Prof Björn Schuller

Professor of Artificial Intelligence

Imperial College London

United Kingdom

Mobile Digital Health for COVID-19: Ready, Set, ... Go?

​

Mobile Health holds the promise to serve almost anyone, anywhere, anytime. In times of crises like the present COVID-19 one, such promises are put to the test under highest time pressure, but also expectations involved. In this talk, we will see the example of computer audition and intelligent sensor-signal analysis on wearable devices for the diagnosis and monitoring in the context of COVID-19. This includes the assessment of COVID-19 and patient wellbeing from the speech signal to symptom onset detection from heartbeat measurement, to symptom histogram building in the wild from sounds such as of a congested nose, coughing, sneezing, or shortness of breath. Advanced deep learning enables immediate interpretation of such features, but also comes at the cost of a number of challenges: Key issues such as reliability, explainability, and privacy will be touched upon prior to discussing major hurdles on the road to broad usage “in the wild”. Ultimately, the question arises, whether AI and mobile Digital Health can keep the promises made?