Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

FedSA: A Unified Representation Learning via Semantic Anchors for Prototype-based Federated Learning

About

Prototype-based federated learning has emerged as a promising approach that shares lightweight prototypes to transfer knowledge among clients with data heterogeneity in a model-agnostic manner. However, existing methods often collect prototypes directly from local models, which inevitably introduce inconsistencies into representation learning due to the biased data distributions and differing model architectures among clients. In this paper, we identify that both statistical and model heterogeneity create a vicious cycle of representation inconsistency, classifier divergence, and skewed prototype alignment, which negatively impacts the performance of clients. To break the vicious cycle, we propose a novel framework named Federated Learning via Semantic Anchors (FedSA) to decouple the generation of prototypes from local representation learning. We introduce a novel perspective that uses simple yet effective semantic anchors serving as prototypes to guide local models in learning consistent representations. By incorporating semantic anchors, we further propose anchor-based regularization with margin-enhanced contrastive learning and anchor-based classifier calibration to correct feature extractors and calibrate classifiers across clients, achieving intra-class compactness and inter-class separability of prototypes while ensuring consistent decision boundaries. We then update the semantic anchors with these consistent and discriminative prototypes, which iteratively encourage clients to collaboratively learn a unified data representation with robust generalization. Extensive experiments under both statistical and model heterogeneity settings show that FedSA significantly outperforms existing prototype-based FL methods on various classification tasks.

Yanbing Zhou, Xiangmou Qu, Chenlong You, Jiyang Zhou, Jingyue Tang, Xin Zheng, Chunmao Cai, Yingbo Wu• 2025

Related benchmarks

TaskDatasetResultRank
Image ClassificationCIFAR-10 long-tailed (test)--
211
Image ClassificationOffice-Caltech
Average Accuracy0.6057
44
Image ClassificationSUN397
Accuracy70.86
28
Image ClassificationCIFAR10 (Non-IID 2)
Accuracy77.86
17
Image ClassificationTiny-ImageNet Non-IID 2
Accuracy38.35
13
Federated Image ClassificationOffice-Caltech
Accuracy (Caltech)62.23
9
Image ClassificationCIFAR10 NID1, α=0.2
Accuracy84.27
9
Image ClassificationHAM10000 NID1, α=0.5
Accuracy51.28
9
Image ClassificationHAM10000 NID2
Accuracy44.85
9
Image ClassificationTinyImageNet NID1, α=0.5
Accuracy45.56
9
Showing 10 of 34 rows

Other info

Follow for update