pFedBBN: A Personalized Federated Test-Time Adaptation with Balanced Batch Normalization for Class-Imbalanced Data

TL;DR

pFedBBN introduces a federated test-time adaptation method using balanced batch normalization to improve performance on class-imbalanced data without requiring labeled data, enhancing robustness and minority-class accuracy.

Contribution

It proposes a novel personalized federated TTA framework with balanced batch normalization and class-aware aggregation, addressing domain shifts and class imbalance without data sharing.

Findings

Consistently improves robustness over state-of-the-art methods.

Enhances minority-class performance in federated settings.

Supports fully unsupervised local adaptation.

Abstract

Test-time adaptation (TTA) in federated learning (FL) is crucial for handling unseen data distributions across clients, particularly when faced with domain shifts and skewed class distributions. Class Imbalance (CI) remains a fundamental challenge in FL, where rare but critical classes are often severely underrepresented in individual client datasets. Although prior work has addressed CI during training through reliable aggregation and local class distribution alignment, these methods typically rely on access to labeled data or coordination among clients, and none address class unsupervised adaptation to dynamic domains or distribution shifts at inference time under federated CI constraints. Revealing the failure of state-of-the-art TTA in federated client adaptation in CI scenario, we propose pFedBBN,a personalized federated test-time adaptation framework that employs balanced batch normalization (BBN) during local client adaptation to mitigate prediction bias by treating all classes equally, while also enabling client collaboration guided by BBN similarity, ensuring that clients with similar balanced representations reinforce each other and that adaptation remains aligned with domain-specific characteristics. pFedBBN supports fully unsupervised local adaptation and introduces a class-aware model aggregation strategy that enables personalized inference without compromising privacy. It addresses both distribution shifts and class imbalance through balanced feature normalization and domain-aware collaboration, without requiring any labeled or raw data from clients. Extensive experiments across diverse baselines show that pFedBBN consistently enhances robustness and minority-class performance over state-of-the-art FL and TTA methods.