ICANN ’93 pp 421-421 | Cite as

Weighted Distance Measure for Speaker-Independent Digit Recognition with Hidden-Control Neural Network

  • Dou-Seok Kim
  • Soo-Young Lee
Conference paper

Abstract

Hidden control neural network (HCNN) incorporates dynamic state transition of the hidden Markov model (HMM) with Viterbi segmentation. [1] In this paper we report a different hidden-state representation and weighted distance measure for better recognition rates.

Keywords

Covariance 

References

  1. [1]
    E. Levin, “Modeling time varying systems using hidden control neural architecture,” Proc. NIPS-3, pp. 147–154, 1991.Google Scholar
  2. [2]
    Y. Tohkura, “A weighted cepstral distance measure for speech recognition,” IEEE Trans. ASSP, pp. 1414–1422, 1987.Google Scholar

Copyright information

© Springer-Verlag London Limited 1993

Authors and Affiliations

  • Dou-Seok Kim
    • 1
  • Soo-Young Lee
    • 1
  1. 1.Department of Electrical EngineeringKorea Advanced Institute of Science and TechnologyYusong-gu, TaejonKorea (South)

Personalised recommendations