_version_ 1866909303025172480
author Ma, Yinghao
Øland, Anders
Ragni, Anton
Del Sette, Bleiz MacSen
Saitis, Charalampos
Donahue, Chris
Lin, Chenghua
Plachouras, Christos
Benetos, Emmanouil
Shatri, Elona
Morreale, Fabio
Zhang, Ge
Fazekas, György
Xia, Gus
Zhang, Huan
Manco, Ilaria
Huang, Jiawen
Guinot, Julien
Lin, Liwei
Marinelli, Luca
Lam, Max W. Y.
Sharma, Megha
Kong, Qiuqiang
Dannenberg, Roger B.
Yuan, Ruibin
Wu, Shangda
Wu, Shih-Lun
Dai, Shuqi
Lei, Shun
Kang, Shiyin
Dixon, Simon
Chen, Wenhu
Huang, Wenhao
Du, Xingjian
Qu, Xingwei
Tan, Xu
Li, Yizhi
Tian, Zeyue
Wu, Zhiyong
Wu, Zhizheng
Ma, Ziyang
Wang, Ziyu
author_facet Ma, Yinghao
Øland, Anders
Ragni, Anton
Del Sette, Bleiz MacSen
Saitis, Charalampos
Donahue, Chris
Lin, Chenghua
Plachouras, Christos
Benetos, Emmanouil
Shatri, Elona
Morreale, Fabio
Zhang, Ge
Fazekas, György
Xia, Gus
Zhang, Huan
Manco, Ilaria
Huang, Jiawen
Guinot, Julien
Lin, Liwei
Marinelli, Luca
Lam, Max W. Y.
Sharma, Megha
Kong, Qiuqiang
Dannenberg, Roger B.
Yuan, Ruibin
Wu, Shangda
Wu, Shih-Lun
Dai, Shuqi
Lei, Shun
Kang, Shiyin
Dixon, Simon
Chen, Wenhu
Huang, Wenhao
Du, Xingjian
Qu, Xingwei
Tan, Xu
Li, Yizhi
Tian, Zeyue
Wu, Zhiyong
Wu, Zhizheng
Ma, Ziyang
Wang, Ziyu
contents In recent years, foundation models (FMs) such as large language models (LLMs) and latent diffusion models (LDMs) have profoundly impacted diverse sectors, including music. This comprehensive review examines state-of-the-art (SOTA) pre-trained models and foundation models in music, spanning from representation learning, generative learning and multimodal learning. We first contextualise the significance of music in various industries and trace the evolution of AI in music. By delineating the modalities targeted by foundation models, we discover many of the music representations are underexplored in FM development. Then, emphasis is placed on the lack of versatility of previous methods on diverse music applications, along with the potential of FMs in music understanding, generation and medical application. By comprehensively exploring the details of the model pre-training paradigm, architectural choices, tokenisation, finetuning methodologies and controllability, we emphasise the important topics that should have been well explored, like instruction tuning and in-context learning, scaling law and emergent ability, as well as long-sequence modelling etc. A dedicated section presents insights into music agents, accompanied by a thorough analysis of datasets and evaluations essential for pre-training and downstream tasks. Finally, by underscoring the vital importance of ethical considerations, we advocate that following research on FM for music should focus more on such issues as interpretability, transparency, human responsibility, and copyright issues. The paper offers insights into future challenges and trends on FMs for music, aiming to shape the trajectory of human-AI collaboration in the music realm.
format Preprint
id arxiv_https___arxiv_org_abs_2408_14340
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Foundation Models for Music: A Survey
Ma, Yinghao
Øland, Anders
Ragni, Anton
Del Sette, Bleiz MacSen
Saitis, Charalampos
Donahue, Chris
Lin, Chenghua
Plachouras, Christos
Benetos, Emmanouil
Shatri, Elona
Morreale, Fabio
Zhang, Ge
Fazekas, György
Xia, Gus
Zhang, Huan
Manco, Ilaria
Huang, Jiawen
Guinot, Julien
Lin, Liwei
Marinelli, Luca
Lam, Max W. Y.
Sharma, Megha
Kong, Qiuqiang
Dannenberg, Roger B.
Yuan, Ruibin
Wu, Shangda
Wu, Shih-Lun
Dai, Shuqi
Lei, Shun
Kang, Shiyin
Dixon, Simon
Chen, Wenhu
Huang, Wenhao
Du, Xingjian
Qu, Xingwei
Tan, Xu
Li, Yizhi
Tian, Zeyue
Wu, Zhiyong
Wu, Zhizheng
Ma, Ziyang
Wang, Ziyu
Sound
Artificial Intelligence
Computation and Language
Machine Learning
Audio and Speech Processing
In recent years, foundation models (FMs) such as large language models (LLMs) and latent diffusion models (LDMs) have profoundly impacted diverse sectors, including music. This comprehensive review examines state-of-the-art (SOTA) pre-trained models and foundation models in music, spanning from representation learning, generative learning and multimodal learning. We first contextualise the significance of music in various industries and trace the evolution of AI in music. By delineating the modalities targeted by foundation models, we discover many of the music representations are underexplored in FM development. Then, emphasis is placed on the lack of versatility of previous methods on diverse music applications, along with the potential of FMs in music understanding, generation and medical application. By comprehensively exploring the details of the model pre-training paradigm, architectural choices, tokenisation, finetuning methodologies and controllability, we emphasise the important topics that should have been well explored, like instruction tuning and in-context learning, scaling law and emergent ability, as well as long-sequence modelling etc. A dedicated section presents insights into music agents, accompanied by a thorough analysis of datasets and evaluations essential for pre-training and downstream tasks. Finally, by underscoring the vital importance of ethical considerations, we advocate that following research on FM for music should focus more on such issues as interpretability, transparency, human responsibility, and copyright issues. The paper offers insights into future challenges and trends on FMs for music, aiming to shape the trajectory of human-AI collaboration in the music realm.
title Foundation Models for Music: A Survey
topic Sound
Artificial Intelligence
Computation and Language
Machine Learning
Audio and Speech Processing
url https://arxiv.org/abs/2408.14340