Directed by Jason Palamara, the Machine Musician Lab at Herron School of Art and Design develops increasingly autonomous music systems that make music with humans and other machines. Fusing research conducted in artificial intelligence, machine learning, music cognition, and music information retrieval with concepts from traditional music theory, MML systems encourage humans to be better musicians by creating machines that make better music. MML machine musicians (both carbon and silicon-based) improve themselves over time. The MML is the developer of the AVATAR interactive music performance software, and the open-source MML repo of ML tools.
AVATAR
AVATAR is a machine-learning aided performance technology which plays along with a human player. By listening to live audio, avatar plays along dynamically, matching the playing style it hears around it in collaboration with a human musician. AVATAR is a research project jointly pursued by Jason Palamara and Scott Deal of the Tavel Lab.
Autonomous Music Systems
This course serves as an introduction to the intersection of music and artificial intelligence. Course topics, combining machine learning, artificial intelligence, data management, and automation with music theory concepts (low-level concepts like modes, harmonies, rhythms, and dynamics and high-level concepts like style). Students will build musical working digital systems via computer programming environments, culminating in an autonomously generated piece of music to be performed live.
DISEnsemble (Destructive / Inventive Systems Ensemble)
This ensemble uses improvisatory hardware and software hacking techniques as an approach to music making. By finding novel solutions to musical performance problems, students cultivate an attitude of creative freedom. Students experiment with improvisation, circuit-bending and destructive/creative instrument design while crafting a performance of live works. For the Machine Musician Lab, this student ensemble acts as a test subject to prove the Lab’s research findings in real-life performance situations.
The Artificial Intelligence Work Group (AIMWoG)
A collection of faculty and staff with a common interest in AI and music. AIMWoG meets biweekly to discuss topics in AI, ML, and how these things work in music, and we act as a safe open environment for research support, for group learning/activities, and open discussions about AI in music.
Current Students
Jerrelle Austin Sr., Ph.D. Student
Dana Goot, Ph.D. Student
Jessica Gunderson, M.S. Student
Matt Vice, Ph.D. Student