FedFM: Anchor-based Feature Matching for Data Heterogeneity in Federated Learning, TSP

Introduction

This repo is the official implementation of FedFM: Anchor-based Feature Matching for Data Heterogeneity in Federated Learning (IEEE Link / ArXiv Link), which has been accepted by IEEE Transactions on Signal Processing.

We have open-sourced our codes on three libraries, including Pytorch, PaddlePaddle, and MindSpore! Please refer to the corresponding sub-folder for detailed README.

See other projects and papers at Rui Ye's Homepage.

Citation

Please cite our paper if you find the repository helpful.

@article{ye2023fedfm,
  title={Fedfm: Anchor-based feature matching for data heterogeneity in federated learning},
  author={Ye, Rui and Ni, Zhenyang and Xu, Chenxin and Wang, Jianyu and Chen, Siheng and Eldar, Yonina C},
  journal={IEEE Transactions on Signal Processing},
  year={2023},
  publisher={IEEE}
}

Acknowledgement

• MOON repo - Our Pytorch code is adapted from this repo.
• @Xinyu Zhu - Transfer codes from Pytorch to PaddlePaddle and MindSpore.