Unifying Symbolic Music Arrangement: Track-Aware Reconstruction and Structured Tokenization

TL;DR

This paper introduces a unified, pre-trained symbolic music model capable of handling diverse multitrack arrangement tasks through a novel tokenization scheme and a segment-level reconstruction objective, outperforming specialized models.

Contribution

The authors propose REMI-z, a structured tokenization scheme, and a segment-level reconstruction approach enabling a single model to perform various arrangement tasks effectively.

Findings

Outperforms state-of-the-art models in band, piano, and drum arrangements

Enhances modeling efficiency and effectiveness for multiple tasks

Demonstrates broad applicability in symbolic music transformation

Abstract

We present a unified framework for automatic multitrack music arrangement that enables a single pre-trained symbolic music model to handle diverse arrangement scenarios, including reinterpretation, simplification, and additive generation. At its core is a segment-level reconstruction objective operating on token-level disentangled content and style, allowing for flexible any-to-any instrumentation transformations at inference time. To support track-wise modeling, we introduce REMI-z, a structured tokenization scheme for multitrack symbolic music that enhances modeling efficiency and effectiveness for both arrangement tasks and unconditional generation. Our method outperforms task-specific state-of-the-art models on representative tasks in different arrangement scenarios -- band arrangement, piano reduction, and drum arrangement, in both objective metrics and perceptual evaluations. Taken together, our framework demonstrates strong generality and suggests broader applicability in symbolic music-to-music transformation.