Music generation with variational recurrent autoencoder supported by history

Ivan P Yamshchikov,Alexey Tikhonov

doi:10.1007/s42452-020-03715-w

Abstract

A new artificial neural network architecture that helps generating longer melodic patterns is introduced alongside with methods for post-generation filtering. The proposed approach, called variational autoencoder supported by history, is based on a recurrent highway gated network combined with a variational autoencoder. The combination of this architecture with filtering heuristics allows the generation of pseudo-live, acoustically pleasing, melodically diverse music.

Highlights

The rapid progress of artificial neural networks is gradually erasing the border between the arts and the sciences
A variety of different recurrent neural networks such as hierarchical Recurrent neural networks (RNNs), gated RNN, long short-term memory (LSTM) network, and recurrent highway network were successfully used for music generation in [4,5,6, 10, 20, 28] or [23]
The contribution of this paper is twofold: (1) we suggest a new architecture for the algorithmic composition of monotonic music called Variational Recurrent Autoencoder Supported by History (VRASH) and (2) we demonstrate that, when paired with simple filtering heuristics, VRASH can generate pseudo-live, acoustically pleasing, melodically diverse melodies

Summary

Introduction

The rapid progress of artificial neural networks is gradually erasing the border between the arts and the sciences. A significant number of results demonstrate how areas previously regarded as entirely human due to their creative or intuitive nature are being opened up for algorithmic approaches [24]. There were a number of attempts to automate the process of music composition long before the era of artificial neural networks. Well-developed theory of music inspired a number of heuristic approaches to automated music composition. In the middle of the twentieth century, a Markov-chain approach for music composition was developed in [8]. Despite these advances, Lin and Tegmark [14] have demonstrated that music, as well as some other types of human-generated discrete time series, tends to have long-distance dependencies that cannot be captured by models based on Markov chains.

Methods

Findings

Discussion

Conclusion