Development of a Machine Learning Model with a Function of Amygdala for Rapid Image Processing

Abstract

In this paper, a method of implementing convolutional neural network that can quickly deal with dangerous situations is designed based on the processing rules of the amygdala of human brain. By studying the processing rules of the amygdala in the human brain, we can understand neuron activity when humans are at risk, build similar models, and test relevant data.  A neural network model can be constructed by changing structure, loss function and number of filters the general convolutional neural network model. The designed neural network model can quickly and accurately predict dangers. It was used to test data set and good results were obtained.