Real-time Filled Pause Detection for Spontaneous Speech Dialogue
This project is proposed and researched by
Masataka Goto and
Katunobu Itou.
Japanese version is here.
We have developed
a method for automatically detecting filled (vocalized) pauses,
which are one of the hesitation phenomena
that current speech recognizers typically cannot handle.
The detection of these pauses is
important in spontaneous speech dialogue systems
because they play valuable roles,
such as helping a speaker keep a conversational turn,
in oral communication.
Although a few speech recognition systems
have processed filled pauses within
subword-based connected word recognition or word-spotting frameworks,
they did not detect the pauses individually and
consequently could not consider their roles.
We propose a method that detects filled pauses
and word lengthening
on the basis of small fundamental frequency transition and
small spectral envelope deformation
under the assumption that
speakers do not change articulator parameters during filled pauses.
Experimental results
for a Japanese spoken dialogue corpus show that
our real-time filled-pause-detection system yielded
good recall and precision rates.
References:
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
A Real-time System Detecting Filled Pauses
in Spontaneous Speech,
The Transactions of the Institute of Electronics,
Information and Communication Engineers D-II,
Vol.J83-D-II, No.11, pp.2330-2340, November 2000. (in Japanese)
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
Evaluation of a Real-time Filled Pause Detection System,
The 2000 Spring Meeting of The Acoustical Society of Japan, 3-8-8,
pp.81-82, March 2000. (in Japanese)
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
A Real-time Filled Pause Detection System:
Toward Spontaneous Speech Dialogue,
2000 RWC Symposium
Proceedings (RWC Technical Report),
TR-99-002, pp.187-192, January 2000.
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
Real-time Detection of Filled Pauses in Spontaneous Speech,
The 1999 Autumn Meeting of The Acoustical Society of Japan, 3-1-5,
pp.105-106, October 1999. (in Japanese)
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
A Real-time Filled Pause Detection System
for Spontaneous Speech Recognition,
Proceedings of
the 6th European Conference on Speech Communication and Technology
(Eurospeech '99),
pp.227-230, September 1999.
- Masataka Goto, Katunobu Itou, and Satoru Hayamizu:
A Real-time System Detecting Filled Pauses
in Spontaneous Speech,
Information Processing Society of Japan SIG Notes,
99-SLP-27-2, Vol.99, No.64, July 1999. (in Japanese)
Back to:
Please E-mail comments and questions to
Masataka GOTO
<m.goto [at] aist.go.jp>
last update: May 11, 2000