Notably, when examining SVM’s audio signals ROC-AUC scores across all 4 different datasets, we observed an interesting trend in the predictive efficiency of the four feature categories (ComParE, Coswara, Virufy, COUGHVID): time domain (79%, 64%, 56%, 56.80%), tonal (83%, 73%, 69%, 72%), spectral (85%, 74%, 72%, 72.32%), and cepstral (87%, 76%, 71%, 70%).
← all excerpts
Cough2COVID-19 detection using an enhanced multi layer ensemble deep learning framework and CoughFeatureRanker.
1
—
—