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

While semi-supervised learning (SSL) has yielded promising results, the morerealistic SSL scenario remains to be explored, in which the unlabeled dataexhibits extremely high recognition difficulty, e.g., fine-grained visualclassification in the context of SSL (SS-FGVC). The increased recognitiondifficulty on fine-grained unlabeled data spells disaster for pseudo-labelingaccuracy, resulting in poor performance of the SSL model. To tackle thischallenge, we propose Soft Label Selection with Confidence-Aware Clusteringbased on Class Transition Tracking (SoC) by reconstructing the pseudo-labelselection process by jointly optimizing Expansion Objective and ShrinkageObjective, which is based on a soft label manner. Respectively, the formerobjective encourages soft labels to absorb more candidate classes to ensure theattendance of ground-truth class, while the latter encourages soft labels toreject more noisy classes, which is theoretically proved to be equivalent toentropy minimization. In comparisons with various state-of-the-art methods, ourapproach demonstrates its superior performance in SS-FGVC. Checkpoints andsource code are available at https://github.com/NJUyued/SoC4SS-FGVC.

Roll With the Punches: Expansion and Shrinkage of Soft Label Selection for Semi-supervised Fine-Grained Learning