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) emerges as a decentralized learning framework whichtrains models from multiple distributed clients without sharing their data topreserve privacy. Recently, large-scale pre-trained models (e.g., VisionTransformer) have shown a strong capability of deriving robust representations.However, the data heterogeneity among clients, the limited computationresources, and the communication bandwidth restrict the deployment oflarge-scale models in FL frameworks. To leverage robust representations fromlarge-scale models while enabling efficient model personalization forheterogeneous clients, we propose a novel personalized FL framework ofclient-specific Prompt Generation (pFedPG), which learns to deploy apersonalized prompt generator at the server for producing client-specificvisual prompts that efficiently adapts frozen backbones to local datadistributions. Our proposed framework jointly optimizes the stages ofpersonalized prompt adaptation locally and personalized prompt generationglobally. The former aims to train visual prompts that adapt foundation modelsto each client, while the latter observes local optimization directions togenerate personalized prompts for all clients. Through extensive experiments onbenchmark datasets, we show that our pFedPG is favorable againststate-of-the-art personalized FL methods under various types of dataheterogeneity, allowing computation and communication efficient modelpersonalization.

Efficient Model Personalization in Federated Learning via Client-Specific Prompt Generation