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

Multi-label classification is a widely encountered problem in daily life,where an instance can be associated with multiple classes. In theory, this is asupervised learning method that requires a large amount of labeling. However,annotating data is time-consuming and may be infeasible for huge labelingspaces. In addition, label imbalance can limit the performance of multi-labelclassifiers, especially when some labels are missing. Therefore, it ismeaningful to study how to train neural networks using partial labels. In thiswork, we address the issue of label imbalance and investigate how to trainclassifiers using partial labels in large labeling spaces. First, we introducethe pseudo-labeling technique, which allows commonly adopted networks to beapplied in partially labeled settings without the need for additional complexstructures. Then, we propose a novel loss function that leverages statisticalinformation from existing datasets to effectively alleviate the label imbalanceproblem. In addition, we design a dynamic training scheme to reduce thedimension of the labeling space and further mitigate the imbalance. Finally, weconduct extensive experiments on some publicly available multi-label datasetssuch as COCO, NUS-WIDE, CUB, and Open Images to demonstrate the effectivenessof the proposed approach. The results show that our approach outperformsseveral state-of-the-art methods, and surprisingly, in some partial labelingsettings, our approach even exceeds the methods trained with full labels.

Towards Imbalanced Large Scale Multi-label Classification with Partially Annotated Labels