Hardware Implementation of a Bio-plausible Neuron Model for Evolution and Growth of Spiking Neural Networks on FPGA

Author

Shayani, Hooman ; Bentley, Peter J. ; Tyrrell, Andrew M.

Author_Institution

Dept. of Comput. Sci., UCL, London

fYear

2008

fDate

22-25 June 2008

Firstpage

236

Lastpage

243

Abstract

We propose a digital neuron model suitable for evolving and growing heterogeneous spiking neural networks on FPGAs by introducing a novel flexible dendrite architecture and the new PLAQIF (piecewise-linear approximation of quadratic integrate and fire) soma model. A network of 161 neurons and 1610 synapses was simulated, implemented, and verified on a Virtex-5 chip with 4210 times real-time speed with 1 ms resolution. The parametric flexibility of the soma model was shown through a set of experiments.

Keywords

electronic engineering computing; field programmable gate arrays; neural nets; piecewise linear techniques; FPGA; Virtex-5 chip; bioplausible neuron model; flexible dendrite architecture; hardware implementation; parametric flexibility; piecewise-linear approximation; quadratic integrate and fire soma model; spiking neural networks; Artificial neural networks; Computer science; Evolution (biology); Face detection; Field programmable gate arrays; Neural network hardware; Neural networks; Neurons; Object detection; Parallel processing; Development; Digital Spiking Neuron; FPGA; Growth;

fLanguage

English

Publisher

ieee

Conference_Titel

Adaptive Hardware and Systems, 2008. AHS '08. NASA/ESA Conference on

Conference_Location

Noordwijk

Print_ISBN

978-0-7695-3166-3

Type

conf

DOI

10.1109/AHS.2008.13

Filename

4584279