<abstract xmlns="http://www.w3.org/1999/xhtml"><p>In the study the efficiency of three features for trajectory-based endpoint detection is experimentally evaluated in the fixed-text Dynamic Time Warping (DTW) - a based speaker verification task with short phrases of telephone speech. The employed features are Modified Teager Energy (MTE), Energy-Entropy (EE) feature and Mean-Delta (MD) feature. The utterance boundaries in the endpoint detector are provided by means of state automaton and a set of thresholds based only on trajectory characteristics. The training and testing have been done with noisy telephone speech (short phrases in Bulgarian language with length of about 2 s) selected from BG-SRDat corpus. The results of the experiments have shown that the MD feature demonstrates the best performance in the endpoint detection tests in terms of the verification rate.</p></abstract>

In the study the efficiency of three features for trajectory-based endpoint detection is experimentally evaluated in the fixed-text Dynamic Time Warping (DTW) - a based speaker verification task with short phrases of telephone speech. The employed features are Modified Teager Energy (MTE), Energy-Entropy (EE) feature and Mean-Delta (MD) feature. The utterance boundaries in the endpoint detector are provided by means of state automaton and a set of thresholds based only on trajectory characteristics. The training and testing have been done with noisy telephone speech (short phrases in Bulgarian language with length of about 2 s) selected from BG-SRDat corpus. The results of the experiments have shown that the MD feature demonstrates the best performance in the endpoint detection tests in terms of the verification rate.

Telephone Speech Endpoint Detection using Mean-Delta Feature

Institute of Information and Communication Technologies, 1113 Sofia

Cybernetics and Information Technologies

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

{"article-title":"Telephone Speech Endpoint Detection using Mean-Delta Feature"}

In the study the efficiency of three features for trajectory-based endpoint detection is experimentally evaluated in the fixed-text Dynamic Time Warping (DTW)...

Telephone Speech Endpoint Detection using Mean-Delta Feature

Publié en ligne: 15 juil. 2014

Pages: 127 - 139

DOI: https://doi.org/10.2478/cait-2014-0025

Mots clés
Speech detection, endpoint detection, spectral entropy

© 2014 Atanas Ouzounov

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Telephone Speech Endpoint Detection using Mean-Delta Feature

Publié en ligne: 15 juil. 2014

Pages: 127 - 139

DOI: https://doi.org/10.2478/cait-2014-0025

Mots clésSpeech detection, endpoint detection, spectral entropy

© 2014 Atanas Ouzounov

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Mots clés
Speech detection, endpoint detection, spectral entropy