He started with a guitar and a synthesiser; now Paris Smaragdis is teaching machines to understand sound beyond human hearing
MIT professor Paris Smaragdis is developing AI systems that can interpret sound, separate overlapping audio sources and help machines understand their surroundings. His journey from classical guitar and electronic music to machine learning has led to applications ranging from audio analysis to restoring archival Beatles recordings for the Emmy-winning documentary series The Beatles: Get Back. Now, his research focuses on improving audio processing, speech recognition and tools that empower musicians, with the aim of making AI a creative aid rather than a replacement.
