Tied Mixture modeling optimization for Korean-digit in the embedded ASR system

2004 IEEE International Symposium on Consumer Electronics - Proceedings

Volumn , Issue , 2004, Pages 595-599

  Kim, Kihyeon ^a   Ko, Hanseok ^a  

^a Korea University   (South Korea)

Author keywords

Embedded ASR System; Exclusive HMM's; Korean Digits; Tied Mixture Model

Indexed keywords

MATHEMATICAL MODELS; PARAMETER ESTIMATION; PROBABILITY DENSITY FUNCTION; PROBLEM SOLVING; SPEECH RECOGNITION; VECTOR QUANTIZATION;

EMBEDDED ASR SYSTEM; EXCLUSIVE HIDDEN MARKOV MODEL (HMM); KOREAN DIGITS; TIED MIXTURE MODEL;

EMBEDDED SYSTEMS;

EID: 10444277350     PISSN: None     EISSN: None     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (5)

1. S. J. Young, The HTK book, Ver. 3.2, University of Cambridge, 2002.
2. X. Huang, et al, "Semi-continuous Hidden Markov Models with Maximum Likelihood VQ", IEEE Workshop on Speech Recognition, New York, 1998.
3. S. J. Young, "Large Vocabulary Continuous Speech Recognition: a review", IEEE Signal Processing Magazine, Vol. 13, Issue. 5, Sept. 1996
4. J.H. Park, et al, "Compact Acoustic Model for Embedded Implementation", accepted to Proc. ICSLP'04, Oct. 2004.
5. B. -H. Juang, et al, "Minimum Classfication Error Rate Methods for Speech Recognition", IEEE Transactions on Speech and Audio Processing, Vol. 5, No. 3, pp. 257-265, 1997.