Efficient Continuous Vocabulary Clustering Modeling forTying Model Recognition Performance Improvement

안찬식; 오상엽

Journal of The Korea Society of Computer and Information

Abbr : JKSCI
2010, 15(1), pp.177-183
Publisher : The Korean Society Of Computer And Information
Research Area : Computer Science

안찬식 ¹ , 오상엽 ²

¹ 광운대학교
² 경원대학교

ABSTRACT

In continuous vocabulary recognition system by statistical method vocabulary recognition to be performed using probability distribution it also modeling using phoneme clustering for based sample probability parameter presume. When vocabulary search that low recognition rate problem happened in express vocabulary result from presumed probability parameter by not defined phoneme and insert phoneme and it has it's bad points of gaussian model the accuracy unsecure for one clustering modeling. To improve suggested probability distribution mixed gaussian model to optimized for based resemble Euclidean and Bhattacharyya distance measurement method mixed clustering modeling that system modeling for be searching phoneme probability model in clustered model. System performance as a result of represent vocabulary dependence recognition rate of 98.63%, vocabulary independence recognition rate of 97.91%.
KEYWORDS

clustering modeling , phoneme modeling , tying model , gaussian model , vocabulary recognition
REFERENCES (20)

[book] S.Young / 2002 / The HTK Book / Cambridge University Engineering Department

[book] L.R.Rabiner / 1993 / Fundamentals of speech recognition / Prentice Hall

[journal] 안태옥 / 2005 / A Study on the Optimization of State Tying Acoustic Models using Mixture Gaussian Clustering / 전자공학회논문지 - SP / 대한전자공학회 42 (6) : 937~946

[book] D.Jurafsky / 2000 / Speech and Language Processing / Prentice-Hall

[journal] 우인성 / 2007 / The Study of Korean Speech Recognition for Various Continue HMM / 전기전자학회논문지 / 한국전기전자학회 11 (2) : 89~94

[journal] 이호웅 / 2008 / A Study on Vocabulary-independent Continuous Speech Recognition System for Intelligent Home Network System / 한국ITS학회 논문지 / 한국ITS학회 7 (2) : 37~42

[journal] K.Demuynck, / 1997 / A static lexicon network repre-sentation for cross-word context dependent phones / Proc.EUROSPEECH 1 : 143~146

[journal] 조영수 / 1994 / HMM을 이용한 한국어 음소인식 / 대한전자공학회 학술발표회 논문집 16 (1) : 81~84

[thesis] M.F.Gales / 1995 / Model-based techniques for nosie robust speech recognition / Ph. D / University of Cambridge

[journal] 안찬식 / 2009 / Vocabulary Recognition Retrieval Optimized System using MLHF Model / 한국컴퓨터정보학회논문지 / 한국컴퓨터정보학회 14 (10) : 217~223

[thesis] A.S.Manos / 1996 / A study on out-of-vocabulary word modeling for a segment-based keyword spotting system / Master / MIT

[journal] 김광호 / 2008 / Vocabulary Coverage Improvement for Embedded Continuous Speech Recognition Using Knowledgebase / 말소리 / 대한음성학회 1 (68) : 115~126

[journal] 김동주 / 2006 / Context-Weighted Metrics for Example Matching / 전자공학회논문지 - CI / 대한전자공학회 43 (6) : 43~51

[journal] 김기백 / 2007 / 음성인식 기반 컨텐츠 네비게이션 시스템 / 한국컴퓨터정보학회지 15 (1) : 99~102

[confproc] S.Ortmanns / 1997 / Look-ahead Techniques for Fast Beam Search / Proc.IEEE ICASSP-1997 : 1783~1786

[confproc] Kris Demuynck / 2003 / FLaVor:a flexible architecture for LVCSR / EUROSPEECH-2003 : 1973~1976

[confproc] Justin Zobel / 1996 / Phonetic String Matching:Lessons from Information Retrival / SIGIR'96 : 166~173

[confproc] T.Jitsuhiro,S. / 1998 / Rejection of out-of-vocabulary words using phoneme confidence likelihood / ICASSP : 217~220

[confproc] L.R.Bahl / 1992 / A Fast Match for Continuous Speech Recognition Using Allophonic Models / Proc.IEEE ICASSP-92, : 17~21

[confproc] W.Daelemans / 1999 / Memorybased shallow parsing / Proc.CoNLL : 53~60
KCI Citation (14)

KJC Korea
Journal Central

Statistics

Tools