Alexander Lerch teaches computers to listen to and comprehend music, which puts him at the crossroads of multiple disciplines: signal processing, machine learning, music psychology, and systematic musicology.

Lerch literally wrote the book on audio content analysis. His textbook, An Introduction to Audio Content Analysis, published in 2012, has helped define educational practice in his field. The second edition has just been released, and there is also a companion website.

“When we listen to music, we can extract a lot of information from it,” Lerch said. “We might recognize the artists, and we can tap with the tempo, and if we are musicians, we might hear the chord progressions, and so on. So, the question is, how can we enable a computer to extract this kind of information?”

Lerch and his colleagues in the Music Informatics Group use AI-driven methods to expand the limits of how a machine listens to and understands music and develop the next generation of analysis technology. 

“The work has many different potential applications, including music discovery, music education, creative listening, and music generation,” said Lerch, who co-founded zplane, a research-focused company that provides state-of-the-art music processing and analysis technology, and know-how, to the music industry. 

The work also is helping graduates from the School of Music land great careers with a who’s who list of companies, like Apple, Bose, Dolby, Spotify, and SiriusXM, among others.