Improved hybrid microphone array post-filter by integrating a robust speech absence probability estimator for speech enhancement

Junfeng Li, Masato Akagi, Yoiti Suzuki

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

To improve the performance of multi-channel speech enhancement algorithms, we previously proposed a hybrid Wiener postfilter for microphone arrays under the assumption of a diffuse noise field [4]. In this paper, considering the speech presence uncertainty, we further improve the hybrid post-filter presented before by integrating a novel robust estimator for the a priori speech absence probability, which makes full use of the correlation characteristics of the noises on different microphone pairs and hence offers the much more accurate speech absence probability estimates. The effectiveness of this improved hybrid post-filter was finally confirmed by the experiments using multi-channel recordings in various car environments.

Original languageEnglish
Title of host publicationINTERSPEECH 2006 and 9th International Conference on Spoken Language Processing, INTERSPEECH 2006 - ICSLP
PublisherInternational Speech Communication Association
Pages2130-2133
Number of pages4
Volume5
ISBN (Print)9781604234497
Publication statusPublished - 2006 Jan 1
EventINTERSPEECH 2006 and 9th International Conference on Spoken Language Processing, INTERSPEECH 2006 - ICSLP - Pittsburgh, PA, United States
Duration: 2006 Sep 172006 Sep 21

Other

OtherINTERSPEECH 2006 and 9th International Conference on Spoken Language Processing, INTERSPEECH 2006 - ICSLP
CountryUnited States
CityPittsburgh, PA
Period06/9/1706/9/21

Keywords

  • Microphone array
  • Post-filtering
  • Speech absence probability
  • Speech enhancement

ASJC Scopus subject areas

  • Computer Science(all)

Fingerprint Dive into the research topics of 'Improved hybrid microphone array post-filter by integrating a robust speech absence probability estimator for speech enhancement'. Together they form a unique fingerprint.

Cite this