Unsupervised speaker segmentation in telephone conversations

Arnon Cohen; Vladimir Lapidus

Unsupervised speaker segmentation in telephone conversations

Arnon Cohen, Vladimir Lapidus

Department of Electrical & Computer Engineering

Research output: Contribution to conference › Paper › peer-review

8 Scopus citations

Abstract

Speaker recognition and verification has been used in a variety of commercial, forensic and military applications. The classical problem is that of supervised recognition, in which there is sufficient a priori information on the speakers to be identified. This paper deals with the problem of unsupervised speech segmentation and speaker classification, where no a priori speaker information is available. The algorithm accepts dual-speaker conversation telephone speech data, detects events of simultaneous speakers, and segment the signal by assigning each speech segment to its speaker. Discrete HMM are used, with 12th order cepstral coefficients. Correct recognition rates of more than 90% are demonstrated.

Original language	English
Pages	102-105
Number of pages	4
State	Published - 1 Dec 1996
Event	Proceedings of the 1996 19th Convention of Electrical and Electronics Engineers in Israel - Jerusalem, Isr Duration: 5 Nov 1996 → 6 Nov 1996

Conference

Conference	Proceedings of the 1996 19th Convention of Electrical and Electronics Engineers in Israel
City	Jerusalem, Isr
Period	5/11/96 → 6/11/96

ASJC Scopus subject areas

General Engineering

Cite this

@conference{1a765021bba34f96bef0e9d3dc0aeec1,

title = "Unsupervised speaker segmentation in telephone conversations",

abstract = "Speaker recognition and verification has been used in a variety of commercial, forensic and military applications. The classical problem is that of supervised recognition, in which there is sufficient a priori information on the speakers to be identified. This paper deals with the problem of unsupervised speech segmentation and speaker classification, where no a priori speaker information is available. The algorithm accepts dual-speaker conversation telephone speech data, detects events of simultaneous speakers, and segment the signal by assigning each speech segment to its speaker. Discrete HMM are used, with 12th order cepstral coefficients. Correct recognition rates of more than 90% are demonstrated.",

author = "Arnon Cohen and Vladimir Lapidus",

year = "1996",

month = dec,

day = "1",

language = "English",

pages = "102--105",

note = "Proceedings of the 1996 19th Convention of Electrical and Electronics Engineers in Israel ; Conference date: 05-11-1996 Through 06-11-1996",

}

TY - CONF

T1 - Unsupervised speaker segmentation in telephone conversations

AU - Cohen, Arnon

AU - Lapidus, Vladimir

PY - 1996/12/1

Y1 - 1996/12/1

N2 - Speaker recognition and verification has been used in a variety of commercial, forensic and military applications. The classical problem is that of supervised recognition, in which there is sufficient a priori information on the speakers to be identified. This paper deals with the problem of unsupervised speech segmentation and speaker classification, where no a priori speaker information is available. The algorithm accepts dual-speaker conversation telephone speech data, detects events of simultaneous speakers, and segment the signal by assigning each speech segment to its speaker. Discrete HMM are used, with 12th order cepstral coefficients. Correct recognition rates of more than 90% are demonstrated.

AB - Speaker recognition and verification has been used in a variety of commercial, forensic and military applications. The classical problem is that of supervised recognition, in which there is sufficient a priori information on the speakers to be identified. This paper deals with the problem of unsupervised speech segmentation and speaker classification, where no a priori speaker information is available. The algorithm accepts dual-speaker conversation telephone speech data, detects events of simultaneous speakers, and segment the signal by assigning each speech segment to its speaker. Discrete HMM are used, with 12th order cepstral coefficients. Correct recognition rates of more than 90% are demonstrated.

UR - http://www.scopus.com/inward/record.url?scp=0030381663&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0030381663

SP - 102

EP - 105

T2 - Proceedings of the 1996 19th Convention of Electrical and Electronics Engineers in Israel

Y2 - 5 November 1996 through 6 November 1996

ER -

Unsupervised speaker segmentation in telephone conversations

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this