Multivariate Statistical Analysis and One-Pass Vector Quantization

  • Qi (Peter) LiEmail author
Part of the Signals and Communication Technology book series (SCT)


Current speaker authentication algorithms are largely based on multivariate statistical theory. In this chapter, we introduce the most important technical components and concepts of multivariate analysis as they apply to speaker authentication: the multivariate Gaussian (also called normal) distribution, principal component analysis (PCA), vector quantization (VQ), and segmental K-means. These fundamental techniques have been used for statistical pattern recognition and will be used in our further discussions throughout this book. Understanding the basic concepts of these techniques is essential for understanding and developing speaker authentication algorithms.


Gaussian Mixture Model Multivariate Statistical Analysis Vector Quantization Speaker Recognition Training Vector 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Duda, R. O., Hart, P. E., Stork, D. G.: Pattern Classification, Second edn. John & Wiley, New York (2001)Google Scholar
  2. 2.
    Fischer, T. R., Dicharry, R. M.: “Vector quantizer design for gaussian, gamma, and laplacian sources”. IEEE Transactions on Communications. COM-32, 1065–1069 (1984)CrossRefGoogle Scholar
  3. 3.
    Gray, R. M., Linde, Y.: “Vector quantizers and predictive quantizers for Gauss-Markov sources”. IEEE Transactions on Communications. COM-30, 381–389 (1982)CrossRefGoogle Scholar
  4. 4.
    Huang, C. M., Harris, R. W.: “A comparison of several vector quantization codebook generation approaches”. IEEE Transactions on Image Processing 2, 108–112 (1993)CrossRefGoogle Scholar
  5. 5.
    Huang, X., Acero, A., Acero, A., Acero, A.: Spoken language processing. Prentice Hall PTR, NJ (2001)Google Scholar
  6. 6.
    Johnson, R. A., Wichern, D. W.: Applied Multivariate Statistical Analysis. Prentice Hall, New Jersey (1988)zbMATHGoogle Scholar
  7. 7.
    Juang, B.-H., Rabiner, L. R.: “The segmental k-means algorithm for estimating parameters of hidden Markove models”. IEEE Trans. Acoustics, Speech and Signal Processing 38, 1639–1641 (1990)zbMATHCrossRefGoogle Scholar
  8. 8.
    Li, Q., Swaszek, P. F., “One-pass vector quantizer design by sequential pruning of the training data,” in Proceedings of International Conference on Image Processing (Washington, DC), October 1995Google Scholar
  9. 9.
    Li, Q. and Tufts, D. W., “Improving discriminant neural network (DNN) design by the use of principal component analysis,” in Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing (Detroit, MI), pp. 3375–3379, May 1995Google Scholar
  10. 10.
    Li, Q. and Tufts, D. W., “Synthesizing neural networks by sequential addition of hidden nodes,” in Proceedings of the IEEE International Conference on Neural Networks (Orlando, FL), pp. 708–713, June 1994Google Scholar
  11. 11.
    Li, Q., Tufts, D. W., Duhaime, R., August, P., “Fast training algorithms for large data sets with application to classification of multispectral images,” in Proceedings of the IEEE 28th Asilomar Conference (Pacific Grove), October 1994Google Scholar
  12. 12.
    Linde, Y., Buzo, A., Gray, R. M.: “An algorithm for vector quantizer design”. IEEE Transactions on Communications. COM-28, 84–95 (1980)CrossRefGoogle Scholar
  13. 13.
    MacQueen, J., “Some methods for classification and analysis of multivariate observations,” in Proc. 5th Berkeley Symp. Math. Stat., Prob. pp. 281–296 (1967)Google Scholar
  14. 14.
    Max, J.: “Quantizing for minimum distortion”. IEEE Transactions on Information Theory. IT-6, 7–12 (1960)MathSciNetCrossRefGoogle Scholar
  15. 15.
    Paez, M. D., Glisson, T. H.: “Minimum mean-square-error quantization in speech pcm and dpcm systems”. IEEE Transactions on Communications. COM-20, 225–230 (1972)CrossRefGoogle Scholar
  16. 16.
    Rabiner, L. R., Wilpon, J. G., Wilpon, J. G., Wilpon, J. G.: “A segmental k-means training procedure for connected word recognition”. AT&T Technical Journal 65, 21–31 (1986)Google Scholar
  17. 17.
    Soong, F. K., Rosenberg, A. E., Juang, B.-H.: “A vector quantization approach to speaker recognition”. AT&T Technical Journal 66, 14–26 (1987)Google Scholar
  18. 18.
    Swaszek, P. F., “Low dimension/moderate bitrate vector quantizers for the laplace source,” in Abstracts of IEEE International Symposium on Information Theory, p. 74, (1990)Google Scholar
  19. 19.
    Swaszek, P. F., “Vector quantization for image compression,” in Proceedings of Princeton Conference on Information Sciences and Systems (Princeton, NJ), pp. 254–259, March 1986Google Scholar
  20. 20.
    Swaszek, P. F., Narasimhan, A., “Quantization of the correlated gaussian source,” in Proceedings of Princeton Conference on Information Sciences and Systems (Princeton, NJ), pp. 784–789, March 1988Google Scholar
  21. 21.
    Swaszek, P. F., Thomas, J. B.: “Optimal circularly symmetric quantizers”. Journal of Franklin Institute 313, 373–384 (1982)CrossRefGoogle Scholar
  22. 22.
    Tufts, D. W. and Li, Q., “Principal feature classification,” in Neural Networks for Signal Processing V, Proceedings of the 1995 IEEE Workshop (Cambridge, MA), August 1995Google Scholar
  23. 23.
    Viterbi, A. J.: “Error bounds for convolutional codes and an asymptotically optimal decoding algorithm”. IEEE Transactions on Information Theory. IT-13, 260–269 (1967)CrossRefGoogle Scholar
  24. 24.
    Wilson, S. G.: “Magnitude/phase quantization of independent gaussian variates”. IEEE Transactions on Communications. COM-28, 1924–1929 (1980)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg  2012

Authors and Affiliations

  1. 1.Li Creative Technologies (LcT), IncFlorham ParkUSA

Personalised recommendations