High Performance Implementation of an FPGA-Based Sequential DT-CNN

  • J. Javier Martínez-Alvarez
  • F. Javier Garrigós-Guerrero
  • F. Javier Toledo-Moreo
  • J. Manuel Ferrández-Vicente
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4528)

Abstract

In this paper an FPGA-based implementation of a sequential discrete time cellular neural network (DT-CNN) with 3×3 templates is described. The architecture is based on a single pipelined cell which is employed to emulate a CNN with larger number of neurons. This solution diminishes the use of hardware resources on the FPGA and allows the cell to process real time input data in a sequential mode. Highly efficient FPGA implementation has been achieved by manual design based on low level instantiation and placement of hardware primitives. The Intellectual Property Core offers an appropriate tradeoff between area and speed. Our architecture has been developed to assist designers implementing discrete CNN models with performance equivalent to hundreds or millions of neurons on low cost FPGA-based systems.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Chua, L.O., Yang, L.: Cellular neural networks: theory. IEEE Trans. Circuits and Systems 35 (1988)Google Scholar
  2. 2.
    Crounse, K.R., Chua, L.O.: Methods for image processing and pattern formation in Cellular Neural Networks: a tutorial. IEEE Transactions on Circuits and Systems I 42(10), 583–601 (1995)CrossRefMathSciNetGoogle Scholar
  3. 3.
    Balya, D., Rekeczky, C., Roska, T.: A realistic mammalian retinal model implemented on complex cell CNN universal machine. In: IEEE Intern. Symp. on Circuits and Systems, pp. 26–29 (2002)Google Scholar
  4. 4.
    Chua, L.O.: The CNN: a brain-like computer Neural Networks. In: IEEE Intern. Joint Conference, vol. 1, pp. 25–29 (2004)Google Scholar
  5. 5.
    Roska, T., Rodriguez-Vazquez, A.: Review of CMOS implementations of the CNN universal machine-type visual microprocessors. In: IEEE Int. Symp. on Circuits and Systems, ISCAS 2000, Geneva, Italia, vol. 2, pp. 120–123 (2000)Google Scholar
  6. 6.
    Linan, G., Rodriguez-Vazquez, A., Espejo, S., Dominguez-Castro, R.: ACE16K: a 128x128 focal plane analog processor with digital I/O. In: IEEE Int. Work. on Cellular Neural Networks and Their Applications, CNNA 2002, pp. 132–139 (2002)Google Scholar
  7. 7.
    Crounse, K.R., Chua, L.O.: The CNN Universal Machine is as universal as a Turing Machine. IEEE Transactions on Circuits and Systems I 43(4), 353–355 (1996)CrossRefMathSciNetGoogle Scholar
  8. 8.
    Hsin, C., Yung, H., Chang, C., Teh, L., Chun, C.: Image-processing algorithms realized by discrete-time cellular neural networks and their circuit implementations. Chaos Solitons and Fractals Journal 29(5), 1100–1108 (2006)CrossRefGoogle Scholar
  9. 9.
    Martínez, J.J., Toledo, F.J., Ferrández, J.M.: Architecture Implementation of a Discrete Cellular Neuron Model (DT-CNN) on FPGA. In: The Int. Society for Optical Engineering, Bioengineered and bioinspired systems II, pp. 332–340 (2005)Google Scholar
  10. 10.
    Nagy, Z., Szolgay, P.: Configurable multilayer CNN-UM emulator on FPGA. IEEE Trans. on Circuits and Systems I 50(6), 774–778 (2003)CrossRefGoogle Scholar
  11. 11.
    Xilinx Inc. Virtex-4 User Guide, data sheet(ug070) (2004), http://www.xilinx.com

Copyright information

© Springer Berlin Heidelberg 2007

Authors and Affiliations

  • J. Javier Martínez-Alvarez
    • 1
  • F. Javier Garrigós-Guerrero
    • 1
  • F. Javier Toledo-Moreo
    • 1
  • J. Manuel Ferrández-Vicente
    • 1
  1. 1.Dpto. Electrónica, Tecnología de Computadoras y Proyectos, Universidad Politécnica de Cartagena, 30202 CartagenaSpain

Personalised recommendations