StemGen: A music generation model that listens

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Parker, Julian D., Spijkervet, Janne, Kosta, Katerina, Yesiler, Furkan, Kuznetsov, Boris, Wang, Ju-Chiang, Avent, Matt, Chen, Jitong, Le, Duc
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916091389804544
author Parker, Julian D.
Spijkervet, Janne
Kosta, Katerina
Yesiler, Furkan
Kuznetsov, Boris
Wang, Ju-Chiang
Avent, Matt
Chen, Jitong
Le, Duc
author_facet Parker, Julian D.
Spijkervet, Janne
Kosta, Katerina
Yesiler, Furkan
Kuznetsov, Boris
Wang, Ju-Chiang
Avent, Matt
Chen, Jitong
Le, Duc
contents End-to-end generation of musical audio using deep learning techniques has seen an explosion of activity recently. However, most models concentrate on generating fully mixed music in response to abstract conditioning information. In this work, we present an alternative paradigm for producing music generation models that can listen and respond to musical context. We describe how such a model can be constructed using a non-autoregressive, transformer-based model architecture and present a number of novel architectural and sampling improvements. We train the described architecture on both an open-source and a proprietary dataset. We evaluate the produced models using standard quality metrics and a new approach based on music information retrieval descriptors. The resulting model reaches the audio quality of state-of-the-art text-conditioned models, as well as exhibiting strong musical coherence with its context.
format Preprint
id arxiv_https___arxiv_org_abs_2312_08723
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle StemGen: A music generation model that listens
Parker, Julian D.
Spijkervet, Janne
Kosta, Katerina
Yesiler, Furkan
Kuznetsov, Boris
Wang, Ju-Chiang
Avent, Matt
Chen, Jitong
Le, Duc
Sound
Machine Learning
Audio and Speech Processing
End-to-end generation of musical audio using deep learning techniques has seen an explosion of activity recently. However, most models concentrate on generating fully mixed music in response to abstract conditioning information. In this work, we present an alternative paradigm for producing music generation models that can listen and respond to musical context. We describe how such a model can be constructed using a non-autoregressive, transformer-based model architecture and present a number of novel architectural and sampling improvements. We train the described architecture on both an open-source and a proprietary dataset. We evaluate the produced models using standard quality metrics and a new approach based on music information retrieval descriptors. The resulting model reaches the audio quality of state-of-the-art text-conditioned models, as well as exhibiting strong musical coherence with its context.
title StemGen: A music generation model that listens
topic Sound
Machine Learning
Audio and Speech Processing
url https://arxiv.org/abs/2312.08723