Advertisement

Image Compression by Vector Quantization with Recurrent Discrete Networks

  • Domingo López-Rodríguez
  • Enrique Mérida-Casermeiro
  • Juan M. Ortiz-de-Lazcano-Lobato
  • Ezequiel López-rubio
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4132)

Abstract

In this work we propose a recurrent multivalued network, generalizing Hopfield’s model, which can be interpreted as a vector quantifier. We explain the model and establish a relation between vector quantization and sum-of-squares clustering. To test the efficiency of this model as vector quantifier, we apply this new technique to image compression. Two well-known images are used as benchmark, allowing us to compare our model to standard competitive learning. In our simulations, our new technique clearly outperforms the classical algorithm for vector quantization, achieving not only a better distortion rate, but even reducing drastically the computational time.

Keywords

Mean Square Error Image Compression Distortion Function Lossless Compression Competitive Learning 
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.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Ahalt, S.C., Krishnamurthy, A.K., Chen, P., Melton, D.E.: Competitive learning algorithms for vector quantization. Neural Networks 3, 277–290 (1990)CrossRefGoogle Scholar
  2. 2.
    Amerijckx, C., Verleysen, M., Thissen, P., et al.: Image compression by self-organized Kohonen map. IEEE Trans. Neural Networks 9(3), 503–507 (1998)CrossRefGoogle Scholar
  3. 3.
    Daugman, J.G.: Complete discrete 2-D Gabor transforms by neural networks for image analysis and compression. IEEE Trans. Acoustics, Speech Signal Process. 36(7), 1169–1179 (1988)MATHCrossRefGoogle Scholar
  4. 4.
    Dony, R.D., Haykin, S.: Optimally adaptive transform coding. IEEE Trans. Image Process. 4(10), 1358–1370 (1995)CrossRefGoogle Scholar
  5. 5.
    Egmont-Petersen, M., de Ridder, D., Handels, H.: Image processing with neural networksa review. Pattern Recognition 35, 2279–2301 (2002)MATHCrossRefGoogle Scholar
  6. 6.
    Erdem, M.H., Ozturk, Y.: A New family of Multivalued Networks. Neural Networks 9(6), 979–989 (1996)CrossRefGoogle Scholar
  7. 7.
    Hauske, G.: A self organizing map approach to image quality. Biosystems 40(1-2), 93–102 (1997)CrossRefGoogle Scholar
  8. 8.
    Hopfield, J.J.: Neural networks and physical systems with emergent collective computational abilities. Proc. of National Academy of Sciences USA 79, 2254–2558 (1982)MathSciNetGoogle Scholar
  9. 9.
    Mérida Casermeiro, E.: Red Neuronal recurrente multivaluada para el reconocimiento de patrones y la optimización combinatoria, Ph. D. dissertation. Univ. Málaga, España (2000) (in Spanish)Google Scholar
  10. 10.
    Mérida, E., Muñoz, J., Benítez, R.: A recurrent multivalued neural network for the N-queens problem. In: Mira, J., Prieto, A.G. (eds.) IWANN 2001. LNCS, vol. 2084, pp. 522–529. Springer, Heidelberg (2001)CrossRefGoogle Scholar
  11. 11.
    Mérida-Casermeiro, E., López-Rodríguez, D.: Multivalued Neural Network for Graph MaxCut Problem. Lecture Series on Computer and Computational Sciences, vol. 1, pp. 375–378 (2004)Google Scholar
  12. 12.
    Mitra, S., Yang, S.Y.: High fidelity adaptive vector quantization at very low bit rates for progressive transmission of radiographic images. J. Electron. Imaging 8(1), 23–35 (1999)CrossRefGoogle Scholar
  13. 13.
    Muñoz-Perez, J., Gomez-Ruiz, J.A., Lopez-Rubio, E., Garcia-Bernal, M.A.: Expansive and Competitive Learning for Vector Quantization. Neural Processing Letters 15, 261–273 (2002)MATHCrossRefGoogle Scholar
  14. 14.
    Ozturk, Y., Abut, H.: System of associative relationships (SOAR). In: Proceedings of ASILOMAR (1997)Google Scholar
  15. 15.
    Pal, N.R., Bezdek, J.C., Tsao, E.C.: Generalized clustering networks and Kohonens self-organizing scheme. IEEE Trans. Neural Networks 4(4), 549–557 (1993)CrossRefGoogle Scholar
  16. 16.
    Rizvi, S.A., Wang, L.C., Nasrabadi, N.M.: Nonlinear vector prediction using feedforward neural networks. IEEE Trans. Image Process 6(10), 1431–1436 (1997)CrossRefGoogle Scholar
  17. 17.
    Skarbek, W., Cichocki, A.: Robust image association by recurrent neural subnetworks. Neural Process. Lett. 3, 131–138 (1996)CrossRefGoogle Scholar
  18. 18.
    Tzovaras, D., Strintzis, M.G.: Use of nonlinear principal component analysis and vector quantization for image coding. IEEE Trans. Image Process. 7(8), 1218–1223 (1998)MATHCrossRefMathSciNetGoogle Scholar
  19. 19.
    Ueda, N., Nakano, R.: A new competitive learning approach based on an equidistortion principle for designing optimal vector quantizers. Neural Networks 7(8), 1211–1227 (1994)CrossRefGoogle Scholar
  20. 20.
    Wang, L.C., Rizvi, S.A., Nasrabadi, N.M.: A modular neural network vector predictor for predictive image coding. IEEE Trans. Image Process. 7(8), 1198–1217 (1998)MATHCrossRefMathSciNetGoogle Scholar
  21. 21.
    Weingessel, A., Bischof, H., Hornik, K., et al.: Adaptive combination of PCA and VQ networks. IEEE Trans. Neural Networks 8(5), 1208–1211 (1997)CrossRefGoogle Scholar
  22. 22.
    Yair, E.K., Zeger, K., Gersho, A.: Competitive learning and soft competition for vector quantizer design. IEEE Trans. Signal Processing 40(2), 294–308 (1992)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2006

Authors and Affiliations

  • Domingo López-Rodríguez
    • 1
  • Enrique Mérida-Casermeiro
    • 1
  • Juan M. Ortiz-de-Lazcano-Lobato
    • 2
  • Ezequiel López-rubio
    • 2
  1. 1.Department of Applied MathematicsUniversity of MálagaMálagaSpain
  2. 2.Department of Computer Science and Artificial IntelligenceUniversity of MálagaMálagaSpain

Personalised recommendations