A DCT based nonlinear predictive coding for feature extraction in speech recognition systems

Mahmood Yousefi Azar; Farbod Razzazi

doi:10.1109/CIMSA.2008.4595825

A DCT based nonlinear predictive coding for feature extraction in speech recognition systems

Mahmood Yousefi Azar, Farbod Razzazi

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

6 Citations (Scopus)

Abstract

Speech representation strategies play a key role in automatic speech recognition systems. In this study, a nonlinear procedure has been proposed to overcome the complexities of speech sequence representations. The proposed method may he considered as an extension of nonlinear predictive coding representation procedure in cosine transform domain. The best results belong to classification of nonlinear behaved stop phonemes (i.e. /b/, /d/, /g/) in TIMIT database which show good performance while reducing the computational complexity in comparison to standard NPC.

Original language	English
Title of host publication	CIMSA 2008 - IEEE Conference on Computational Intelligence for Measurement Systems and Applications Proceedings
Pages	19-22
Number of pages	4
DOIs	https://doi.org/10.1109/CIMSA.2008.4595825
Publication status	Published - 26 Sept 2008
Externally published	Yes
Event	2008 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, IEEE CIMSA 2008 - Istanbul, Turkey Duration: 14 Jul 2008 → 16 Jul 2008

Conference

Conference	2008 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, IEEE CIMSA 2008
Country/Territory	Turkey
City	Istanbul
Period	14/07/08 → 16/07/08

Keywords

Automatic feature extraction
Automatic speech recognition
Cosine transform
Neural network

Access to Document

10.1109/CIMSA.2008.4595825

Cite this

@inproceedings{a4695e023fe143128f3fb5e348c13965,

title = "A DCT based nonlinear predictive coding for feature extraction in speech recognition systems",

abstract = "Speech representation strategies play a key role in automatic speech recognition systems. In this study, a nonlinear procedure has been proposed to overcome the complexities of speech sequence representations. The proposed method may he considered as an extension of nonlinear predictive coding representation procedure in cosine transform domain. The best results belong to classification of nonlinear behaved stop phonemes (i.e. /b/, /d/, /g/) in TIMIT database which show good performance while reducing the computational complexity in comparison to standard NPC.",

keywords = "Automatic feature extraction, Automatic speech recognition, Cosine transform, Neural network",

author = "Azar, \{Mahmood Yousefi\} and Farbod Razzazi",

year = "2008",

month = sep,

day = "26",

doi = "10.1109/CIMSA.2008.4595825",

language = "English",

isbn = "9781424423064",

pages = "19--22",

booktitle = "CIMSA 2008 - IEEE Conference on Computational Intelligence for Measurement Systems and Applications Proceedings",

note = "2008 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, IEEE CIMSA 2008 ; Conference date: 14-07-2008 Through 16-07-2008",

}

Azar, MY & Razzazi, F 2008, A DCT based nonlinear predictive coding for feature extraction in speech recognition systems. in CIMSA 2008 - IEEE Conference on Computational Intelligence for Measurement Systems and Applications Proceedings., 4595825, pp. 19-22, 2008 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, IEEE CIMSA 2008, Istanbul, Turkey, 14/07/08. https://doi.org/10.1109/CIMSA.2008.4595825

A DCT based nonlinear predictive coding for feature extraction in speech recognition systems. / Azar, Mahmood Yousefi; Razzazi, Farbod.
CIMSA 2008 - IEEE Conference on Computational Intelligence for Measurement Systems and Applications Proceedings. 2008. p. 19-22 4595825.

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

TY - GEN

T1 - A DCT based nonlinear predictive coding for feature extraction in speech recognition systems

AU - Azar, Mahmood Yousefi

AU - Razzazi, Farbod

PY - 2008/9/26

Y1 - 2008/9/26

N2 - Speech representation strategies play a key role in automatic speech recognition systems. In this study, a nonlinear procedure has been proposed to overcome the complexities of speech sequence representations. The proposed method may he considered as an extension of nonlinear predictive coding representation procedure in cosine transform domain. The best results belong to classification of nonlinear behaved stop phonemes (i.e. /b/, /d/, /g/) in TIMIT database which show good performance while reducing the computational complexity in comparison to standard NPC.

AB - Speech representation strategies play a key role in automatic speech recognition systems. In this study, a nonlinear procedure has been proposed to overcome the complexities of speech sequence representations. The proposed method may he considered as an extension of nonlinear predictive coding representation procedure in cosine transform domain. The best results belong to classification of nonlinear behaved stop phonemes (i.e. /b/, /d/, /g/) in TIMIT database which show good performance while reducing the computational complexity in comparison to standard NPC.

KW - Automatic feature extraction

KW - Automatic speech recognition

KW - Cosine transform

KW - Neural network

UR - https://www.scopus.com/pages/publications/52249085416

U2 - 10.1109/CIMSA.2008.4595825

DO - 10.1109/CIMSA.2008.4595825

M3 - Conference proceeding contribution

AN - SCOPUS:52249085416

SN - 9781424423064

SP - 19

EP - 22

BT - CIMSA 2008 - IEEE Conference on Computational Intelligence for Measurement Systems and Applications Proceedings

T2 - 2008 IEEE International Conference on Computational Intelligence for Measurement Systems and Applications, IEEE CIMSA 2008

Y2 - 14 July 2008 through 16 July 2008

ER -

A DCT based nonlinear predictive coding for feature extraction in speech recognition systems

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this