Amazon's AI uses a microphone array to localize multiple speakers in a room

A team of Amazon researchers proposes a machine learning model that's able to localize multiple sound sources from microphone data alone.