Evolutionary Algorithms Research Group (EARes)

Switch to Russian


Wellcome to the web-site of Evolutionary Algorithms Research Group (EARes) at Tomsk Polytechnic University!

(under construction)


Scientific Interests:

  • Evolutionary algorithms (EA)
  • Neuroevolutionary algorithms (NEA)
  • Digital images processing using EAs and NEAs


    Projects:

  • Development of automation technologies of digital images enhancement on the base of evolving artificial neural network application (funded by Russian Foundation for Basic Researches, project 06-08-00840)

      A new method for monochrome and color images enhancement is proposed. The method uses artificial neural network (ANN) which is designed and trained with use of genetic algorithm. For this purpose an adaptive neuroevolutionary algorithm is developed, that implements evolutionary approach to design and training of neural network for images enhancement. The topology of ANN is tuned simultaneously with connections weights. For implementation of neuroevolutionary algorithm we used genuine adaptive operators of crossing (recombination) and mutation (variation), which respect structure of ANN. Also an original strategy of population size adaptation with respect to the characteristics of evolutionary search process is used as well. The way of the training of ANN for local-adaptive images enhancement is proposed. Alternative methods of images enhancement on the basis of wavelet transform and genetic algorithm are also introduced.

      Series of experiments for monochrome and color images enhancement using developed processing methods were hold, which results show that the processing speed of three-stage neural processing method exceeds by the order of 1 method on the basis of genetic algorithm only. The results of processing are comparable with that of more sophisticated technology adopting model of human color perception ("Multi-Scale Retinex", NASA) and the processing speed for the developed method is most likely higher.

      We plan further development of proposed methods and approaches and also implementation of neural image quality evaluation technique that adapts to the subjective demands and perception properties of user. We also plan to extend application of developed approach for color-correction and edge detection problems. Creation of software complex for automated real-time images processing on the basis of developed methods is planned.

  • Genetic algorithms research
  • Neuroevolutionary algorithms research

    Back to top


      Selected publications:

    1. Tsoy Y.R., Spitsyn V.G., Chernyavsky A.V. Method of monochrome and color images enhancement with use of evolving neural network // Information technologies, 2006, No. 7, pp. 27-33. (in Russian)
    2. Tsoy Y.R., Spitsyn V.G. Using genetic algorithm with adaptive mutation mechanism for neural networks design and training // Optical memory and neural networks, 2004, Vol. 13, No. 4, pp. 225-232.
    3. Red'ko V.G., Tsoy Yu. R. Estimation of the Efficiency of Evolution Algorithms // Doklady Mathematics (Doklady Akademii nauk), 2005. No. 72, pp. 810-813.
    4. Tsoy Y.R., Spitsyn V.G. Digital images enhancement with use of evolving neural networks // Proceedings of the IX International Conference Parallel Problems Solving from Nature(PPSN-IX), Reykjavik, Iceland, September 9-13, 2006. Lecture Notes in Computer Science, Vol. 4193. Berlin: Springer-Verlag, 2006. pp. 593-602. http://www.springerlink.com/content/m20qx311u75p5937/.
    5. Belousov A.A., Sidorov D.V., Spitsyn V.G. Applying wavelets and evolutionary algorithms to automatic image enhancement // XIII International Symposium "Atmospheric and Ocean Optics. Atmospheric Physics", Tomsk, July 2-6, 2006, pp. 104.
    6. Chernyavskii A.V., Tsoy Yu.R, Spitsyn V.G. Image processing using evolving neural network // XIII International Symposium "Atmospheric and Ocean Optics. Atmospheric Physics", Tomsk, July 2-6, 2006, pp. 104.
    7. Spitsyn V.G., Tsoy Yu. R. Application of Evolving Artificial Neural Network for Image Processing // USNC/URSI National Radio Science and AMEREM Meetings, 9-14 July 2006, Albuquerque, New Mexico, USA, P. 745
    8. Tsoy Y.R. Evolutionary Algorithms Design: State of the Art and Future Perspectives // Proceedings of IEEE East-West Design and Test Workshop (EWDTW'06). - Sochi, Russia, September 15-19, 2006. - P. 375-379.
    9. Tsoy Yu.R., Spitsyn V.G. Using genetic algorithm with adaptive mutation mechanism for neural networks design and training // Proceedings of the 9-th Korea - Russia International Symposium on Science and Technology "Korus-2005". Novosibirsk. Russia. June 26-July 2. 2005. Vol. 1. P. 237-241.
    10. Yankovskaya A.E., Tsoy Y.R. Optimization of a Set of Tests Selection Satisfying the Criteria Prescribed Using Compensatory Genetic Algorithm // Proceedings of IEEE East-West Design & Test Workshop (EWDTW'05), Ukraine, 2005. P. 123-126.
    11. Spitsyn V.G., Fedotov I.V., Tsoy Yu.R. Stochastic model of electromagnetic wave propagation in stratified absorbing media inclusive a semitransparent objects // 2004 National Radio Science Meeting (URSI 2004). Boulder, USA, 2004. P. 364.
    12. Spitsyn V.G., Tsoy Y.R., Fedotov I.V. Application of optimization methods for modeling of nonlinear electromagnetic wave interaction with random discrete media // 2004 USNC/ URSI National Radio Science Meeting Digest, Monterey, California. USA. June 20-26. 2004. P. 364.
    13. Spitsyn V.G. Stochastic model of electromagnetic wave propagation in periodic and stratified absorbing media inclusive a semitransparent object // IEEE Antennas and Propagation Society International Symposium Digest. Monterey. California. USA. June 20-26. 2004. Vol. 1. P. 930-933.
    14. Tsoy Yu.R., Spitsyn V.G. Using design patterns for design of software environment for researches in genetic algorithms // Proceedings of the 8-th Korea - Russia International Symposium on Science and Technology "Korus-2004". Tomsk. Russia. June 28-July 6. 2004. Vol. 1, P. 166-168.
    15. Spitsyn V.G. Stochastic model of interaction of electromagnetic signal with oscillator neural network // Journal of Applied Electromagnetism. 1998. Vol. 1. No 3. P. 12 - 19.

    Back to top


    People:

  • Spitsyn Vladimir Grigorievich, full Dr.Sc., PhD -- group leader
      Scientific interests: electromagnetic waves propagation, modelling, artificial neural networks, genetic algorithms, image processing
  • Tsoy Yury Robertovich, MSc, PhD -- researcher
      Scientific interests: genetic algorithms, artificial neural networks, artificial life, image processing and analysis, self-organization
  • Belousov Artem Anatolievich, MSc, PhD student -- researcher
      Scientific interests: image processing and analysis algorithms, artificial intelligence methods
  • Chernyavsky Alexander Valerievich, MSc, PhD student -- researcher
      Scientific interests: fuzzy logic, artificial neural networks, image enhancement
  • Fedotov Iliya Victorovich, PhD student -- researcher
      Scientific interests: information technologies, image processing, mathematical modelling

    Back to top


    Contacts:

  • Phone: 8-3822-418-912

  • E-mail: eares [AT] narod.ru

    Back to top

    • Home

    Projects

    Publications

    People

    Contacts

    Links:

  • Computer Engineering Department
  • "Cybernetic Center" Institute
  • Tomsk Polytechnic University
    • Copyright © EARes Group. 2006.

    Сайт создан в системе uCoz