Clustering In Tied Mixture HMM Using Homogeneous Centroid Neural Network 


Vol. 31,  No. 9, pp. 853-858, Sep.  2006


PDF
  Abstract

TMHMM(Tied Mixture Hidden Markov Model) is an important approach to reduce the number of free parameters in speech recognition. However, this model suffers from a degradation in recognition accuracy due to its GPDF (Gaussian Probability Density Function) clustering error. This paper proposes a clustering algorithm, called HCNN(Homogeneous Centroid Neural network), to cluster acoustic feature vectors in TMHMM. Moreover, the HCNN uses the heterogeneous distance measure to allocate more code vectors in the heterogeneous areas where probability densities of different states overlap each other. When applied to Korean digit isolated word recognition, the HCNN reduces the error rate by 9.39% over CNN clustering, and 14.63% over the traditional K-means clustering.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

D. Park and W. Kim, "Clustering In Tied Mixture HMM Using Homogeneous Centroid Neural Network," The Journal of Korean Institute of Communications and Information Sciences, vol. 31, no. 9, pp. 853-858, 2006. DOI: .

[ACM Style]

Dong-Chul Park and Woo-Sung Kim. 2006. Clustering In Tied Mixture HMM Using Homogeneous Centroid Neural Network. The Journal of Korean Institute of Communications and Information Sciences, 31, 9, (2006), 853-858. DOI: .

[KICS Style]

Dong-Chul Park and Woo-Sung Kim, "Clustering In Tied Mixture HMM Using Homogeneous Centroid Neural Network," The Journal of Korean Institute of Communications and Information Sciences, vol. 31, no. 9, pp. 853-858, 9. 2006.

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880