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 Domain Adaptation (UDA) approaches address the covariate shiftproblem by minimizing the distribution discrepancy between the source andtarget domains, assuming that the label distribution is invariant acrossdomains. However, in the imbalanced domain adaptation (IDA) scenario, covariateand long-tailed label shifts both exist across domains. To tackle the IDAproblem, some current research focus on minimizing the distributiondiscrepancies of each corresponding class between source and target domains.Such methods rely much on the reliable pseudo labels' selection and the featuredistributions estimation for target domain, and the minority classes withlimited numbers makes the estimations more uncertainty, which influences themodel's performance. In this paper, we propose a cross-domain class discrepancyminimization method based on accumulative class-centroids for IDA (centroIDA).Firstly, class-based re-sampling strategy is used to obtain an unbiasedclassifier on source domain. Secondly, the accumulative class-centroidsalignment loss is proposed for iterative class-centroids alignment acrossdomains. Finally, class-wise feature alignment loss is used to optimize thefeature representation for a robust classification boundary. A series ofexperiments have proved that our method outperforms other SOTA methods on IDAproblem, especially with the increasing degree of label shift.

centroIDA: Cross-Domain Class Discrepancy Minimization Based on Accumulative Class-Centroids for Imbalanced Domain Adaptation