Chord-conditioned Melody and Bass Generation

TL;DR

This paper evaluates five Transformer-based models for chord-conditioned melody and bass generation, demonstrating that chord-conditioning enhances stylistic pitch content and chord tone usage, especially with bass-first generation.

Contribution

It systematically compares different chord-conditioning strategies in Transformer models for music generation, highlighting the effectiveness of bass-first conditioning.

Findings

Chord-conditioning improves stylistic pitch content.

Bass-first model best replicates chord tone usage.

Transformer strategies vary in effectiveness for music generation.

Abstract

We evaluate five Transformer-based strategies for chord-conditioned melody and bass generation using a set of music theory-motivated metrics capturing pitch content, pitch interval size, and chord tone usage. The evaluated models include (1) no chord conditioning, (2) independent line chord-conditioned generation, (3) bass-first chord-conditioned generation, (4) melody-first chord-conditioned generation, and (5) chord-conditioned co-generation. We show that chord-conditioning improves the replication of stylistic pitch content and chord tone usage characteristics, particularly for the bass-first model.