Discriminative analysis of distortion sequences in speech recognition

Pao Chung Chang; Sin-Horng Chen; Biing Hwang Juang

doi:10.1109/ICASSP.1991.150398

Discriminative analysis of distortion sequences in speech recognition

Pao Chung Chang^*, Sin-Horng Chen, Biing Hwang Juang

^*Corresponding author for this work

National Chiao Tung University

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

17 Scopus citations

Abstract

The authors suggest a linear discriminant function to complete the distance score instead of a conventional average distance. Several discriminative algorithms are proposed to learn the discriminant function. These include one heuristic method, two methods based on the error propagation algorithm, and one method based on the generalized probabilistic descent (GPD) algorithm. The authors study these methods in a speaker-independent speech recognition task involving utterances of the highly confusable English E-set. The results show that the best performance is obtained by using the GPD method, which achieved a 78.1% accuracy, compared to 67.6% with the traditional average method.

Original language	English
Title of host publication	Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing
Editors	Anon
Publisher	Publ by IEEE
Pages	549-552
Number of pages	4
ISBN (Print)	078030033
DOIs	https://doi.org/10.1109/ICASSP.1991.150398
State	Published - 1 Dec 1991
Event	Proceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91 - Toronto, Ont, Can Duration: 14 May 1991 → 17 May 1991

Publication series

Name	Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing
Volume	1
ISSN (Print)	0736-7791

Conference

Conference	Proceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91
City	Toronto, Ont, Can
Period	14/05/91 → 17/05/91

Access to Document

10.1109/ICASSP.1991.150398

Link to publication in Scopus

Fingerprint Dive into the research topics of 'Discriminative analysis of distortion sequences in speech recognition'. Together they form a unique fingerprint.

Cite this

Chang, P. C., Chen, S-H., & Juang, B. H. (1991). Discriminative analysis of distortion sequences in speech recognition. In Anon (Ed.), Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing (pp. 549-552). (Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing; Vol. 1). Publ by IEEE. https://doi.org/10.1109/ICASSP.1991.150398

Chang, Pao Chung ; Chen, Sin-Horng ; Juang, Biing Hwang. / Discriminative analysis of distortion sequences in speech recognition. Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. editor / Anon. Publ by IEEE, 1991. pp. 549-552 (Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing).

@inproceedings{c85a34c50d5f40068a1bdde99ae7392a,

title = "Discriminative analysis of distortion sequences in speech recognition",

abstract = "The authors suggest a linear discriminant function to complete the distance score instead of a conventional average distance. Several discriminative algorithms are proposed to learn the discriminant function. These include one heuristic method, two methods based on the error propagation algorithm, and one method based on the generalized probabilistic descent (GPD) algorithm. The authors study these methods in a speaker-independent speech recognition task involving utterances of the highly confusable English E-set. The results show that the best performance is obtained by using the GPD method, which achieved a 78.1% accuracy, compared to 67.6% with the traditional average method.",

author = "Chang, {Pao Chung} and Sin-Horng Chen and Juang, {Biing Hwang}",

year = "1991",

month = dec,

day = "1",

doi = "10.1109/ICASSP.1991.150398",

language = "English",

isbn = "078030033",

series = "Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing",

publisher = "Publ by IEEE",

pages = "549--552",

editor = "Anon",

booktitle = "Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing",

note = "null ; Conference date: 14-05-1991 Through 17-05-1991",

}

Chang, PC, Chen, S-H & Juang, BH 1991, Discriminative analysis of distortion sequences in speech recognition. in Anon (ed.), Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing, vol. 1, Publ by IEEE, pp. 549-552, Proceedings of the 1991 International Conference on Acoustics, Speech, and Signal Processing - ICASSP 91, Toronto, Ont, Can, 14/05/91. https://doi.org/10.1109/ICASSP.1991.150398

Discriminative analysis of distortion sequences in speech recognition. / Chang, Pao Chung; Chen, Sin-Horng; Juang, Biing Hwang.

Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. ed. / Anon. Publ by IEEE, 1991. p. 549-552 (Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing; Vol. 1).

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

TY - GEN

T1 - Discriminative analysis of distortion sequences in speech recognition

AU - Chang, Pao Chung

AU - Chen, Sin-Horng

AU - Juang, Biing Hwang

PY - 1991/12/1

Y1 - 1991/12/1

N2 - The authors suggest a linear discriminant function to complete the distance score instead of a conventional average distance. Several discriminative algorithms are proposed to learn the discriminant function. These include one heuristic method, two methods based on the error propagation algorithm, and one method based on the generalized probabilistic descent (GPD) algorithm. The authors study these methods in a speaker-independent speech recognition task involving utterances of the highly confusable English E-set. The results show that the best performance is obtained by using the GPD method, which achieved a 78.1% accuracy, compared to 67.6% with the traditional average method.

AB - The authors suggest a linear discriminant function to complete the distance score instead of a conventional average distance. Several discriminative algorithms are proposed to learn the discriminant function. These include one heuristic method, two methods based on the error propagation algorithm, and one method based on the generalized probabilistic descent (GPD) algorithm. The authors study these methods in a speaker-independent speech recognition task involving utterances of the highly confusable English E-set. The results show that the best performance is obtained by using the GPD method, which achieved a 78.1% accuracy, compared to 67.6% with the traditional average method.

UR - http://www.scopus.com/inward/record.url?scp=0026368472&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.1991.150398

DO - 10.1109/ICASSP.1991.150398

M3 - Conference contribution

AN - SCOPUS:0026368472

SN - 078030033

T3 - Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing

SP - 549

EP - 552

BT - Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing

A2 - Anon, null

PB - Publ by IEEE

Y2 - 14 May 1991 through 17 May 1991

ER -

Chang PC, Chen S-H, Juang BH. Discriminative analysis of distortion sequences in speech recognition. In Anon, editor, Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing. Publ by IEEE. 1991. p. 549-552. (Proceedings - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing). https://doi.org/10.1109/ICASSP.1991.150398