A New Voice Recognition Topology Based on the Human Auditory Cortex

Human brain consists of 4 lobes where each lobe has a special task. Each lobe is divided into several cortexes each of them constructed of millions of neurons. The brain functionality depends on these neurons and their interconnections. One of these cortexes is auditory cortex which is constructed of 6 layers and deals with the processing of auditory information transmitted from ear to the brain. The artificial neural network presented in this paper consisting of four modular parts recognizes the frequency of the human voice. The first part of the network eliminates frequencies which have weak intensities. The second part recognizes the pitch or the first harmony of the voice. The third part prepares inputs based on other harmonies of the frequency for the last part, and the last part which is an associative memory neural network, maps the input set to a desire output set.