Robust localization and tracking of multiple speakers in real environments for binaural robot audition

Ui Hyun Kim; Hiroshi G. Okuno

doi:10.1109/WIAMIS.2013.6616137

Robust localization and tracking of multiple speakers in real environments for binaural robot audition

Ui Hyun Kim, Hiroshi G. Okuno

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

4 Citations (Scopus)

Abstract

This paper presents a multisource sound localization method based on the generalized cross-correlation (GCC) method weighted by the phase transform (PHAT) and a novel multisource speech tracking method consisting of voice activity detection (VAD) and K-means clustering algorithm for binaural robot audition. The standard K-means clustering algorithm was improved for the purpose of multisource speech tracking by adding two additional steps. Experiments conducted on the SIG-2 humanoid robot in a real environment show that our method can track multiple speakers in real-time with tracking error below 4.35°.

Original language	English
Title of host publication	International Workshop on Image Analysis for Multimedia Interactive Services
DOIs	https://doi.org/10.1109/WIAMIS.2013.6616137
Publication status	Published - 2013
Externally published	Yes
Event	2013 14th International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS 2013 - Paris Duration: 2013 Jul 3 → 2013 Jul 5

Other

Other	2013 14th International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS 2013
City	Paris
Period	13/7/3 → 13/7/5

ASJC Scopus subject areas

Computer Graphics and Computer-Aided Design
Human-Computer Interaction
Software

Access to Document

10.1109/WIAMIS.2013.6616137

Cite this

Kim, UH & Okuno, HG 2013, Robust localization and tracking of multiple speakers in real environments for binaural robot audition. in International Workshop on Image Analysis for Multimedia Interactive Services., 6616137, 2013 14th International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS 2013, Paris, 13/7/3. https://doi.org/10.1109/WIAMIS.2013.6616137

@inproceedings{06ec2365ae43400ab478e983e129e60c,

title = "Robust localization and tracking of multiple speakers in real environments for binaural robot audition",

abstract = "This paper presents a multisource sound localization method based on the generalized cross-correlation (GCC) method weighted by the phase transform (PHAT) and a novel multisource speech tracking method consisting of voice activity detection (VAD) and K-means clustering algorithm for binaural robot audition. The standard K-means clustering algorithm was improved for the purpose of multisource speech tracking by adding two additional steps. Experiments conducted on the SIG-2 humanoid robot in a real environment show that our method can track multiple speakers in real-time with tracking error below 4.35°.",

author = "Kim, {Ui Hyun} and Okuno, {Hiroshi G.}",

year = "2013",

doi = "10.1109/WIAMIS.2013.6616137",

language = "English",

isbn = "9781479908332",

booktitle = "International Workshop on Image Analysis for Multimedia Interactive Services",

note = "2013 14th International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS 2013 ; Conference date: 03-07-2013 Through 05-07-2013",

}

TY - GEN

T1 - Robust localization and tracking of multiple speakers in real environments for binaural robot audition

AU - Kim, Ui Hyun

AU - Okuno, Hiroshi G.

PY - 2013

Y1 - 2013

N2 - This paper presents a multisource sound localization method based on the generalized cross-correlation (GCC) method weighted by the phase transform (PHAT) and a novel multisource speech tracking method consisting of voice activity detection (VAD) and K-means clustering algorithm for binaural robot audition. The standard K-means clustering algorithm was improved for the purpose of multisource speech tracking by adding two additional steps. Experiments conducted on the SIG-2 humanoid robot in a real environment show that our method can track multiple speakers in real-time with tracking error below 4.35°.

AB - This paper presents a multisource sound localization method based on the generalized cross-correlation (GCC) method weighted by the phase transform (PHAT) and a novel multisource speech tracking method consisting of voice activity detection (VAD) and K-means clustering algorithm for binaural robot audition. The standard K-means clustering algorithm was improved for the purpose of multisource speech tracking by adding two additional steps. Experiments conducted on the SIG-2 humanoid robot in a real environment show that our method can track multiple speakers in real-time with tracking error below 4.35°.

UR - http://www.scopus.com/inward/record.url?scp=84887222569&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887222569&partnerID=8YFLogxK

U2 - 10.1109/WIAMIS.2013.6616137

DO - 10.1109/WIAMIS.2013.6616137

M3 - Conference contribution

AN - SCOPUS:84887222569

SN - 9781479908332

BT - International Workshop on Image Analysis for Multimedia Interactive Services

T2 - 2013 14th International Workshop on Image Analysis for Multimedia Interactive Services, WIAMIS 2013

Y2 - 3 July 2013 through 5 July 2013

ER -

Robust localization and tracking of multiple speakers in real environments for binaural robot audition

Abstract

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this