Hypernetwork-Driven Model Fusion for Federated Domain Generalization
This addresses domain generalization in federated settings, an under-explored area, but appears incremental as it builds on existing hypernetwork and gradient alignment techniques.
The paper tackles the problem of domain shifts in Federated Learning by proposing a hypernetwork-based framework for non-linear model aggregation, achieving the highest in-domain and out-of-domain accuracy on datasets like PACS, Office-Home, and VLCS.
Federated Learning (FL) faces significant challenges with domain shifts in heterogeneous data, degrading performance. Traditional domain generalization aims to learn domain-invariant features, but the federated nature of model averaging often limits this due to its linear aggregation of local learning. To address this, we propose a robust framework, coined as hypernetwork-based Federated Fusion (hFedF), using hypernetworks for non-linear aggregation, facilitating generalization to unseen domains. Our method employs client-specific embeddings and gradient alignment techniques to manage domain generalization effectively. Evaluated in both zero-shot and few-shot settings, hFedF demonstrates superior performance in handling domain shifts. Comprehensive comparisons on PACS, Office-Home, and VLCS datasets show that hFedF consistently achieves the highest in-domain and out-of-domain accuracy with reliable predictions. Our study contributes significantly to the under-explored field of Federated Domain Generalization (FDG), setting a new benchmark for performance in this area.