An isolated speech endpoint detector using multiple speech features

Energy and zero crossing rate of the speech signal have been the two most widely used features for detecting the endpoints of an utterance. This paper proposed a new approach for locating the endpoint for isolated speech, which significantly improve the endpoint detector performance. The proposed al...

Full description

Bibliographic Details
Main Authors: Ahmad, Abdul Manan, Goh, Kia Eng, Mohamed Shaharoun, Awaluddin, Tan, Chiu Yeek, Jarni, Muhamad Hafiz
Format: Book Section
Language:English
Published: IEEE 2004
Subjects:
Online Access:http://eprints.utm.my/9804/
http://eprints.utm.my/9804/1/AbdulMananAhmad2004_an_isolated_speech_endpoint_detector.pdf
_version_ 1848891943157760000
author Ahmad, Abdul Manan
Goh, Kia Eng
Mohamed Shaharoun, Awaluddin
Tan, Chiu Yeek
Jarni, Muhamad Hafiz
author_facet Ahmad, Abdul Manan
Goh, Kia Eng
Mohamed Shaharoun, Awaluddin
Tan, Chiu Yeek
Jarni, Muhamad Hafiz
author_sort Ahmad, Abdul Manan
building UTeM Institutional Repository
collection Online Access
description Energy and zero crossing rate of the speech signal have been the two most widely used features for detecting the endpoints of an utterance. This paper proposed a new approach for locating the endpoint for isolated speech, which significantly improve the endpoint detector performance. The proposed algorithm relies on multiple speech features: root mean square energy (rmse), zero crossing rate (zcr) and cepstral coefficient (cepstrum) where the Euclidean distance measure is adopted to accurately detect the endpoint of an isolated utterance. This algorithm offers better performance than conventional algorithm which using energy only. The vocabulary for the experiment includes English digit from 1 to 9. These experimental results were conducted by 360 utterances from a male speaker. Experimental results show that the accuracy of the algorithm is quite acceptable.
first_indexed 2025-11-15T21:05:59Z
format Book Section
id utm-9804
institution Universiti Teknologi Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T21:05:59Z
publishDate 2004
publisher IEEE
recordtype eprints
repository_type Digital Repository
spelling utm-98042011-03-07T07:28:41Z http://eprints.utm.my/9804/ An isolated speech endpoint detector using multiple speech features Ahmad, Abdul Manan Goh, Kia Eng Mohamed Shaharoun, Awaluddin Tan, Chiu Yeek Jarni, Muhamad Hafiz QA75 Electronic computers. Computer science Energy and zero crossing rate of the speech signal have been the two most widely used features for detecting the endpoints of an utterance. This paper proposed a new approach for locating the endpoint for isolated speech, which significantly improve the endpoint detector performance. The proposed algorithm relies on multiple speech features: root mean square energy (rmse), zero crossing rate (zcr) and cepstral coefficient (cepstrum) where the Euclidean distance measure is adopted to accurately detect the endpoint of an isolated utterance. This algorithm offers better performance than conventional algorithm which using energy only. The vocabulary for the experiment includes English digit from 1 to 9. These experimental results were conducted by 360 utterances from a male speaker. Experimental results show that the accuracy of the algorithm is quite acceptable. IEEE 2004-11-21 Book Section PeerReviewed application/pdf en http://eprints.utm.my/9804/1/AbdulMananAhmad2004_an_isolated_speech_endpoint_detector.pdf Ahmad, Abdul Manan and Goh, Kia Eng and Mohamed Shaharoun, Awaluddin and Tan, Chiu Yeek and Jarni, Muhamad Hafiz (2004) An isolated speech endpoint detector using multiple speech features. In: Tencon 2004 - 2004 IEEE Region 10 Conference, Vols A-D, Proceedings - Analog And Digital Techniques in Electrical Engineering. TENCON-IEEE Region 10 Conference Proceedings, B . IEEE, USA, pp. 403-406. ISBN 0-7803-8560-8 10.1109/TENCON.2004.1414617
spellingShingle QA75 Electronic computers. Computer science
Ahmad, Abdul Manan
Goh, Kia Eng
Mohamed Shaharoun, Awaluddin
Tan, Chiu Yeek
Jarni, Muhamad Hafiz
An isolated speech endpoint detector using multiple speech features
title An isolated speech endpoint detector using multiple speech features
title_full An isolated speech endpoint detector using multiple speech features
title_fullStr An isolated speech endpoint detector using multiple speech features
title_full_unstemmed An isolated speech endpoint detector using multiple speech features
title_short An isolated speech endpoint detector using multiple speech features
title_sort isolated speech endpoint detector using multiple speech features
topic QA75 Electronic computers. Computer science
url http://eprints.utm.my/9804/
http://eprints.utm.my/9804/
http://eprints.utm.my/9804/1/AbdulMananAhmad2004_an_isolated_speech_endpoint_detector.pdf