A STUDY ON SPEECH SUMMARY USING DEMI-PHONEME SYMBOLS GENERATED FROM SPEECH WAVES
-
- NAKAZAWA Masayuki
- Tsukuba Research Center, Real World Computing Partnership
-
- ENDO Takashi
- Tsukuba Research Center, Real World Computing Partnership
-
- FURUKAWA Kiyoshi
- Tsukuba Research Center, Real World Computing Partnership
-
- TOYOURA Jun
- Tsukuba Research Center, Real World Computing Partnership
-
- OKA Ryuichi
- Tsukuba Research Center, Real World Computing Partnership
Bibliographic Information
- Other Title
-
- 音声波形からの音素片記号系列を用いた音声要約と話題要約の検討
Search this article
Description
We propose a new approach to the speech interface based on superficial speech information and natural language information through generating a topic summary. We know, it is difficult for system to understand a spontaneous speech. The major problem of speech recognition is that difficulty of creating grammar and using as a man-machine interface. This paper describes a method for segmentation of topics from a spontaneous speech using by superficial speech information and natural language information.
Journal
-
- IEICE technical report. Speech
-
IEICE technical report. Speech 96 (93), 61-68, 1996-06-14
The Institute of Electronics, Information and Communication Engineers
- Tweet
Details 詳細情報について
-
- CRID
- 1571417127317075328
-
- NII Article ID
- 110003296379
-
- NII Book ID
- AN10013221
-
- Text Lang
- ja
-
- Data Source
-
- CiNii Articles