FedDefender: Backdoor Attack Defense in Federated Learning

Waris Gill, Ali Anwar, Muhammad Ali Gulzar
7/2/2023
cs.CRcs.AIcs.CV

Official code: warisgill/FedDefender (16 stars, MIT, last commit 2024-07-06).

Abstract

Federated Learning (FL) is a privacy-preserving distributed machine learning technique that enables individual clients (e.g., user participants, edge devices, or organizations) to train a model on their local data in a secure environment and then share the trained model with an aggregator to build a global model collaboratively. In this work, we propose FedDefender, a defense mechanism against targeted poisoning attacks in FL by leveraging differential testing. Our proposed method fingerprints the neuron activations of clients' models on the same input and uses differential testing to identify a potentially malicious client containing a backdoor. We evaluate FedDefender using MNIST and FashionMNIST datasets with 20 and 30 clients, and our results demonstrate that FedDefender effectively mitigates such attacks, reducing the attack success rate (ASR) to 10\% without deteriorating the global model performance.

View on arXivView PDF

Code Implementations(1)

warisgill/FedDefenderOfficial
MIT

FedDefender is a novel defense mechanism designed to safeguard Federated Learning from the poisoning attacks (i.e., backdoor attacks).

164Python, Jupyter NotebookJul 22, 20232 years agoMIT
backdoor-attacksdatadifferential-testingfederated-learningpoisoning-attack

Similar Papers

On the Vulnerability of Backdoor Defenses for Federated Learning

92% match1 repo412023

Backdoor Attacks and Defenses in Federated Learning: Survey, Challenges and Future Research Directions

91% match2023

FedBAP: Backdoor Defense via Benign Adversarial Perturbation in Federated Learning

91% match2025

Client-Side Patching against Backdoor Attacks in Federated Learning

91% match2024

Concealing Backdoor Model Updates in Federated Learning by Trigger-Optimized Data Poisoning

90% match2024

Fedward: Flexible Federated Backdoor Defense Framework with Non-IID Data

90% match2023

Cite this paper

@article{gill2023feddefender,
  title  = {FedDefender: Backdoor Attack Defense in Federated Learning},
  author = {Waris Gill and Ali Anwar and Muhammad Ali Gulzar},
  year   = {2023},
  eprint = {2307.08672},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2307.08672v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2307.08672.svg)](https://codeofpaper.com/paper/2307.08672)

Discussion