Xiaobo Wang, and Jianzhu Guo.

Citation

Please consider citing our paper in your publications if the project helps your research. BibTeX reference is as follows.

@inproceedings{du2020semi,
  title={Semi-Siamese Training for Shallow Face Learning},
  author={Du, Hang and Shi, Hailin and Liu, Yuchi and Wang, Jun and Lei, Zhen and Zeng, Dan and Mei, Tao},
  booktitle={European Conference on Computer Vision},
  pages={36--53},
  year={2020},
  organization={Springer}
}