Speech separation is the process of separating multiple speakers from an audio recording. In this work we propose to separate the sources using a Speaker LOcalization Guided Deflation (SLOGD) approach wherein we estimate the sources iteratively. In each iteration we first estimate the location of the speaker and use it to estimate a mask corresponding to the localized speaker. The estimated source is removed from the mixture before estimating the location and mask of the next source. Experiments are conducted on a reverberated, noisy multichannel version of the well-studied WSJ-2MIX dataset using word error rate (WER) as a metric. The proposed method achieves a WER of 44.2%, a 34% relative improvement over the system without separation and 17% relative improvement over Conv-TasNet.
@article{arxiv.1910.11131,
title = {SLOGD: Speaker LOcation Guided Deflation approach to speech separation},
author = {Sunit Sivasankaran and Emmanuel Vincent and Dominique Fohr},
journal= {arXiv preprint arXiv:1910.11131},
year = {2019}
}