EIHW-MTG: Second DiCOVA Challenge System Report

TL;DR

This paper explores deep learning methods, including CNN and ResNet18, with attention mechanisms and demographic info, to improve COVID-19 detection from cough, breath, and speech spectrograms, achieving around 84% AUC.

Contribution

It introduces an outer product-based fusion approach and compares CNN and ResNet18 architectures with attention mechanisms for COVID-19 detection from acoustics.

Findings

Fusion of breath and speech improves detection accuracy.

ResNet18 with attention mechanisms achieves highest AUC of 84.26%.

Using patient sex and contextual attention benefits model performance.

Abstract

This work presents an outer product-based approach to fuse the embedded representations generated from the spectrograms of cough, breath, and speech samples for the automatic detection of COVID-19. To extract deep learnt representations from the spectrograms, we compare the performance of a CNN trained from scratch and a ResNet18 architecture fine-tuned for the task at hand. Furthermore, we investigate whether the patients' sex and the use of contextual attention mechanisms is beneficial. Our experiments use the dataset released as part of the Second Diagnosing COVID-19 using Acoustics (DiCOVA) Challenge. The results suggest the suitability of fusing breath and speech information to detect COVID-19. An Area Under the Curve (AUC) of 84.06% is obtained on the test partition when using a CNN trained from scratch with contextual attention mechanisms. When using the ResNet18 architecture for feature extraction, the baseline model scores the highest performance with an AUC of 84.26%.