English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Federated Learning (FL) has become a popular distributed learning paradigmthat involves multiple clients training a global model collaboratively in adata privacy-preserving manner. However, the data samples usually follow along-tailed distribution in the real world, and FL on the decentralized andlong-tailed data yields a poorly-behaved global model severely biased to thehead classes with the majority of the training samples. To alleviate thisissue, decoupled training has recently been introduced to FL, considering ithas achieved promising results in centralized long-tailed learning byre-balancing the biased classifier after the instance-balanced training.However, the current study restricts the capacity of decoupled training infederated long-tailed learning with a sub-optimal classifier re-trained on aset of pseudo features, due to the unavailability of a global balanced datasetin FL. In this work, in order to re-balance the classifier more effectively, weintegrate the local real data with the global gradient prototypes to form thelocal balanced datasets, and thus re-balance the classifier during the localtraining. Furthermore, we introduce an extra classifier in the training phaseto help model the global data distribution, which addresses the problem ofcontradictory optimization goals caused by performing classifier re-balancinglocally. Extensive experiments show that our method consistently outperformsthe existing state-of-the-art methods in various settings.

Integrating Local Real Data with Global Gradient Prototypes for Classifier Re-Balancing in Federated Long-Tailed Learning