Continuous Interactions & Generative Machine Learning for Digital Musical Instruments

This project is on integrating Machine Learning techniques to the design, creation, and performative aspect of Digital Musical Instruments (DMIs).

More specifically we are interested in how Generative Machine Learning can be used as a creative tool for musicians that supports their individual workflows as well as fosters a sense of artistic agency and creative ownership over their results. This research therefore is agnostic to musical techniques, genres, or performance style and is instead focused on the how musicians could choose to rely on this technology.

This project features a mixed-methods approach spanning multiple research methods. From the development of software prototypes, implementation of machine learning architectures to the construction of hardware interfaces, each are highly important research artifacts emerging as a result of this work. Furthermore, we evaluate each of these artifacts with user-studies, spanning both qualitative and quantitative methods.

Current Researchers

Current Publications

We aim to actively share our work with the larger research community. Here is a selected list of our publications to date.

Matthew Peachey; Sageev Oore; Joseph Malloch

Evaluating Low-Dimensional Latent Representations as a Creative Interface for Digital Synthesizers Proceedings Article

In: The Conference on AI Music Creativity (AIMC), 2025.

Links | BibTeX

Matthew Peachey; Sageev Oore; Joseph Malloch

Creating Latent Representations of Synthesizer Patches using Variational Autoencoders Proceedings Article

In: Proceedings of the 4th International Symposium on the Internet of Sounds (IS2), Pisa, Italy, 2023.

Links | BibTeX

Continuous Interactions & Generative Machine Learning for Digital Musical Instruments