A New Step in Arabic Speech Identification: Spoken Digit Recognition

  • Khalid Saeed
  • Mohammad K. Nammous


This work presents a new Algorithm to recognize separate voices of some Arabic words, the digits form zero to ten. Firstly we prepare our signal by pre-processing trial. Next the speech signal is processed as an image by Power Spectrum Estimation. For feature extraction, transformation and hence recognition, the algorithm of minimal eigenvalues of Toeplitz matrices together with other methods of speech processing and recognition are used. At the stage of classification many methods are tested from classical ones, which depend on the matrix theory, to different types of neuron networks, mainly radial basis functions neural networks. The success rate obtained in the presented experiments is almost ideal and exceeded 98% for many cases. The results have shown flexibility to extend the algorithm to speaker identification.


Speech Processing Recognition Burg's and Toeplitz Models Neural Networks 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

6 References

  1. [1]
    K. Saeed, M. Nammous, “Experimental Image-Based Algorithm for Spoken Arabic Digits Identification,” Computer Information Systems and Applications, Vol.1, pp.55–66, WSFiZ Press, Bialystok, Poland 2004.Google Scholar
  2. [2]
    K. Saeed, “Computer Graphics Analysis: A Criterion for Image Feature Extraction and Recognition,” Vol. 10, Issue 2, 2001, pp. 185–194, MGV-International Journal on Machine Graphics and Vision, Institute of Computer Science, Polish Academy of Sciences, Warsaw.Google Scholar
  3. [3]
    R. W. Schafer, L. R. Rabiner, “System for Automatic Formant Analysis of Voiced Speech,” /J. Acoust. Soc. Amer. Vol.47, Feb. 1970.Google Scholar
  4. [4]
    Andreas A., “Digital Filters: Analysis and Design,” McGraw-Hill, New York 1979.Google Scholar
  5. [5]
    Cz. Basztura, “Modele analizy i procedury w komputerowym rozpoznawaniu głosów,” (in Polish), prace naukowe ITiA Politechniki Wrocławskiej, no. 30, Wrocław 1989.Google Scholar
  6. [6]
    L. S. Marple, “Digital Spectral Analysis,” Englewood Cliffs, NJ: Prentice Hall, 1987.Google Scholar
  7. [7]
    Sadaoki Furui, “Digital Speech Processing, Synthesis, and Recognition,” Marcel Dekker, Inc. 2001.Google Scholar
  8. [8]
    R. Tadeusiewicz, “Sygnał mowy,” WKiŁ (in Polish), Warsaw 1988.Google Scholar
  9. [9]
    V. K. Ingle, J. G. Proakis, “Digital Signal Processing Using MATLAB,” Brooks Cole, July 1999.Google Scholar
  10. [10]
    K. Saeed, M. Kozłowski, A. Kaczanowski, “Metoda do rozpoznawania obrazów akustycznych izolowanych liter mowy”, Zeszyty Politechniki Białostockiej (in Polish), I-1/2002, pp. 181–207, Bialystok 2002.Google Scholar
  11. [11]
    K. Saeed, M. Kozłowski, “An Image-Based System for Spoken-Letter Recognition,” 10th Int. Conference CAIP'03 on Computer Analysis of Images and Patterns, August 2003, Groningen. Proceedings published in: Lecture Notes in Computer Science, Petkov and Westenberg (Eds.), pp. 494–502, LNCS 2756, Springer-Verlag Heidelberg: Berlin 2003.Google Scholar
  12. [12]
    K. Saeed, M. Tabedzki, “A New Hybrid System for Recognition of Handwritten-Script,” Invited for publication in International Scientific Journal of Computing, Institute of Computer Information Technologies, Volume 3, Issue 1, pp. 50–57, 2004, Ternopil, Ukraine 2004.Google Scholar
  13. [13]
    Shigeru Katagiri, “Handbook of Neural Networks for Speech Processing,” Artech House, Boston 2000.Google Scholar
  14. [14]
    M.W. Mak, W.G. Allen and G.G. Sexton, “Speaker identification using radial basis functions”, The Third International Conference on Artifical Neural networks, University of Northumbria at Newcastle, U.K 1998.Google Scholar

Copyright information

© Springer Science+Business Media, Inc. 2005

Authors and Affiliations

  • Khalid Saeed
    • 1
  • Mohammad K. Nammous
    • 1
  1. 1.Faculty of Computer ScienceBialystok Technical UniversityBialystokPoland

Personalised recommendations