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

TL;DR

This paper explores how generative adversarial networks can be used to produce classical music that closely resembles compositions by renowned masters, addressing key technical and musical challenges.

Contribution

It introduces a novel approach combining deep learning and musical principles to improve AI-generated classical music quality.

Findings

Generated compositions are nearly indistinguishable from masterworks

Identified key parameters for high-quality music generation

Addressed implementation issues both programmatically and musically

Abstract

Current AI-generated music lacks fundamental principles of good compositional techniques. By narrowing down implementation issues both programmatically and musically, we can create a better understanding of what parameters are necessary for a generated composition nearly indistinguishable from that of a master composer.