FREQUENCY SMOOTHING WITH OPTIMAL WEIGHTING FOR IMPROVED SPEAKER LOCALIZATION

Hanan Beit-On, Tom Shlomo, Vladimir Tourbabin, Boaz Rafaely

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Many speaker localization methods apply frequency smoothing to decorrelate the direct-path signal and coherent reflections. However, in practice, only modest decorrelation is accomplished, which may result in performance loss. In this paper, frequency smoothing weights for improved decorrelation are derived and shown to be inversely proportional to the source signal power at the specified frequency. An experimental study demonstrates the added performance when incorporating the proposed weights in a direct path dominance test-based speaker localization method.

Original languageEnglish
Title of host publicationForum Acusticum 2023 - 10th Convention of the European Acoustics Association, EAA 2023
PublisherEuropean Acoustics Association, EAA
ISBN (Electronic)9788888942674
StatePublished - 1 Jan 2023
Event10th Convention of the European Acoustics Association, EAA 2023 - Torino, Italy
Duration: 11 Sep 202315 Sep 2023

Publication series

NameProceedings of Forum Acusticum
ISSN (Print)2221-3767

Conference

Conference10th Convention of the European Acoustics Association, EAA 2023
Country/TerritoryItaly
CityTorino
Period11/09/2315/09/23

Keywords

  • Speaker localization
  • coherent signal subspace
  • direct path dominance test
  • frequency smoothing

ASJC Scopus subject areas

  • Acoustics and Ultrasonics

Fingerprint

Dive into the research topics of 'FREQUENCY SMOOTHING WITH OPTIMAL WEIGHTING FOR IMPROVED SPEAKER LOCALIZATION'. Together they form a unique fingerprint.

Cite this