Extending Deep Rhythm for Tempo and Genre Estimation Using Complex Convolutions, Multitask Learning and Multi-input Network

Abstract

Tempo and genre are two inter-leaved aspects of music, genres are often associated to rhythm patterns which are played in specific tempo ranges.

In this paper, we focus on the Deep Rhythm system based on a harmonic representation of rhythm used as an input to a convolutional neural network.

To consider the relationships between frequency bands, we process complex-valued inputs through complex-convolutions.

We also study the joint estimation of tempo/genre using a multitask learning approach. 

Finally, we study the addition of a second input convolutional branch to the system applied to a mel-spectrogram input dedicated to the timbre.

This multi-input approach allows to improve the performances for tempo and genre estimation.

Keywords

multi-input, multitask, complex network, deep-learning, genre classification, tempo estimation

How to Cite

Foroughmand Aarabi, H. & Peeters, G., (2022) “Extending Deep Rhythm for Tempo and Genre Estimation Using Complex Convolutions, Multitask Learning and Multi-input Network”, Journal of Creative Music Systems 1(1). doi: https://doi.org/10.5920/jcms.887

414

Views

404

Downloads

Share

Authors

Hadrien Foroughmand Aarabi (IRCAM)
Geoffroy Peeters (LTCI - Télécom Paris - Institut Polytechnique)

Download

Issue

Dates

Licence

Creative Commons Attribution 4.0

Identifiers

Peer Review

This article has been peer reviewed.

File Checksums (MD5)

  • Post-Copyedit (Reduced APA bibliography, as style guide): ec330ee0b4a2b4b8ec5ca082bd28d830