Machine Learning for Music Composition
I'm interested in modelling symbolic music using machine learning methods such as deep neural networks and the Long Short-Term Memory
cell. The goal is to model the note choices (as represented by sheet music or midi files), rather than modelling audio waveforms. This topic has generated a lot of interest lately, with several notable examples including the exciting Magenta
project from the Google Brain team.
Please take a look at my page for more information:
I have several interesting projects to offer, and I'm open to suggestions for new ones. We would need to negotiate this based on your background and level of enthusiasm. Here are a few candidates:
- machine learning modelling
- interface design, to prototype tools which allow human composers to exploit our existing models
- evaluation framework, including the psychophysical design and software implementation of online web surveys for music evaluation
These should be really fun projects. I'm looking forward to hearing from you.