대한음성학회지:말소리 (MALSORI)
- 제63호
- /
- Pages.113-124
- /
- 2007
- /
- 1226-1173(pISSN)
신뢰도 벡터 기반의 다단계 음성인식
Multi-stage Speech Recognition Using Confidence Vector
-
전형배
(한국전자통신연구원(ETRI)음성/언어정보연구센터 음성처리연구팀) ;
- 황규웅 (한국전자통신연구원(ETRI)음성/언어정보연구센터 자동통역연구팀) ;
-
정훈
(한국전자통신연구원(ETRI)음성/언어정보연구센터 음성처리연구팀) ;
-
김승희
(한국전자통신연구원(ETRI)음성/언어정보연구센터 자동통역연구팀) ;
-
박준
(한국전자통신연구원(ETRI)음성/언어정보연구센터 자동통역연구팀) ;
-
이윤근
(한국전자통신연구원(ETRI)음성/언어정보연구센터 음성처리연구팀)
-
Jeon, Hyung-Bae
(ETRI) ;
- Hwang, Kyu-Woong (ETRI) ;
-
Chung, Hoon
(ETRI) ;
-
Kim, Seung-Hi
(ETRI) ;
-
Park, Jun
(ETRI) ;
-
Lee, Yun-Keun
(ETRI)
- 발행 : 2007.09.30
초록
In this paper, we propose a use of confidence vector as an intermediate input feature for multi-stage based speech recognition architecture to improve recognition accuracy. A multi-stage speech recognition structure is introduced as a method to reduce the computational complexity of the decoding procedure and then accomplish faster speech recognition. Conventional multi-stage speech recognition is usually composed of three stages, acoustic search, lexical search, and acoustic re-scoring. In this paper, we focus on improving the accuracy of the lexical decoding by introducing a confidence vector as an input feature instead of phoneme which was used typically. We take experimental results on 220K Korean Point-of-Interest (POI) domain and the experimental results show that the proposed method contributes on improving accuracy.