English

Multitrack Music Transformer

Sound 2023-05-26 v4 Artificial Intelligence Machine Learning Multimedia Audio and Speech Processing

Abstract

Existing approaches for generating multitrack music with transformer models have been limited in terms of the number of instruments, the length of the music segments and slow inference. This is partly due to the memory requirements of the lengthy input sequences necessitated by existing representations. In this work, we propose a new multitrack music representation that allows a diverse set of instruments while keeping a short sequence length. Our proposed Multitrack Music Transformer (MMT) achieves comparable performance with state-of-the-art systems, landing in between two recently proposed models in a subjective listening test, while achieving substantial speedups and memory reductions over both, making the method attractive for real time improvisation or near real time creative applications. Further, we propose a new measure for analyzing musical self-attention and show that the trained model attends more to notes that form a consonant interval with the current note and to notes that are 4N beats away from the current step.

Keywords

Cite

@article{arxiv.2207.06983,
  title  = {Multitrack Music Transformer},
  author = {Hao-Wen Dong and Ke Chen and Shlomo Dubnov and Julian McAuley and Taylor Berg-Kirkpatrick},
  journal= {arXiv preprint arXiv:2207.06983},
  year   = {2023}
}

Comments

Accepted by ICASSP 2023. Demo: https://salu133445.github.io/mmt/ . Code: https://github.com/salu133445/mmt

R2 v1 2026-06-25T00:55:09.211Z