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

Open-set recognition (OSR) aims to simultaneously detect unknown-classsamples and classify known-class samples. Most of the existing OSR methods areinductive methods, which generally suffer from the domain shift problem thatthe learned model from the known-class domain might be unsuitable for theunknown-class domain. Addressing this problem, inspired by the success oftransductive learning for alleviating the domain shift problem in many othervisual tasks, we propose an Iterative Transductive OSR framework, calledIT-OSR, which implements three explored modules iteratively, including areliability sampling module, a feature generation module, and a baseline updatemodule. Specifically, at each iteration, a dual-space consistent samplingapproach is presented in the explored reliability sampling module for selectingsome relatively more reliable ones from the test samples according to theirpseudo labels assigned by a baseline method, which could be an arbitraryinductive OSR method. Then, a conditional dual-adversarial generative networkunder an orthogonal coding condition is designed in the feature generationmodule to generate discriminative sample features of both known and unknownclasses according to the selected test samples with their pseudo labels.Finally, the baseline method is updated for sample re-prediction in thebaseline update module by jointly utilizing the generated features, theselected test samples with pseudo labels, and the training samples. Extensiveexperimental results on both the standard-dataset and the cross-datasetsettings demonstrate that the derived transductive methods, by introducing twotypical inductive OSR methods into the proposed IT-OSR framework, achievebetter performances than 15 state-of-the-art methods in most cases.

Orthogonal-Coding-Based Feature Generation for Transductive Open-Set Recognition via Dual-Space Consistent Sampling