The current coronavirus pandemic is sweeping the globe and as more and more people get sick, test kits are running short. That’s why a Carnegie Mellon University research team created a simpler, more open way to potentially identify virus signs: Voice Detection.
The team has been working on this solution for some time, and now an early version of a web app called the COVID Voice Detector has been released. The app uses an algorithm built in-house to attempt to identify signs of infection in the user’s voice.
All you need is a smartphone or a device with a microphone, if you want to try it for yourself. Once the account setup process is complete, the test will prompt you to cough multiple times, record a series of vowel sounds, and recite the alphabet. The app then offers a score indicating how likely the algorithm thinks that you have contracted COVID-19.
Before we start, we need to make it clear: This software is not a substitute for a clinical coronavirus study, and should not be viewed as such. If you encounter any virus symptoms please contact your doctor. Again do not use this online app instead of a doctor’s appointment.
Although COVID-19 patients ‘cough is very distinctive, the challenge for the research team is to get audio samples from confirmed patients for training the algorithm. Since Carnegie Mellon’s campus is closed, the team has been operating from home so it has been reaching out to colleagues around the world who have contracted not only COVID-19 but other viruses as well. They have also used news interviews to expand their data collection with patients.
Because of this intermittent sampling, the team says it’s hard to measure the exactness of the app’s current version. Yeah, the more people who use the app — sick or not — the more data the team needs to prepare.
It may be a very useful tool even in its early stages of helping people monitor coronavirus spread if they don’t have access to an official check.