Diagnosing COVID-19 from Sounds
The aim of this research is to develop machine learning algorithms to automatically detect if a person is suffering of COVID-19, based primarily on sounds of their voice, their breathing and coughing. In order to enable this research we are launching a large scale, crowdsourced data collection through a mobile app. The app will collect some basic demographics and medical history data, as well as some voice samples (while you read text on the screen) through a questionnaire and a few seconds of breathing and coughing through the phone microphone. We will additionally collect one location sample. The app will also ask if you have tested positive for the virus. The app won't be tracking you and only collect this data when you actively interact with it. The data will be stored on University servers and be used solely for research purposes. We hope to release the dataset we are collecting to other researchers after the initial analysis. The app will not give medical advice and any reports of symptoms will not be responded to by medical assistance.

To help the algorithms, we need many samples so we hope a large number of participants will download and report their symptoms and uplaod sounds using the app. We are releasing the Android version and are working on the iOS one. To obtain and install the Android App, click here from your phone, or open Google Play Store from the phone and look for the Covid-19-Sounds App.

Please contact us at covid19sounds@gmail.com if you want clarification regarding the study.

The study has been approved by the Ethics Committee of the Department of Computer Science and Technology, University of Cambridge.

The study is partly funded by the European Research Council through Project EAR.