A filter-based approach to detect end-of-utterances from prosody in dialog systems

Authors:
Olac Fuentes;David Vera;Thamar Solorio
Affiliations:
University of Texas at El Paso, El Paso, TX;University of Texas at El Paso, El Paso, TX;University of Texas at El Paso, El Paso, TX
Venue:
NAACL-Short '07 Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics; Companion Volume, Short Papers
Year:
2007

Citing 5
Cited 1

Simplifying decision trees

International Journal of Man-Machine Studies - Special Issue: Knowledge Acquisition for Knowledge-based Systems. Part 5
C4.5: programs for machine learning

C4.5: programs for machine learning
Summed-area tables for texture mapping

SIGGRAPH '84 Proceedings of the 11th annual conference on Computer graphics and interactive techniques
Computer Vision for Music Identification

CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Data Mining: Practical Machine Learning Tools and Techniques, Second Edition (Morgan Kaufmann Series in Data Management Systems)

Data Mining: Practical Machine Learning Tools and Techniques, Second Edition (Morgan Kaufmann Series in Data Management Systems)

Multimodal end-of-turn prediction in multi-party meetings

Proceedings of the 2009 international conference on Multimodal interfaces

Quantified Score

Hi-index	0.01

Visualization

Abstract

We propose an efficient method to detect end-of-utterances from prosodic information in conversational speech. Our method is based on the application of a large set of binary and ramp filters to the energy and fundamental frequency signals obtained from the speech signal. These filter responses, which can be computed very efficiently, are used as input to a learning algorithm that generates the final detector. Preliminary experiments using data obtained from conversations show that an accurate classifier can be trained efficiently and that good results can be obtained without requiring a speech recognition system.