CANDIDATE SELECTION BASED ON SIGNIFICANCE TESTING AND ITS USE IN NORMALISATION AND SCORING

Ji Hwan Kim; Gil Jin Jang; Seong Jin Yun; Yung Hwan Oh

CANDIDATE SELECTION BASED ON SIGNIFICANCE TESTING AND ITS USE IN NORMALISATION AND SCORING

Ji Hwan Kim, Gil Jin Jang, Seong Jin Yun, Yung Hwan Oh

School of Electronics Engineering

Korea Advanced Institute of Science and Technology

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

2 Scopus citations

Abstract

Log likelihood ratio normalisation and scoring methods have been studied by many researchers and have improved the performance of speaker identification systems. However, these studies have disadvantages: the recognised distorted speech segments are different for each speaker. Also the background model in log likelihood ratio normalisation is changed in each speech segment even for the same speaker. This paper presents two techniques. Firstly, candidate selection based on significance testing, which designs the background speaker model more accurately. And secondly, the scoring method, which recognises the same distorted speech segments for every speaker. We perform a number of experiments with the SPIDRE database.

Original language	English
State	Published - 1998
Event	5th International Conference on Spoken Language Processing, ICSLP 1998 - Sydney, Australia Duration: 30 Nov 1998 → 4 Dec 1998

Conference

Conference	5th International Conference on Spoken Language Processing, ICSLP 1998
Country/Territory	Australia
City	Sydney
Period	30/11/98 → 4/12/98

Cite this

@conference{ed3d59c7e08b433bb7182d651978d79f,

title = "CANDIDATE SELECTION BASED ON SIGNIFICANCE TESTING AND ITS USE IN NORMALISATION AND SCORING",

abstract = "Log likelihood ratio normalisation and scoring methods have been studied by many researchers and have improved the performance of speaker identification systems. However, these studies have disadvantages: the recognised distorted speech segments are different for each speaker. Also the background model in log likelihood ratio normalisation is changed in each speech segment even for the same speaker. This paper presents two techniques. Firstly, candidate selection based on significance testing, which designs the background speaker model more accurately. And secondly, the scoring method, which recognises the same distorted speech segments for every speaker. We perform a number of experiments with the SPIDRE database.",

author = "Kim, {Ji Hwan} and Jang, {Gil Jin} and Yun, {Seong Jin} and Oh, {Yung Hwan}",

note = "Publisher Copyright: {\textcopyright} 1998. 5th International Conference on Spoken Language Processing, ICSLP 1998. All rights reserved.; 5th International Conference on Spoken Language Processing, ICSLP 1998 ; Conference date: 30-11-1998 Through 04-12-1998",

year = "1998",

language = "English",

}

TY - CONF

T1 - CANDIDATE SELECTION BASED ON SIGNIFICANCE TESTING AND ITS USE IN NORMALISATION AND SCORING

AU - Kim, Ji Hwan

AU - Jang, Gil Jin

AU - Yun, Seong Jin

AU - Oh, Yung Hwan

PY - 1998

Y1 - 1998

N2 - Log likelihood ratio normalisation and scoring methods have been studied by many researchers and have improved the performance of speaker identification systems. However, these studies have disadvantages: the recognised distorted speech segments are different for each speaker. Also the background model in log likelihood ratio normalisation is changed in each speech segment even for the same speaker. This paper presents two techniques. Firstly, candidate selection based on significance testing, which designs the background speaker model more accurately. And secondly, the scoring method, which recognises the same distorted speech segments for every speaker. We perform a number of experiments with the SPIDRE database.

AB - Log likelihood ratio normalisation and scoring methods have been studied by many researchers and have improved the performance of speaker identification systems. However, these studies have disadvantages: the recognised distorted speech segments are different for each speaker. Also the background model in log likelihood ratio normalisation is changed in each speech segment even for the same speaker. This paper presents two techniques. Firstly, candidate selection based on significance testing, which designs the background speaker model more accurately. And secondly, the scoring method, which recognises the same distorted speech segments for every speaker. We perform a number of experiments with the SPIDRE database.

UR - http://www.scopus.com/inward/record.url?scp=33947433246&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:33947433246

T2 - 5th International Conference on Spoken Language Processing, ICSLP 1998

Y2 - 30 November 1998 through 4 December 1998

ER -

CANDIDATE SELECTION BASED ON SIGNIFICANCE TESTING AND ITS USE IN NORMALISATION AND SCORING

Abstract

Conference

Other files and links

Fingerprint

Cite this