Cognitive Consistency Routing Algorithm of Capsule-Network

Abstract

Artificial Neural Networks (ANNs) are computational models inspired by the central nervous system (especially the brain) of animals and are used to estimate or generate unknown approximation functions that rely on a large number of inputs. The Capsule Neural Network [1] is a novel structure of Convolutional Neural Networks (CNN) which simulates the visual processing system of the human brain. In this paper, we introduce a psychological theory which is called Cognitive Consistency to optimize the routing algorithm of Capsnet to make it more close to the working pattern of the human brain. Our experiments show that progress had been made compared with the baseline.