DART: A Server-side Plug-in for Resource-efficient Robust Federated Learning

TL;DR

This paper introduces DART, a server-side plug-in for federated learning that improves robustness against common corruptions without adding computational overhead to clients, making robust FL deployment more practical.

Contribution

DART is a novel server-side robust training plug-in that enhances federated learning robustness without increasing client-side computational costs.

Findings

DART significantly improves robustness of FL models against corruptions.

DART requires no private data access and adds zero client overhead.

Experiments confirm DART's effectiveness and scalability in real-world scenarios.

Abstract

Federated learning (FL) emerged as a popular distributed algorithm to train machine learning models on edge devices while preserving data privacy. However, FL systems face challenges due to client-side computational constraints and from a lack of robustness to naturally occurring common corruptions such as noise, blur, and weather effects. Existing robust training methods are computationally expensive and unsuitable for resource-constrained clients. We propose a novel data-agnostic robust training (DART) plug-in that can be deployed in any FL system to enhance robustness at zero client overhead. DART operates at the server-side and does not require private data access, ensuring seamless integration in existing FL systems. Extensive experiments showcase DART's ability to enhance robustness of state-of-the-art FL systems, establishing it as a practical and scalable solution for real-world robust FL deployment.