Taming Cross-Domain Representation Variance in Federated Prototype Learning with Heterogeneous Data Domains
This addresses the problem of cross-domain representation variance in federated learning for clients with heterogeneous data, offering an incremental improvement over existing methods.
The paper tackled performance gaps in federated prototype learning due to heterogeneous data domains by introducing FedPLVM, which uses variance-aware dual-level prototypes clustering and an α-sparsity prototype loss, achieving superior results on datasets like Digit-5, Office-10, and DomainNet.
Federated learning (FL) allows collaborative machine learning training without sharing private data. While most FL methods assume identical data domains across clients, real-world scenarios often involve heterogeneous data domains. Federated Prototype Learning (FedPL) addresses this issue, using mean feature vectors as prototypes to enhance model generalization. However, existing FedPL methods create the same number of prototypes for each client, leading to cross-domain performance gaps and disparities for clients with varied data distributions. To mitigate cross-domain feature representation variance, we introduce FedPLVM, which establishes variance-aware dual-level prototypes clustering and employs a novel $α$-sparsity prototype loss. The dual-level prototypes clustering strategy creates local clustered prototypes based on private data features, then performs global prototypes clustering to reduce communication complexity and preserve local data privacy. The $α$-sparsity prototype loss aligns samples from underrepresented domains, enhancing intra-class similarity and reducing inter-class similarity. Evaluations on Digit-5, Office-10, and DomainNet datasets demonstrate our method's superiority over existing approaches.