A framework to compare music generative models using automatic evaluation metrics extended to rhythm

TL;DR

This paper extends an existing music evaluation framework to include rhythm, enabling automatic quantitative assessment of monophonic music generated by RNN models with different memory cells.

Contribution

It introduces rhythm support into an existing evaluation framework and compares RNN memory cells for monophonic music generation.

Findings

Rhythm support improves evaluation accuracy.

Different RNN cells show varying performance.

Framework effectively assesses generated music quality.

Abstract

To train a machine learning model is necessary to take numerous decisions about many options for each process involved, in the field of sequence generation and more specifically of music composition, the nature of the problem helps to narrow the options but at the same time, some other options appear for specific challenges. This paper takes the framework proposed in a previous research that did not consider rhythm to make a series of design decisions, then, rhythm support is added to evaluate the performance of two RNN memory cells in the creation of monophonic music. The model considers the handling of music transposition and the framework evaluates the quality of the generated pieces using automatic quantitative metrics based on geometry which have rhythm support added as well.