VoiceFilter: Targeted Voice Separation by Speaker-Conditioned Spectrogram Masking
Audio and Speech Processing
2019-06-20 v6 Machine Learning
Signal Processing
Machine Learning
Abstract
In this paper, we present a novel system that separates the voice of a target speaker from multi-speaker signals, by making use of a reference signal from the target speaker. We achieve this by training two separate neural networks: (1) A speaker recognition network that produces speaker-discriminative embeddings; (2) A spectrogram masking network that takes both noisy spectrogram and speaker embedding as input, and produces a mask. Our system significantly reduces the speech recognition WER on multi-speaker signals, with minimal WER degradation on single-speaker signals.
Cite
@article{arxiv.1810.04826,
title = {VoiceFilter: Targeted Voice Separation by Speaker-Conditioned Spectrogram Masking},
author = {Quan Wang and Hannah Muckenhirn and Kevin Wilson and Prashant Sridhar and Zelin Wu and John Hershey and Rif A. Saurous and Ron J. Weiss and Ye Jia and Ignacio Lopez Moreno},
journal= {arXiv preprint arXiv:1810.04826},
year = {2019}
}
Comments
To appear in Interspeech 2019