FedFusion: Federated Learning with Diversity- and Cluster-Aware Encoders for Robust Adaptation under Label Scarcity
This addresses robust federated learning for clients with heterogeneous and label-scarce data, representing an incremental improvement by harmonizing existing techniques like personalization and domain adaptation.
The paper tackled the challenges of heterogeneous feature spaces, non-IID data, and scarce labels in federated learning by introducing FedFusion, a framework that unifies domain adaptation and frugal labeling with diversity- and cluster-aware encoders, resulting in consistent outperformance of state-of-the-art baselines in accuracy, robustness, and fairness across benchmarks.
Federated learning in practice must contend with heterogeneous feature spaces, severe non-IID data, and scarce labels across clients. We present FedFusion, a federated transfer-learning framework that unifies domain adaptation and frugal labelling with diversity-/cluster-aware encoders (DivEn, DivEn-mix, DivEn-c). Labelled teacher clients guide learner clients via confidence-filtered pseudo-labels and domain-adaptive transfer, while clients maintain personalised encoders tailored to local data. To preserve global coherence under heterogeneity, FedFusion employs similarity-weighted classifier coupling (with optional cluster-wise averaging), mitigating dominance by data-rich sites and improving minority-client performance. The frugal-labelling pipeline combines self-/semi-supervised pretext training with selective fine-tuning, reducing annotation demands without sharing raw data. Across tabular and imaging benchmarks under IID, non-IID, and label-scarce regimes, FedFusion consistently outperforms state-of-the-art baselines in accuracy, robustness, and fairness while maintaining comparable communication and computation budgets. These results show that harmonising personalisation, domain adaptation, and label efficiency is an effective recipe for robust federated learning under real-world constraints.