CS 229 Final Project: Predicting Imagined Meters in Musical Patterns from MEG Data
Authors: Aashna Shroff, Ben Limonchik, Zoe Robert
Abstract: Musical data is often interpreted within a metrical framework that integrates hierarchical timing information. It has been shown previously that listening to metronomes and imagining different meters such as march, waltz and hemiola (described below) modulated the resulting auditory evoked responses in the temporal lobe and motor- related brain areas such as the motor cortex, basal ganglia, and cerebellum. We wish to use MEG data of imagined meters to classify the metrical framework (i.e. march, waltz or hemiola) that the participating was imagining. The data for this project was collected in collaboration with Dr. Takako Fujioka of the Stanford CCRMA Lab. She has conducted several experiments to study the neural correlates of musical meters without the application of machine learning techniques. Our goal is to use advanced machine learning techniques to classify the imagined meters and gain quantitative insights from her findings.