Grid Neuron Model for Spatial Navigation

Abstract: Cognitive robots are required to work in dangerous areas since humans are unable to due to health-related restrictions. Robot interaction with the working environment is hampered by the fact that a robot cannot learn the spatial semantics of the environment or an object. In this work, a computational agent is created to address this issue. This agent learns cognitive maps from input spatial data of an environment or an item by simulating the behaviour of place neurons and grids. It is suggested that a novel quadrant-based modelling strategy be used to simulate the behaviour of the grid neuron, which, like the real grid neuron, can produce periodic hexagonal grid-like output patterns from the input body movement.