Speaking Corona? Human and Machine Recognition of COVID-19 from Voice

Abstract

With the COVID-19 pandemic, several research teams have reported successful advances in automated recognition of COVID-19 by voice. Resulting voice-based screening tools for COVID-19 could support large-scale testing efforts. While capabilities of machines on this task are progressing, we approach the so far unexplored aspect whether human raters can distinguish COVID-19 positive and negative tested speakers from voice samples, and compare their performance to a machine learning baseline. To account for the challenging symptom similarity between COVID-19 and other respiratory diseases, we use a carefully balanced dataset of voice samples, in which COVID-19 positive and negative tested speakers are matched by their symptoms alongside COVID-19 negative speakers without symptoms. Both human raters and the machine struggle to reliably identify COVID-19 positive speakers in our dataset. These results indicate that particular attention should be paid to the distribution of symptoms across all speakers of a dataset when assessing the capabilities of existing systems. The identification of acoustic aspects of COVID-19-related symptom manifestations might be the key for a reliable voice-based COVID-19 detection in the future by both trained human raters and machine learning models.

Voice Analysis for Neurological Disorder Recognition–A Systematic Review and Perspective on Emerging Trends

“…Since in anonymous data collection efforts [e.g., Zhang et al ( 45 )], it cannot be ruled out that the same speaker donates multiple samples, evaluation of the system's performance might be biased. Recruiting a clear set of speakers can be a solution, or using a setup in which the user has to register with a unique ID [e.g., via email address, Hecker et al ( 86 )].…”

The Acoustic Dissection of Cough: Diving into Machine Listening-based COVID-19 Analysis and Detection

“…A promising approach for a screening tool fulfilling these requirements could be based on bioacoustic signals such as speech sounds or cough sounds (Brown et al, 2020;Hecker et al, 2021;. Several studies have reported acoustic peculiarities in the speech of patients who have diseases associated with symptoms affecting anatomical correlates of speech production, such as bronchial asthma (Balamurali et al, 2020;Dogan et al, 2007) or vocal cord disorders (Falk et al, 2021;Jesus et al, 2015;Petrović-Lazić et al, 2011).…”

“…A promising approach for a screening tool fulfilling these requirements could be based on bioacoustic signals such as speech sounds or cough sounds (Brown et al, 2020; Hecker et al, 2021; B. W. Schuller, Batliner, et al, 2021; B.…”

The Acoustic Dissection of Cough: Diving Into Machine Listening-based COVID-19 Analysis and Detection

“…A promising approach for a screening tool fulfilling these requirements could be based on bioacoustic signals such as speech sounds or cough sounds. 11 , 12 , 13 , 14 Several studies have reported acoustic peculiarities in the speech of patients who have diseases associated with symptoms affecting anatomical correlates of speech production, such as bronchial asthma 15 , 16 or vocal cord disorders. 17 , 18 , 19 Differences in various acoustic parameters were also found in recent studies comparing speech samples of COVID-19 positive and COVID-19 negative individuals.…”