PULSE UNIT CALCULATION OF THE ACTIVATION FUNCTION OF THE ARTIFICIAL NEURAL NETWORK

摘要

The utility model relates to the field of computing, namely, a pulse unit for calculating the activation function of an artificial neural network, including a sigma-delta modulator, the output of which is connected to the input of the averaging filter, the control input of the multiplexer and two inputs of the three-input adder, the output of the averaging filter is connected to the absolute value calculation unit, the output of which is connected to the first signal input of the multiplexer and to the inverter, which is connected to the third signal input of the multiplexer, the second input of the multiplexer is connected to generator 0, the output of the multiplexer is connected to a sigma-delta multiplier modulator, the output of which is connected to a three-input adder, the output of which is connected to a sigma-delta adder modulator, the output of which is the output of the impulse unit for calculating the activation function of the artificial neural network. The utility model provides a reduction in the resource intensity of the implementation of the activation functions of neurons of artificial neural networks on the basis of the FPGA.