04-11-2017, 05:36 PM
This paper constructs fully parallel NN hardware architecture, FPGA has been used to reduce neuron hardware by design the activation function inside the neuron without using
lookup table as in most researches, to perform an efficient NN. It consist of two main parts; the first part covers network training using MATLAB program, the second part represents the
hardware implementation of the trained network through Xilinx high performance Virtex2 FPGA schematic entry design tools.
lookup table as in most researches, to perform an efficient NN. It consist of two main parts; the first part covers network training using MATLAB program, the second part represents the
hardware implementation of the trained network through Xilinx high performance Virtex2 FPGA schematic entry design tools.