Signal-domain representation of symbolic music for learning embedding spaces

TL;DR

This paper introduces a novel signal-based representation of polyphonic symbolic music that improves feature learning, reconstruction, and generation capabilities in machine learning models, validated through synthetic data evaluation and extensive benchmarking.

Contribution

The paper presents a new continuous signal representation for polyphonic music, enhancing feature learning and generation, with comprehensive evaluation and benchmarking against existing methods.

Findings

Better reconstruction of polyphonic music

More disentangled and meaningful features

Improved music generation quality

Abstract

A key aspect of machine learning models lies in their ability to learn efficient intermediate features. However, the input representation plays a crucial role in this process, and polyphonic musical scores remain a particularly complex type of information. In this paper, we introduce a novel representation of symbolic music data, which transforms a polyphonic score into a continuous signal. We evaluate the ability to learn meaningful features from this representation from a musical point of view. Hence, we introduce an evaluation method relying on principled generation of synthetic data. Finally, to test our proposed representation we conduct an extensive benchmark against recent polyphonic symbolic representations. We show that our signal-like representation leads to better reconstruction and disentangled features. This improvement is reflected in the metric properties and in the generation ability of the space learned from our signal-like representation according to music theory properties.