Differential Music: Automated Music Generation Using LSTM Networks with Representation Based on Melodic and Harmonic Intervals

TL;DR

This paper introduces a novel LSTM-based music generation method that encodes melodies and chords through interval movements rather than absolute pitches, resulting in more musical and tonal compositions.

Contribution

The paper proposes a new interval-based encoding scheme for music generation with LSTM networks, emphasizing movement over absolute pitch for improved musicality.

Findings

Generated music sounds musical and tonal

The method exhibits excessive modulations

Encoding based on intervals influences tonal stability

Abstract

This paper presents a generative AI model for automated music composition with LSTM networks that takes a novel approach at encoding musical information which is based on movement in music rather than absolute pitch. Melodies are encoded as a series of intervals rather than a series of pitches, and chords are encoded as the set of intervals that each chord note makes with the melody at each timestep. Experimental results show promise as they sound musical and tonal. There are also weaknesses to this method, mainly excessive modulations in the compositions, but that is expected from the nature of the encoding. This issue is discussed later in the paper and is a potential topic for future work.