Contrastive Learning of Cough Descriptors for Automatic COVID-19 Preliminary Diagnosis

Abstract

Cough sounds as a descriptor have been used for detecting various respiratory ailments based on its intensity, duration of intermediate phase between two cough sounds, repetitions, dryness etc. However, COVID-19 diagnosis using only cough sounds is challenging because of cough being a common symptom among many non COVID-19 health diseases and inherent data imbalance within the available datasets. As one of the approach in this direction, we explore the robustness of multi-domain representation by performing the early fusion over a wide set of temporal, spectral and tempo-spectral handcrafted features, followed by training a Support Vector Machine (SVM) classifier. In our second approach, using a contrastive loss function we learn a latent space from Mel Filter Cepstral Coefficients (MFCCs) where representations belonging to samples having similar cough characteristics are closer. This helps learn representations for the highly varied COVIDnegative class (healthy and symptomatic COVID-negative), by learning multiple smaller clusters. Using only the DiCOVA data, multi-domain features yields an absolute improvement of 0.74% and 1.07%, whereas our second approach shows an improvement of 2.09% and 3.98%, over the blind test and validation set, respectively, when compared with challenge baseline.

Towards sound based testing of COVID-19—Summary of the first Diagnostics of COVID-19 using Acoustics (DiCOVA) Challenge

“…The challenge turnaround time was days, and the progress made by different teams in this short time span highlighted their efforts. Eleven studies pursued in this challenge ( Muguli et al, 2021 , Das et al, 2021 , Mallol-Ragolta et al, 2021 , Ritwik et al, 2021 , Deshpande and Schuller, 2021 , Karas and Schuller, 2021 , Bhosale et al, 2021 , Södergren et al, 2021 , Harvill et al, 2021 , Kamble et al, 2021 , Avila et al, 2021 ), after going through the peer review process, were presented at the DiCOVA Special Session, Interspeech 2021 Conference (on Aug 2021).…”

Low-shot learning and class imbalance: a survey

“…A 2021 paper by Bhosale et al [123] proposes, among other methods, a contrastive learning approach to deal with the class imbalance present in the task of COVID-19 diagnosis from cough sounds. After pre-processing the audio into "MFCC" sound features, they train their FSL model under a 2-way-K-shot paradigm, aiming to circumvent class imbalance by giving equal shots for each class.…”

Intelligent speech technologies for transcription, disease diagnosis, and medical equipment interactive control in smart hospitals: A review

“…In addition, Bhosale et al [ 188 ] and Casanova et al [ 189 ] also used cough sounds to diagnose COVID-19. Moreover, Gosztolya et al utilized SS to distinguish schizophrenia from bipolar disorder [ 190 ].…”

“…Moreover, the fusion of voice signals with signals of other modalities such as electroacoustic gate signals, EMR, X-ray images, and ultrasound [ 4 , 5 ] will be more valuable for disease diagnosis in smart hospitals. For example, combining the chest X-ray images and cough sounds-based COVID-19 non-contact classification methods will minimize severity and mortality rates during the pandemic [ 5 , 6 , 188 , 189 ]. Furthermore, algorithms in other domains can be used in speech signal processing, such as AlexNet, VGGNet, GoogLeNet, and ResNet in image recognition can be adopted in the spectrum of speech signals.…”