Incorporating Structure and Chord Constraints in Symbolic Transformer-based Melodic Harmonization

TL;DR

This paper introduces B*, an algorithm that integrates chord constraints into transformer-based melodic harmonization, addressing the challenge of satisfying specific harmonic requirements during music generation.

Contribution

It proposes a novel algorithm combining beam search and A* to incorporate chord constraints into transformer models for melodic harmonization.

Findings

The B* algorithm effectively enforces chord constraints during generation.

The approach highlights the complexity of constrained music generation.

Potential for heuristic improvements in future work.

Abstract

Transformer architectures offer significant advantages regarding the generation of symbolic music; their capabilities for incorporating user preferences toward what they generate is being studied under many aspects. This paper studies the inclusion of predefined chord constraints in melodic harmonization, i.e., where a desired chord at a specific location is provided along with the melody as inputs and the autoregressive transformer model needs to incorporate the chord in the harmonization that it generates. The peculiarities of involving such constraints is discussed and an algorithm is proposed for tackling this task. This algorithm is called B* and it combines aspects of beam search and A* along with backtracking to force pretrained transformers to satisfy the chord constraints, at the correct onset position within the correct bar. The algorithm is brute-force and has exponential complexity in the worst case; however, this paper is a first attempt to highlight the difficulties of the problem and proposes an algorithm that offers many possibilities for improvements since it accommodates the involvement of heuristics.