A Cough-based deep learning framework for detecting COVID-19

TL;DR

This paper introduces a deep learning framework that detects COVID-19 from cough sounds by combining pre-trained model embeddings and handcrafted features, achieving high accuracy and robustness in challenge datasets.

Contribution

The study proposes a novel feature extraction approach combining embeddings and handcrafted features, and demonstrates its effectiveness in COVID-19 detection from cough sounds.

Findings

Achieved an AUC score of 81.21 on the DiCOVA dataset.

Outperformed baseline models with an 8.43% increase in AUC.

Secured top rankings in the DiCOVA Challenge.

Abstract

This paper presents a deep learning framework for detecting COVID-19 positive subjects from their cough sounds. In particular, the proposed approach comprises two main steps. In the first step, we generate a feature representing the cough sound by combining an embedding extracted from a pre-trained model and handcrafted features extracted from draw audio recording, referred to as the front-end feature extraction. Then, the combined features are fed into different back-end classification models for detecting COVID-19 positive subjects in the second step. Our experiments on the Track-2 dataset of the Second 2021 DiCOVA Challenge achieved the second top ranking with an AUC score of 81.21 and the top F1 score of 53.21 on a Blind Test set, improving the challenge baseline by 8.43% and 23.4% respectively and showing deployability, robustness and competitiveness with the state-of-the-art systems.