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

Many existing FL methods assume clients with fully-labeled data, while inrealistic settings, clients have limited labels due to the expensive andlaborious process of labeling. Limited labeled local data of the clients oftenleads to their local model having poor generalization abilities to their largerunlabeled local data, such as having class-distribution mismatch with theunlabeled data. As a result, clients may instead look to benefit from theglobal model trained across clients to leverage their unlabeled data, but thisalso becomes difficult due to data heterogeneity across clients. In our work,we propose FedLabel where clients selectively choose the local or global modelto pseudo-label their unlabeled data depending on which is more of an expert ofthe data. We further utilize both the local and global models' knowledge viaglobal-local consistency regularization which minimizes the divergence betweenthe two models' outputs when they have identical pseudo-labels for theunlabeled data. Unlike other semi-supervised FL baselines, our method does notrequire additional experts other than the local or global model, nor requireadditional parameters to be communicated. We also do not assume anyserver-labeled data or fully labeled clients. For both cross-device andcross-silo settings, we show that FedLabel outperforms other semi-supervised FLbaselines by $8$-$24\%$, and even outperforms standard fully supervised FLbaselines ($100\%$ labeled data) with only $5$-$20\%$ of labeled data.

Local or Global: Selective Knowledge Assimilation for Federated Learning with Limited Labels