Least relative entropy for voiced/unvoiced speech classification

Darren K. Emge, Tulay Adali, Mustafa (Kemal) Sonmez

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The aim of this work is to develop a flexible and efficient approach to the classification of the ratio of voiced to unvoiced excitation sources in continuous speech. To achieve this aim we adopt a probabilistic neural network approach. This is accomplished by designing a multi layer perceptron classifier trained by steepest descent minimization of the Least Relative Entropy (LRE) cost function. By using the LRE cost function we can directly output the ratio, as a probability, of excitation source, voiced to unvoiced, for a given speech segment. These output probabilities can then be used directly in other applications, such as low bit rate coders.

Original languageEnglish (US)
Title of host publicationProceedings of the International Joint Conference on Neural Networks
PublisherIEEE
Pages2976-2980
Number of pages5
Volume5
StatePublished - 1999
Externally publishedYes
EventInternational Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, USA
Duration: Jul 10 1999Jul 16 1999

Other

OtherInternational Joint Conference on Neural Networks (IJCNN'99)
CityWashington, DC, USA
Period7/10/997/16/99

Fingerprint

Cost functions
Entropy
Multilayer neural networks
Classifiers
Neural networks

ASJC Scopus subject areas

  • Software

Cite this

Emge, D. K., Adali, T., & Sonmez, M. K. (1999). Least relative entropy for voiced/unvoiced speech classification. In Proceedings of the International Joint Conference on Neural Networks (Vol. 5, pp. 2976-2980). IEEE.

Least relative entropy for voiced/unvoiced speech classification. / Emge, Darren K.; Adali, Tulay; Sonmez, Mustafa (Kemal).

Proceedings of the International Joint Conference on Neural Networks. Vol. 5 IEEE, 1999. p. 2976-2980.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Emge, DK, Adali, T & Sonmez, MK 1999, Least relative entropy for voiced/unvoiced speech classification. in Proceedings of the International Joint Conference on Neural Networks. vol. 5, IEEE, pp. 2976-2980, International Joint Conference on Neural Networks (IJCNN'99), Washington, DC, USA, 7/10/99.
Emge DK, Adali T, Sonmez MK. Least relative entropy for voiced/unvoiced speech classification. In Proceedings of the International Joint Conference on Neural Networks. Vol. 5. IEEE. 1999. p. 2976-2980
Emge, Darren K. ; Adali, Tulay ; Sonmez, Mustafa (Kemal). / Least relative entropy for voiced/unvoiced speech classification. Proceedings of the International Joint Conference on Neural Networks. Vol. 5 IEEE, 1999. pp. 2976-2980
@inproceedings{bf678e5c4cac4001b66b828a547ad93e,
title = "Least relative entropy for voiced/unvoiced speech classification",
abstract = "The aim of this work is to develop a flexible and efficient approach to the classification of the ratio of voiced to unvoiced excitation sources in continuous speech. To achieve this aim we adopt a probabilistic neural network approach. This is accomplished by designing a multi layer perceptron classifier trained by steepest descent minimization of the Least Relative Entropy (LRE) cost function. By using the LRE cost function we can directly output the ratio, as a probability, of excitation source, voiced to unvoiced, for a given speech segment. These output probabilities can then be used directly in other applications, such as low bit rate coders.",
author = "Emge, {Darren K.} and Tulay Adali and Sonmez, {Mustafa (Kemal)}",
year = "1999",
language = "English (US)",
volume = "5",
pages = "2976--2980",
booktitle = "Proceedings of the International Joint Conference on Neural Networks",
publisher = "IEEE",

}

TY - GEN

T1 - Least relative entropy for voiced/unvoiced speech classification

AU - Emge, Darren K.

AU - Adali, Tulay

AU - Sonmez, Mustafa (Kemal)

PY - 1999

Y1 - 1999

N2 - The aim of this work is to develop a flexible and efficient approach to the classification of the ratio of voiced to unvoiced excitation sources in continuous speech. To achieve this aim we adopt a probabilistic neural network approach. This is accomplished by designing a multi layer perceptron classifier trained by steepest descent minimization of the Least Relative Entropy (LRE) cost function. By using the LRE cost function we can directly output the ratio, as a probability, of excitation source, voiced to unvoiced, for a given speech segment. These output probabilities can then be used directly in other applications, such as low bit rate coders.

AB - The aim of this work is to develop a flexible and efficient approach to the classification of the ratio of voiced to unvoiced excitation sources in continuous speech. To achieve this aim we adopt a probabilistic neural network approach. This is accomplished by designing a multi layer perceptron classifier trained by steepest descent minimization of the Least Relative Entropy (LRE) cost function. By using the LRE cost function we can directly output the ratio, as a probability, of excitation source, voiced to unvoiced, for a given speech segment. These output probabilities can then be used directly in other applications, such as low bit rate coders.

UR - http://www.scopus.com/inward/record.url?scp=0033351871&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033351871&partnerID=8YFLogxK

M3 - Conference contribution

VL - 5

SP - 2976

EP - 2980

BT - Proceedings of the International Joint Conference on Neural Networks

PB - IEEE

ER -