Discrete Diffusion Probabilistic Models for Symbolic Music Generation
Matthias Plasser, Silvan Peter, Gerhard Widmer
Abstract
Denoising Diffusion Probabilistic Models (DDPMs) have made great strides in generating high-quality samples in both discrete and continuous domains. However, Discrete DDPMs (D3PMs) have yet to be applied to the domain of Symbolic Music. This work presents the direct generation of Polyphonic Symbolic Music using D3PMs. Our model exhibits state-of-the-art sample quality, according to current quantitative evaluation metrics, and allows for flexible infilling at the note level. We further show, that our models are accessible to post-hoc classifier guidance, widening the scope of possible applications. However, we also cast a critical view on quantitative evaluation of music sample quality via statistical metrics, and present a simple algorithm that can confound our metrics with completely spurious, non-musical samples.
BibTeX
@inproceedings{ijcai2023p648,
title = {Discrete Diffusion Probabilistic Models for Symbolic Music Generation},
author = {Plasser, Matthias and Peter, Silvan and Widmer, Gerhard},
booktitle = {Proceedings of the Thirty-Second International Joint Conference on
Artificial Intelligence, {IJCAI-23}},
publisher = {International Joint Conferences on Artificial Intelligence Organization},
editor = {Edith Elkind},
pages = {5842--5850},
year = {2023},
month = {8},
note = {AI and Arts},
doi = {10.24963/ijcai.2023/648},
url = {https://doi.org/10.24963/ijcai.2023/648},
}