English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Generative AI models for music and the arts in general are increasinglycomplex and hard to understand. The field of eXplainable AI (XAI) seeks to makecomplex and opaque AI models such as neural networks more understandable topeople. One approach to making generative AI models more understandable is toimpose a small number of semantically meaningful attributes on generative AImodels. This paper contributes a systematic examination of the impact thatdifferent combinations of Variational Auto-Encoder models (MeasureVAE andAdversarialVAE), configurations of latent space in the AI model (from 4 to 256latent dimensions), and training datasets (Irish folk, Turkish folk, Classical,and pop) have on music generation performance when 2 or 4 meaningful musicalattributes are imposed on the generative model. To date there have been nosystematic comparisons of such models at this level of combinatorial detail.Our findings show that MeasureVAE has better reconstruction performance thanAdversarialVAE which has better musical attribute independence. Resultsdemonstrate that MeasureVAE was able to generate music across music genres withinterpretable musical dimensions of control, and performs best with lowcomplexity music such a pop and rock. We recommend that a 32 or 64 latentdimensional space is optimal for 4 regularised dimensions when using MeasureVAEto generate music across genres. Our results are the first detailed comparisonsof configurations of state-of-the-art generative AI models for music and can beused to help select and configure AI models, musical features, and datasets formore understandable generation of music.

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