Interpretable Acoustic Representation Learning on Breathing and Speech Signals for COVID-19 Detection

TL;DR

This paper introduces an interpretable deep learning model using cosine-modulated Gaussian filters and self-attention for COVID-19 detection from breathing and speech audio signals, demonstrating improved accuracy and transferability.

Contribution

The study presents a novel interpretable audio representation learning method with cosine-modulated Gaussian filters and self-attention, enhancing COVID-19 detection performance and applicability across different audio types.

Findings

Significant performance improvements over baseline systems.

Effective transfer learning from larger datasets.

Model applicable to both speech and breathing signals.

Abstract

In this paper, we describe an approach for representation learning of audio signals for the task of COVID-19 detection. The raw audio samples are processed with a bank of 1-D convolutional filters that are parameterized as cosine modulated Gaussian functions. The choice of these kernels allows the interpretation of the filterbanks as smooth band-pass filters. The filtered outputs are pooled, log-compressed and used in a self-attention based relevance weighting mechanism. The relevance weighting emphasizes the key regions of the time-frequency decomposition that are important for the downstream task. The subsequent layers of the model consist of a recurrent architecture and the models are trained for a COVID-19 detection task. In our experiments on the Coswara data set, we show that the proposed model achieves significant performance improvements over the baseline system as well as other representation learning approaches. Further, the approach proposed is shown to be uniformly applicable for speech and breathing signals and for transfer learning from a larger data set.