Table 2.

Summary of the datasets used for COVID-19 classification. Cough, breath and speech signals were extracted from the Coswara, ComParE and Sarcos datasets. COVID-19 positive subjects are under-represented in all three.

Type	Dataset	Sampling Rate	Label	Subjects	Total audio	Average per subject	Standard deviation
Cough	Coswara	44.1 kHz	COVID-19 Positive	92	4.24 min	2.77 s	1.62 s
			Healthy	1079	0.98 h	3.26 s	1.66 s
			Total	1171	1.05 h	3.22 s	1.67 s
	ComParE	16 kHz	COVID-19 Positive	119	13.43 min	6.77 s	2.11 s
			Healthy	398	40.89 min	6.16 s	2.26 s
			Total	517	54.32 min	6.31 s	2.24 s
	Sarcos	44.1 kHz	COVID-19 Positive	18	0.87 min	2.91 s	2.23 s
			COVID-19 Negative	26	1.57 min	3.63 s	2.75 s
			Total	44	2.45 min	3.34 s	2.53 s
Breath	Coswara	44.1 kHz	COVID-19 Positive	88	8.58 min	5.85 s	5.05 s
			Healthy	1062	2.77 h	9.39 s	5.23 s
			Total	1150	2.92 h	9.126 s	5.29 s
Speech	Coswara (normal)	44.1 kHz	COVID-19 Positive	88	12.42 min	8.47 s	4.27 s
			Healthy	1077	2.99 h	9.99 s	3.09 s
			Total	1165	3.19 h	9.88 s	3.22 s
	Coswara (fast)	44.1 kHz	COVID-19 Positive	85	7.62 min	5.38 s	2.76 s
			Healthy	1074	1.91 h	6.39 s	1.77 s
			Total	1159	2.03 h	6.31 s	1.88 s
	ComParE	16 kHz	COVID-19 Positive	214	44.02 min	12.34 s	5.35 s
			Healthy	396	1.46 h	13.25 s	4.67 s
			Total	610	2.19 h	12.93 s	4.93 s