College of Charleston professors Bill Manaris, Yiorgos Vassilandonakis, and graduate student David Johnson have developed the Harmonic Navigator, a new computer-music project that combines artificial intelligence and human-computer interaction.
The Harmonic Navigator analyzes the harmony of instrumental or vocal musical works. Based on its library of musical works, it allows the user to explore existing harmonies and create new chord progressions. Once the user selects a starting chord, the system uses advanced artificial-intelligence techniques to explore its data base and suggest chords that may follow, ranging from surprising and tense to predictable and relaxed. This tool provides musicians and students the opportunity to compose music using large musical data sets, such as Bach chorales and John Coltrane jazz pieces.
[Related: See the Harmonic Navigator at work]
This is not the first collaborative project by computer science professor Manaris and music assistant professor Vassilandonakis. The two have previously worked on the Monterey Mirror project. Manaris’ other computer music projects include the Laptop Orchestra and Armonique.
“This technology is the result of a decade of research,” Manaris said. “Recently, I began collaborating with Yiorgos and David. We saw an opportunity to bring something completely novel to this exciting field which combines music, artificial intelligence, and human-computer interaction. I can’t wait to see how people use it, there are endless possibilities.”
The “magic” piece of the Harmonic Navigator, according to Manaris, is the artificial-intelligence genetic algorithm, which evolves different chord progressions aiming to match human musical aesthetics.
“It is amazing to observe the genetic algorithm,” Johnson said, “as it searches the harmonic space quickly and autonomously to find beautiful progressions that a user may not have found on their own.”
Manaris, Johnson and Vassilandonakis have already published one paper on the Harmonic Navigator for the Association for the Advancement of Artificial Intelligence, and have a second paper accepted to next year’s International Human-Computer Interaction conference. They are working on additional publications.
[Related: Watch Manaris and Vassilandonakis demonstrate their earlier project, Monterey Mirror]
“We’ve made something that makes harmonic transition and progression accessible to anyone, not just proficient musical theorists and artists,” Vassilandonakis said. “This program simplifies musical concepts while educating users and building upon their experience with the Harmonic Navigator.”
For more information, contact Bill Manaris at manarisb@cofc.edu, David Johnson at dsjohnso1@g.cofc.edu or Yiorgos Vassilandonakis at vassilandonakisy@cofc.edu.
