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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Unsupervised representation learning has achieved outstanding performancesusing centralized data available on the Internet. However, the increasingawareness of privacy protection limits sharing of decentralized unlabeled imagedata that grows explosively in multiple parties (e.g., mobile phones andcameras). As such, a natural problem is how to leverage these data to learnvisual representations for downstream tasks while preserving data privacy. Toaddress this problem, we propose a novel federated unsupervised learningframework, FedU. In this framework, each party trains models from unlabeleddata independently using contrastive learning with an online network and atarget network. Then, a central server aggregates trained models and updatesclients' models with the aggregated model. It preserves data privacy as eachparty only has access to its raw data. Decentralized data among multipleparties are normally non-independent and identically distributed (non-IID),leading to performance degradation. To tackle this challenge, we propose twosimple but effective methods: 1) We design the communication protocol to uploadonly the encoders of online networks for server aggregation and update themwith the aggregated encoder; 2) We introduce a new module to dynamically decidehow to update predictors based on the divergence caused by non-IID. Thepredictor is the other component of the online network. Extensive experimentsand ablations demonstrate the effectiveness and significance of FedU. Itoutperforms training with only one party by over 5% and other methods by over14% in linear and semi-supervised evaluation on non-IID data.

Collaborative Unsupervised Visual Representation Learning from Decentralized Data