Logit Calibration and Feature Contrast for Robust Federated Learning on Non-IID Data

TL;DR

This paper introduces FatCC, a novel federated learning method that combines logit calibration and feature contrast to improve robustness and accuracy against adversarial attacks in non-IID data settings.

Contribution

It proposes a new approach integrating logit calibration and feature contrast into federated adversarial training to enhance robustness and accuracy on non-IID data.

Findings

FatCC outperforms baseline methods in robustness and accuracy.

The approach effectively handles non-IID data distributions.

Experimental results show significant improvements across multiple datasets.

Abstract

Federated learning (FL) is a privacy-preserving distributed framework for collaborative model training on devices in edge networks. However, challenges arise due to vulnerability to adversarial examples (AEs) and the non-independent and identically distributed (non-IID) nature of data distribution among devices, hindering the deployment of adversarially robust and accurate learning models at the edge. While adversarial training (AT) is commonly acknowledged as an effective defense strategy against adversarial attacks in centralized training, we shed light on the adverse effects of directly applying AT in FL that can severely compromise accuracy, especially in non-IID challenges. Given this limitation, this paper proposes FatCC, which incorporates local logit \underline{C}alibration and global feature \underline{C}ontrast into the vanilla federated adversarial training (\underline{FAT}) process from both logit and feature perspectives. This approach can effectively enhance the federated system's robust accuracy (RA) and clean accuracy (CA). First, we propose logit calibration, where the logits are calibrated during local adversarial updates, thereby improving adversarial robustness. Second, FatCC introduces feature contrast, which involves a global alignment term that aligns each local representation with unbiased global features, thus further enhancing robustness and accuracy in federated adversarial environments. Extensive experiments across multiple datasets demonstrate that FatCC achieves comparable or superior performance gains in both CA and RA compared to other baselines.