The Second DiCOVA Challenge: Dataset and performance analysis for COVID-19 diagnosis using acoustics

The Second DiCOVA Challenge aims at accelerating the research in diagnosing COVID-19 using acoustics (DiCOVA), a topic at the intersection of acoustics signal processing, machine learning, and healthcare. This challenge is an open call to researchers to analyze a dataset of audio recordings, collected from individuals with and without COVID-19, for a two-class classification. The development set audio recordings correspond to breathing, cough, and speech sound samples collected from 965 (172 COVID) individuals. The challenge features four tracks, one associated with each sound category and a fourth fusion track allowing experimentation with combination of the individual sound categories. In this paper, we introduce the challenge and provide a detailed description of the task and a baseline system.