My wife has been following suspected COVID patients and has said since the beggining their cough is different than normal. Makes sense thay ML would work is this scenario.
But if this study serves as a PoC to back up that real-world observation, then this is quite a promising approach!