Generative Deep Learning for Virtuosic Classical Music: Generative Adversarial Networks as Renowned Composers

Szelogowski, Daniel

doi:10.48550/arxiv.2101.00169

Cited by 1 publication

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another solution was to use the Autokeras Python library, which performs a full neural architecture search using multiple combinations of different layers and hyperparameters to find the best model -nonetheless, the absolute best model from this system only achieved 34% validation accuracy at best. We then attempted combining deep learning algorithms with machine learning, such as SVM (Support Vector Machines, a supervised learning algorithm that classifies data with the widest possible margin between each class [19], of which the Neural SVM only achieved 23% test accuracy -see Appendix D) and finally neural decision tree hybrids.…”

Section: Comparison Of Form Analyzing Modelsmentioning

confidence: 99%

“…• Generative Adversarial Network (GAN) -A special neural network comprised of a Generative model (for supervised tasks) and a Discriminative model (for unsupervised tasks, judges the accuracy of the Generative model output) that uses unlabeled datasets to select samples from a distribution of the data developed from competition between the two models to generate new, realistic data; used for generating media such as imagery, video, language, and audio [19].…”

Section: Appendicesmentioning

confidence: 99%

See 1 more Smart Citation

Deep Learning for Musical Form: Recognition and Analysis

Szelogowski¹

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Musical form analysis is a rigorous task that frequently challenges the expertise of human analysts and signal processing algorithms alike. While numerous systems have been proposed to perform the tasks of musical segmentation, genre classification, and single-label segment classification in popular music, none have specifically focused on the analytical process used by classical musicians. Classical music form analysis facilitates a combination of these tasks, including form classification, structural segmentation, and multilabel large- and small-segment classification – tasks that lack feasible algorithms, machine learning models, and extensive research. Form analysis has many applications in the world of music, and a viable analytical system would greatly benefit performing musicians and academic researchers, both in musicology and signal processing. As well, current datasets used for related research tasks lack standardized analytical conventions, including form classification, and suffer from erroneous annotations and extensibility due to the data sources used for the music. In this thesis, we propose a new system to perform the task of automatic musical form analysis using deep learning models, as well as a new standardized dataset.

show abstract

Section: Comparison Of Form Analyzing Modelsmentioning

confidence: 99%

Section: Appendicesmentioning

confidence: 99%