Dimension reduction of the modulation spectrogram for speaker verification

Tomi Kinnunen; Kong Aik Lee; Haizhou Li

Dimension reduction of the modulation spectrogram for speaker verification

Tomi Kinnunen
, Kong Aik Lee
, Haizhou Li

Research output: Unpublished conference presentation (presented paper, abstract, poster) › Conference presentation (not published in journal/proceeding/book) › Academic research › peer-review

38 Citations (Scopus)

Abstract

A so-called modulation spectrogram is obtained from the conventional speech spectrogram by short-term spectral analysis along the temporal trajectories of the frequency bins. In its original definition, the modulation spectrogram is a high-dimensional representation and it is not clear how to extract features from it. In this paper, we define a low-dimensional feature which captures the shape of the modulation spectra. The recognition accuracy of the modulation spectrogram based classifier is improved from our previous result of EER=25.1% to EER=17.4% on the NIST 2001 speaker recognition task.

Original language	English
Publication status	Published - Jan 2008
Externally published	Yes
Event	Speaker and Language Recognition Workshop, Odyssey 2008 - Stellenbosch, South Africa Duration: 21 Jan 2008 → 24 Jan 2008

Conference

Conference	Speaker and Language Recognition Workshop, Odyssey 2008
Country/Territory	South Africa
City	Stellenbosch
Period	21/01/08 → 24/01/08

Keywords

Modulation spectrum
Speaker recognition
Spectro-temporal features

ASJC Scopus subject areas

Human-Computer Interaction
Software
Signal Processing

Cite this

@conference{a7bf6cd03afa4fde9d6a12c004efb7ba,

title = "Dimension reduction of the modulation spectrogram for speaker verification",

abstract = "A so-called modulation spectrogram is obtained from the conventional speech spectrogram by short-term spectral analysis along the temporal trajectories of the frequency bins. In its original definition, the modulation spectrogram is a high-dimensional representation and it is not clear how to extract features from it. In this paper, we define a low-dimensional feature which captures the shape of the modulation spectra. The recognition accuracy of the modulation spectrogram based classifier is improved from our previous result of EER=25.1\% to EER=17.4\% on the NIST 2001 speaker recognition task.",

keywords = "Modulation spectrum, Speaker recognition, Spectro-temporal features",

author = "Tomi Kinnunen and Lee, \{Kong Aik\} and Haizhou Li",

note = "Publisher Copyright: {\textcopyright} Odyssey 2008: Speaker and Language Recognition Workshop. All rights reserved.; Speaker and Language Recognition Workshop, Odyssey 2008 ; Conference date: 21-01-2008 Through 24-01-2008",

year = "2008",

month = jan,

language = "English",

}

Dimension reduction of the modulation spectrogram for speaker verification. / Kinnunen, Tomi; Lee, Kong Aik; Li, Haizhou.
2008. Paper presented at Speaker and Language Recognition Workshop, Odyssey 2008, Stellenbosch, South Africa.

Research output: Unpublished conference presentation (presented paper, abstract, poster) › Conference presentation (not published in journal/proceeding/book) › Academic research › peer-review

TY - CONF

T1 - Dimension reduction of the modulation spectrogram for speaker verification

AU - Kinnunen, Tomi

AU - Lee, Kong Aik

AU - Li, Haizhou

PY - 2008/1

Y1 - 2008/1

N2 - A so-called modulation spectrogram is obtained from the conventional speech spectrogram by short-term spectral analysis along the temporal trajectories of the frequency bins. In its original definition, the modulation spectrogram is a high-dimensional representation and it is not clear how to extract features from it. In this paper, we define a low-dimensional feature which captures the shape of the modulation spectra. The recognition accuracy of the modulation spectrogram based classifier is improved from our previous result of EER=25.1% to EER=17.4% on the NIST 2001 speaker recognition task.

AB - A so-called modulation spectrogram is obtained from the conventional speech spectrogram by short-term spectral analysis along the temporal trajectories of the frequency bins. In its original definition, the modulation spectrogram is a high-dimensional representation and it is not clear how to extract features from it. In this paper, we define a low-dimensional feature which captures the shape of the modulation spectra. The recognition accuracy of the modulation spectrogram based classifier is improved from our previous result of EER=25.1% to EER=17.4% on the NIST 2001 speaker recognition task.

KW - Modulation spectrum

KW - Speaker recognition

KW - Spectro-temporal features

UR - https://www.scopus.com/pages/publications/85084015242

M3 - Conference presentation (not published in journal/proceeding/book)

AN - SCOPUS:85084015242

T2 - Speaker and Language Recognition Workshop, Odyssey 2008

Y2 - 21 January 2008 through 24 January 2008

ER -

Dimension reduction of the modulation spectrogram for speaker verification

Abstract

Conference

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this