Interruption point detection of spontaneous speech using prior knowledge and multiple features

Wei Bin Liang, Jui Feng Yeh, Chung-Hsien Wu, Chi Chiuan Liou

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

1 Citation (Scopus)

Abstract

This paper presents an approach to interruption point (IP) detection of spontaneous speech based on conditional random fields using prior knowledge and multiple features. The features adopted in this study consist of subsyllable boundaries and prosodic features. Conditional random fields (CRFs) and variable-length contextual features are employed for IP modeling. In order to apply the features with continuous values to the CRF models, the K-means clustering algorithm is adopted for the quantization of the prosodic features. In the experimental results, Mandarin Conversional Dialogue Corpus (MCDC) was used to evaluate the proposed method. The IP detection error rate achieved almost 20% reduction in Rt04 measure. The experimental results show that the proposed model can effectively detect the interruption point in spontaneous speech.

Original languageEnglish
Title of host publication2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings
Pages1457-1460
Number of pages4
DOIs
Publication statusPublished - 2008 Oct 23
Event2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Hannover, Germany
Duration: 2008 Jun 232008 Jun 26

Publication series

Name2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings

Other

Other2008 IEEE International Conference on Multimedia and Expo, ICME 2008
CountryGermany
CityHannover
Period08-06-2308-06-26

Fingerprint

Error detection
Clustering algorithms

All Science Journal Classification (ASJC) codes

  • Computer Graphics and Computer-Aided Design
  • Electrical and Electronic Engineering

Cite this

Liang, W. B., Yeh, J. F., Wu, C-H., & Liou, C. C. (2008). Interruption point detection of spontaneous speech using prior knowledge and multiple features. In 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings (pp. 1457-1460). [4607720] (2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings). https://doi.org/10.1109/ICME.2008.4607720
Liang, Wei Bin ; Yeh, Jui Feng ; Wu, Chung-Hsien ; Liou, Chi Chiuan. / Interruption point detection of spontaneous speech using prior knowledge and multiple features. 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings. 2008. pp. 1457-1460 (2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings).
@inproceedings{40fbefc8e47348bf80bd9531d6455826,
title = "Interruption point detection of spontaneous speech using prior knowledge and multiple features",
abstract = "This paper presents an approach to interruption point (IP) detection of spontaneous speech based on conditional random fields using prior knowledge and multiple features. The features adopted in this study consist of subsyllable boundaries and prosodic features. Conditional random fields (CRFs) and variable-length contextual features are employed for IP modeling. In order to apply the features with continuous values to the CRF models, the K-means clustering algorithm is adopted for the quantization of the prosodic features. In the experimental results, Mandarin Conversional Dialogue Corpus (MCDC) was used to evaluate the proposed method. The IP detection error rate achieved almost 20{\%} reduction in Rt04 measure. The experimental results show that the proposed model can effectively detect the interruption point in spontaneous speech.",
author = "Liang, {Wei Bin} and Yeh, {Jui Feng} and Chung-Hsien Wu and Liou, {Chi Chiuan}",
year = "2008",
month = "10",
day = "23",
doi = "10.1109/ICME.2008.4607720",
language = "English",
isbn = "9781424425716",
series = "2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings",
pages = "1457--1460",
booktitle = "2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings",

}

Liang, WB, Yeh, JF, Wu, C-H & Liou, CC 2008, Interruption point detection of spontaneous speech using prior knowledge and multiple features. in 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings., 4607720, 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings, pp. 1457-1460, 2008 IEEE International Conference on Multimedia and Expo, ICME 2008, Hannover, Germany, 08-06-23. https://doi.org/10.1109/ICME.2008.4607720

Interruption point detection of spontaneous speech using prior knowledge and multiple features. / Liang, Wei Bin; Yeh, Jui Feng; Wu, Chung-Hsien; Liou, Chi Chiuan.

2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings. 2008. p. 1457-1460 4607720 (2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings).

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

TY - GEN

T1 - Interruption point detection of spontaneous speech using prior knowledge and multiple features

AU - Liang, Wei Bin

AU - Yeh, Jui Feng

AU - Wu, Chung-Hsien

AU - Liou, Chi Chiuan

PY - 2008/10/23

Y1 - 2008/10/23

N2 - This paper presents an approach to interruption point (IP) detection of spontaneous speech based on conditional random fields using prior knowledge and multiple features. The features adopted in this study consist of subsyllable boundaries and prosodic features. Conditional random fields (CRFs) and variable-length contextual features are employed for IP modeling. In order to apply the features with continuous values to the CRF models, the K-means clustering algorithm is adopted for the quantization of the prosodic features. In the experimental results, Mandarin Conversional Dialogue Corpus (MCDC) was used to evaluate the proposed method. The IP detection error rate achieved almost 20% reduction in Rt04 measure. The experimental results show that the proposed model can effectively detect the interruption point in spontaneous speech.

AB - This paper presents an approach to interruption point (IP) detection of spontaneous speech based on conditional random fields using prior knowledge and multiple features. The features adopted in this study consist of subsyllable boundaries and prosodic features. Conditional random fields (CRFs) and variable-length contextual features are employed for IP modeling. In order to apply the features with continuous values to the CRF models, the K-means clustering algorithm is adopted for the quantization of the prosodic features. In the experimental results, Mandarin Conversional Dialogue Corpus (MCDC) was used to evaluate the proposed method. The IP detection error rate achieved almost 20% reduction in Rt04 measure. The experimental results show that the proposed model can effectively detect the interruption point in spontaneous speech.

UR - http://www.scopus.com/inward/record.url?scp=54049115611&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=54049115611&partnerID=8YFLogxK

U2 - 10.1109/ICME.2008.4607720

DO - 10.1109/ICME.2008.4607720

M3 - Conference contribution

SN - 9781424425716

T3 - 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings

SP - 1457

EP - 1460

BT - 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings

ER -

Liang WB, Yeh JF, Wu C-H, Liou CC. Interruption point detection of spontaneous speech using prior knowledge and multiple features. In 2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings. 2008. p. 1457-1460. 4607720. (2008 IEEE International Conference on Multimedia and Expo, ICME 2008 - Proceedings). https://doi.org/10.1109/ICME.2008.4607720