JamBot: Music Theory Aware Chord Based Generation of Polyphonic Music with LSTMs

TL;DR

This paper introduces JamBot, a two-step LSTM-based system for generating harmonically pleasing polyphonic music with long-term structure, aligning with music theory principles like the circle of fifths.

Contribution

The paper presents a novel two-stage LSTM approach for music generation that incorporates music theory, notably learning the circle of fifths from data.

Findings

Generated music is harmonious with few dissonant notes.

The model captures long-term musical structure similar to human performance.

Chord embeddings reflect key music theory concepts like the circle of fifths.

Abstract

We propose a novel approach for the generation of polyphonic music based on LSTMs. We generate music in two steps. First, a chord LSTM predicts a chord progression based on a chord embedding. A second LSTM then generates polyphonic music from the predicted chord progression. The generated music sounds pleasing and harmonic, with only few dissonant notes. It has clear long-term structure that is similar to what a musician would play during a jam session. We show that our approach is sensible from a music theory perspective by evaluating the learned chord embeddings. Surprisingly, our simple model managed to extract the circle of fifths, an important tool in music theory, from the dataset.