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

Cough audio signal classification is a potentially useful tool in screeningfor respiratory disorders, such as COVID-19. Since it is dangerous to collectdata from patients with such contagious diseases, many research teams haveturned to crowdsourcing to quickly gather cough sound data, as it was done togenerate the COUGHVID dataset. The COUGHVID dataset enlisted expert physiciansto diagnose the underlying diseases present in a limited number of uploadedrecordings. However, this approach suffers from potential mislabeling of thecoughs, as well as notable disagreement between experts. In this work, we use asemi-supervised learning (SSL) approach to improve the labeling consistency ofthe COUGHVID dataset and the robustness of COVID-19 versus healthy cough soundclassification. First, we leverage existing SSL expert knowledge aggregationtechniques to overcome the labeling inconsistencies and sparsity in thedataset. Next, our SSL approach is used to identify a subsample of re-labeledCOUGHVID audio samples that can be used to train or augment future coughclassification models. The consistency of the re-labeled data is demonstratedin that it exhibits a high degree of class separability, 3x higher than that ofthe user-labeled data, despite the expert label inconsistency present in theoriginal dataset. Furthermore, the spectral differences in the user-labeledaudio segments are amplified in the re-labeled data, resulting in significantlydifferent power spectral densities between healthy and COVID-19 coughs, whichdemonstrates both the increased consistency of the new dataset and itsexplainability from an acoustic perspective. Finally, we demonstrate how there-labeled dataset can be used to train a cough classifier. This SSL approachcan be used to combine the medical knowledge of several experts to improve thedatabase consistency for any diagnostic classification task.

A Semi-Supervised Algorithm for Improving the Consistency of Crowdsourced Datasets: The COVID-19 Case Study on Respiratory Disorder Classification