Bayesian extension of MUSIC for sound source localization and tracking

Takuma Otsuka, Kazuhiro Nakadai, Tetsuya Ogata, Hiroshi G. Okuno

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

8 Citations (Scopus)

Abstract

This paper presents a Bayesian extension of MUSIC-based sound source localization (SSL) and tracking method. SSL is important for distant speech enhancement and simultaneous speech separation for improving speech recognition, as well as for auditory scene analysis by mobile robots. One of the draw- backs of existing SSL methods is the necessity of careful param- eter tunings, e.g., the sound source detection threshold depend- ing on the reverberation time and the number of sources. Our contribution consists of (1) automatic parameter estimation in the variational Bayesian framework and (2) tracking of sound sources with reliability. Experimental results demonstrate our method robustly tracks multiple sound sources in a reverberant environment with RT20 = 840 (ms).

Original languageEnglish
Title of host publicationProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Pages3109-3112
Number of pages4
Publication statusPublished - 2011
Externally publishedYes
Event12th Annual Conference of the International Speech Communication Association, INTERSPEECH 2011 - Florence, Italy
Duration: 2011 Aug 272011 Aug 31

Other

Other12th Annual Conference of the International Speech Communication Association, INTERSPEECH 2011
CountryItaly
CityFlorence
Period11/8/2711/8/31

Fingerprint

Source Localization
Acoustic waves
Scene Analysis
Speech Enhancement
Speech enhancement
Reverberation
Parameter Tuning
Speech Recognition
Speech recognition
Mobile Robot
Parameter estimation
Mobile robots
Parameter Estimation
Sound
Localization
Tuning
Experimental Results
Demonstrate

Keywords

  • MUSIC algorithm
  • Particle filter
  • Simultaneous sound source localization
  • Variational Bayes

ASJC Scopus subject areas

  • Language and Linguistics
  • Human-Computer Interaction
  • Signal Processing
  • Software
  • Modelling and Simulation

Cite this

Otsuka, T., Nakadai, K., Ogata, T., & Okuno, H. G. (2011). Bayesian extension of MUSIC for sound source localization and tracking. In Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH (pp. 3109-3112)

Bayesian extension of MUSIC for sound source localization and tracking. / Otsuka, Takuma; Nakadai, Kazuhiro; Ogata, Tetsuya; Okuno, Hiroshi G.

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. 2011. p. 3109-3112.

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

Otsuka, T, Nakadai, K, Ogata, T & Okuno, HG 2011, Bayesian extension of MUSIC for sound source localization and tracking. in Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. pp. 3109-3112, 12th Annual Conference of the International Speech Communication Association, INTERSPEECH 2011, Florence, Italy, 11/8/27.
Otsuka T, Nakadai K, Ogata T, Okuno HG. Bayesian extension of MUSIC for sound source localization and tracking. In Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. 2011. p. 3109-3112
Otsuka, Takuma ; Nakadai, Kazuhiro ; Ogata, Tetsuya ; Okuno, Hiroshi G. / Bayesian extension of MUSIC for sound source localization and tracking. Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH. 2011. pp. 3109-3112
@inproceedings{41882f39848b43c880cd9183440f5047,
title = "Bayesian extension of MUSIC for sound source localization and tracking",
abstract = "This paper presents a Bayesian extension of MUSIC-based sound source localization (SSL) and tracking method. SSL is important for distant speech enhancement and simultaneous speech separation for improving speech recognition, as well as for auditory scene analysis by mobile robots. One of the draw- backs of existing SSL methods is the necessity of careful param- eter tunings, e.g., the sound source detection threshold depend- ing on the reverberation time and the number of sources. Our contribution consists of (1) automatic parameter estimation in the variational Bayesian framework and (2) tracking of sound sources with reliability. Experimental results demonstrate our method robustly tracks multiple sound sources in a reverberant environment with RT20 = 840 (ms).",
keywords = "MUSIC algorithm, Particle filter, Simultaneous sound source localization, Variational Bayes",
author = "Takuma Otsuka and Kazuhiro Nakadai and Tetsuya Ogata and Okuno, {Hiroshi G.}",
year = "2011",
language = "English",
pages = "3109--3112",
booktitle = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",

}

TY - GEN

T1 - Bayesian extension of MUSIC for sound source localization and tracking

AU - Otsuka, Takuma

AU - Nakadai, Kazuhiro

AU - Ogata, Tetsuya

AU - Okuno, Hiroshi G.

PY - 2011

Y1 - 2011

N2 - This paper presents a Bayesian extension of MUSIC-based sound source localization (SSL) and tracking method. SSL is important for distant speech enhancement and simultaneous speech separation for improving speech recognition, as well as for auditory scene analysis by mobile robots. One of the draw- backs of existing SSL methods is the necessity of careful param- eter tunings, e.g., the sound source detection threshold depend- ing on the reverberation time and the number of sources. Our contribution consists of (1) automatic parameter estimation in the variational Bayesian framework and (2) tracking of sound sources with reliability. Experimental results demonstrate our method robustly tracks multiple sound sources in a reverberant environment with RT20 = 840 (ms).

AB - This paper presents a Bayesian extension of MUSIC-based sound source localization (SSL) and tracking method. SSL is important for distant speech enhancement and simultaneous speech separation for improving speech recognition, as well as for auditory scene analysis by mobile robots. One of the draw- backs of existing SSL methods is the necessity of careful param- eter tunings, e.g., the sound source detection threshold depend- ing on the reverberation time and the number of sources. Our contribution consists of (1) automatic parameter estimation in the variational Bayesian framework and (2) tracking of sound sources with reliability. Experimental results demonstrate our method robustly tracks multiple sound sources in a reverberant environment with RT20 = 840 (ms).

KW - MUSIC algorithm

KW - Particle filter

KW - Simultaneous sound source localization

KW - Variational Bayes

UR - http://www.scopus.com/inward/record.url?scp=84865790519&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865790519&partnerID=8YFLogxK

M3 - Conference contribution

SP - 3109

EP - 3112

BT - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

ER -