Section: New Software and Platforms

UMX

open-unmix

Keywords: Source Separation - Audio

Scientific Description: Implements state of the art audio/music source separation with DNNs.

This software is intended to serve as a reference in the domain. It has notably been the object of several scientific communications: 1. An Overview of Lead and Accompaniment Separation in Music https://hal-lirmm.ccsd.cnrs.fr/lirmm-01766781/ 2. Music separation with DNNs: making it work (ISMIR 2018 Tutorial) https://sigsep.github.io/ismir2018_tutorial/index.html#/cover

Functional Description: This software implements audio source separation with deep learning, using pytorch and tensorflow frameworks.

It comprises the code for both training and testing the separation networks, in a flexible manner.

Pre and post-processing around the actual deep neural nets include sophisticated specific multichannel filtering operations.

• Authors: Antoine Liutkus, Fabian Robert Stoter and Emmanuel Vincent

• Contact: Antoine Liutkus

• Publication: An Overview of Lead and Accompaniment Separation in Music