Rethinking recurrent latent variable model for music composition

Eunjeong Stella Kohl, Shlomo Dubnov, Dustin Wright

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

We present a model for capturing musical features and creating novel sequences of music, called the Convolutional-Variational Recurrent Neural Network. To generate sequential data, the model uses an encoder-decoder architecture with latent probabilistic connections to capture the hidden structure of music. Using the sequence-to-sequence model, our generative model can exploit samples from a prior distribution and generate a longer sequence of music. We compare the performance of our proposed model with other types of Neural Networks using the criteria of Information Rate that is implemented by Variable Markov Oracle, a method that allows statistical characterization of musical information dynamics and detection of motifs in a song. Our results suggest that the proposed model has a better statistical resemblance to the musical structure of the training data, which improves the creation of new sequences of music in the style of the originals.

Original languageEnglish
Title of host publication2018 IEEE 20th International Workshop on Multimedia Signal Processing, MMSP 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538660706
DOIs
StatePublished - 26 Nov 2018
Externally publishedYes
Event20th IEEE International Workshop on Multimedia Signal Processing, MMSP 2018 - Vancouver, Canada
Duration: 29 Aug 201831 Aug 2018

Publication series

Name2018 IEEE 20th International Workshop on Multimedia Signal Processing, MMSP 2018

Conference

Conference20th IEEE International Workshop on Multimedia Signal Processing, MMSP 2018
Country/TerritoryCanada
CityVancouver
Period29/08/1831/08/18

Fingerprint

Dive into the research topics of 'Rethinking recurrent latent variable model for music composition'. Together they form a unique fingerprint.

Cite this