Practical applicability of deep neural networks for overlapping speaker separation
Machine Learning
2019-12-20 v1 Sound
Audio and Speech Processing
Machine Learning
Abstract
This paper examines the applicability in realistic scenarios of two deep learning based solutions to the overlapping speaker separation problem. Firstly, we present experiments that show that these methods are applicable for a broad range of languages. Further experimentation indicates limited performance loss for untrained languages, when these have common features with the trained language(s). Secondly, it investigates how the methods deal with realistic background noise and proposes some modifications to better cope with these disturbances. The deep learning methods that will be examined are deep clustering and deep attractor networks.
Cite
@article{arxiv.1912.09261,
title = {Practical applicability of deep neural networks for overlapping speaker separation},
author = {Pieter Appeltans and Jeroen Zegers and Hugo Van hamme},
journal= {arXiv preprint arXiv:1912.09261},
year = {2019}
}
Comments
Interspeech 2019