Generating Music Algorithm with Deep Convolutional Generative Adversarial Networks
Authors:
- 1 May 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 576-580
Abstract
With the extensive development of deep learning, automatic composition has become a vanguard subject exercising the minds of scientists in the area of computer music. This paper proposes an advanced arithmetic for generating music using Generative Adversarial Networks (GAN). The music is divided into tracks and the note segment of tracks is expressed as a piano-roll, through trained a gan model which generator and discriminator continuous zero-sum game to generate a wonderful music integrallty. In most cases, Although GAN excel in image generation, the model adopts a full-channel lateral deep convolutional network structure according to the music data characteristics in this paper, generate music more in line with human hearing and aesthetics.Keywords
This publication has 4 references indexed in Scilit:
- Least Squares Generative Adversarial Networks2017
- Convolutional recurrent neural networks for music classification2017
- GSNs: generative stochastic networksInformation and Inference: A Journal of the IMA, 2016
- A Learning Algorithm for Boltzmann Machines*Cognitive Science, 1985