Generative Adversarial Source Separation
Sound
2017-10-31 v1 Machine Learning
Neural and Evolutionary Computing
Machine Learning
Abstract
Generative source separation methods such as non-negative matrix factorization (NMF) or auto-encoders, rely on the assumption of an output probability density. Generative Adversarial Networks (GANs) can learn data distributions without needing a parametric assumption on the output density. We show on a speech source separation experiment that, a multi-layer perceptron trained with a Wasserstein-GAN formulation outperforms NMF, auto-encoders trained with maximum likelihood, and variational auto-encoders in terms of source to distortion ratio.
Keywords
Cite
@article{arxiv.1710.10779,
title = {Generative Adversarial Source Separation},
author = {Cem Subakan and Paris Smaragdis},
journal= {arXiv preprint arXiv:1710.10779},
year = {2017}
}