Music generation with variational recurrent autoencoder supported by history

Abstract

A new architecture of an artificial neural network that helps to generate 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. Combination of this architecture with filtering heuristics allows generating pseudo-live acoustically pleasing and melodically diverse music.

Cited by in corpus (3)

• PI-VAE: Physics-Informed Variational Auto-Encoder for stochastic differential equations
• Guess who? Multilingual approach for the automated generation of author-stylized poetry
• Variational Autoencoders for Anomaly Detection in Respiratory Sounds