FedNoRo: Towards Noise-Robust Federated Learning by Addressing Class Imbalance and Label Noise Heterogeneity
This work addresses a realistic challenge in privacy-preserving decentralized learning, particularly for medical applications, by handling class imbalance and label noise heterogeneity, though it appears incremental as it builds on existing federated noisy label learning approaches.
The paper tackles the problem of federated learning with class-imbalanced global data and heterogeneous label noise, proposing FedNoRo, a two-stage framework that identifies noisy clients and updates models robustly, achieving superior performance on ICH and ISIC2019 datasets compared to state-of-the-art methods.
Federated noisy label learning (FNLL) is emerging as a promising tool for privacy-preserving multi-source decentralized learning. Existing research, relying on the assumption of class-balanced global data, might be incapable to model complicated label noise, especially in medical scenarios. In this paper, we first formulate a new and more realistic federated label noise problem where global data is class-imbalanced and label noise is heterogeneous, and then propose a two-stage framework named FedNoRo for noise-robust federated learning. Specifically, in the first stage of FedNoRo, per-class loss indicators followed by Gaussian Mixture Model are deployed for noisy client identification. In the second stage, knowledge distillation and a distance-aware aggregation function are jointly adopted for noise-robust federated model updating. Experimental results on the widely-used ICH and ISIC2019 datasets demonstrate the superiority of FedNoRo against the state-of-the-art FNLL methods for addressing class imbalance and label noise heterogeneity in real-world FL scenarios.