FedTracker

This is the official code of our TDSC paper "FedTracker: Furnishing Ownership Verification and Traceability for Federated Learning Model".

Getting Start

First, create a virtual environment using Anaconda.

conda create -n fedtracker python=3.8
conda activate fedtracker

Second, you need to install the necessary packages to run FedTracker.

conda install pytorch==1.13.0 torchvision==0.14.0 pytorch-cuda=11.6 -c pytorch -c nvidia
pip install geneal
pip install quadprog
pip install tqdm

After that, running the bash scripts in /script

bash ./script/vgg16.sh

Citing this work

If you use this repository for academic research, we highly encouraged you to cite our paper.

@article{shao2024fedtracker,
  title={Fedtracker: Furnishing Ownership Verification and Traceability for Federated Learning Model},
  author={Shao, Shuo and Yang, Wenyuan and Gu, Hanlin and Qin, Zhan and Fan, Lixin and Yang, Qiang and Ren, Kui},
  journal={IEEE Transactions on Dependable and Secure Computing},
  year={2024}
}