Impromptune: Symbolic Music Generation with Relative Attention Mechanisms

Author

Connor J. Lennox, University of New Hampshire, DurhamFollow

Date of Award

Spring 2021

Project Type

Senior Honors Thesis

College or School

CEPS

Department

Computer Science

Degree Name

Bachelor of Science

First Advisor

Laura Dietz

Abstract

By combining attention-based mechanisms that have proved beneficial in the field of natural language processing with domain-specific knowledge about the structure of music, better predictions about piece continuations can be made. The goal of this work is to adapt current natural language processing techniques to a musical domain, and to generate new music by predicting continuations on a sequence of notes. An adaptation of traditional attention mechanisms to create a single prediction from sequential input is used to extend musical pieces by appending new elements repeatedly.

Recommended Citation

Lennox, Connor J., "Impromptune: Symbolic Music Generation with Relative Attention Mechanisms" (2021). Honors Theses and Capstones. 564.
https://scholars.unh.edu/honors/564