Abstract

Far-field speech capture systems rely on microphone arrays to spatially filter sound, attenuating unwanted interference and noise and enhancing a speech signal of interest. To design effective spatial filters, we must first estimate the acoustic transfer functions between the source and the microphones. It is difficult to estimate these transfer functions if the source signals are unknown. However, in systems that are activated by a particular speech phrase, we can use that phrase as a pilot signal to estimate the relative transfer functions. Here, we propose a method to estimate relative transfer functions from known speech phrases in the presence of background noise and interference using template matching and time-frequency masking. We find that the proposed method can outperform conventional estimation techniques, but its performance depends on the characteristics of the speech phrase.

Original languageEnglish (US)
Title of host publicationLatent Variable Analysis and Signal Separation - 14th International Conference, LVA/ICA 2018, Proceedings
EditorsSharon Gannot, Yannick Deville, Russell Mason, Mark D. Plumbley, Dominic Ward
PublisherSpringer-Verlag Berlin Heidelberg
Pages238-247
Number of pages10
ISBN (Print)9783319937632
DOIs
StatePublished - 2018
Event14th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2018 - Guildford, United Kingdom
Duration: Jul 2 2018Jul 5 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10891 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other14th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2018
CountryUnited Kingdom
CityGuildford
Period7/2/187/5/18

Keywords

  • Keyword spotting
  • Microphone array
  • Multichannel source separation
  • Relative transfer function

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Relative transfer function estimation from speech keywords'. Together they form a unique fingerprint.

Cite this