Support Material: Unsupervised Interpretable Representation Learningfor Singing Voice Separation
Stylianos Ioannis Mimilakis, Konstantinos Drossos, Gerald Schuller · Zenodo (CERN European Organization for Nuclear Research) · 2020
Support material for the manuscript entitled: "Unsupervised Interpretable Representation Learning for Singing Voice Separation". It contains binary files of the optimized and best performing models described in the corresponding manuscript. The attached file is to be used alongside the provided source code: https://github.com/Js-Mim/rl_singing_voice . We have provided an updated version of the results including additional insights.