Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI

doi:10.1007/s11633-023-1457-1

CORC > 自动化研究所 > 中国科学院自动化研究所 > 学术期刊 > International Journal of Automation and Computing

	Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI
	Nick Bryan-Kinns 1
刊名	Machine Intelligence Research
	2024
卷号	21 期号:1 页码:29-45
关键词	Variational auto-encoder, explainable AI (XAI), generative music, musical features, datasets
ISSN号	2731-538X
DOI	10.1007/s11633-023-1457-1
英文摘要	Generative AI models for music and the arts in general are increasingly complex and hard to understand. The field of explainable AI (XAI) seeks to make complex and opaque AI models such as neural networks more understandable to people. One approach to making generative AI models more understandable is to impose a small number of semantically meaningful attributes on generative AI models. This paper contributes a systematic examination of the impact that different combinations of variational auto-encoder models (measureVAE and adversarialVAE), configurations of latent space in the AI model (from 4 to 256 latent dimensions), and training datasets (Irish folk, Turkish folk, classical, and pop) have on music generation performance when 2 or 4 meaningful musical attributes are imposed on the generative model. To date, there have been no systematic comparisons of such models at this level of combinatorial detail. Our findings show that measureVAE has better reconstruction performance than adversarialVAE which has better musical attribute independence. Results demonstrate that measureVAE was able to generate music across music genres with interpretable musical dimensions of control, and performs best with low complexity music such as pop and rock. We recommend that a 32 or 64 latent dimensional space is optimal for 4 regularised dimensions when using measureVAE to generate music across genres. Our results are the first detailed comparisons of configurations of state-of-the-art generative AI models for music and can be used to help select and configure AI models, musical features, and datasets for more understandable generation of music.
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/54573]
专题	自动化研究所_学术期刊_International Journal of Automation and Computing
作者单位	1.School of Electronic Engineering and Computer Science, Queen Mary University of London, London E1 4NS, UK 2.Computer Science Department, Carleton College, Northfield MN 55057, USA
推荐引用方式 GB/T 7714	Nick Bryan-Kinns. Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI[J]. Machine Intelligence Research,2024,21(1):29-45.
APA	Nick Bryan-Kinns.(2024).Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI.Machine Intelligence Research,21(1),29-45.
MLA	Nick Bryan-Kinns."Exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music Explainable AI".Machine Intelligence Research 21.1(2024):29-45.