IPAB Workshop - 16/01/2020

Title:

Learning Cognitive Map Representations for Navigation by Sensory-motor Integration

Author:

Dongye ZHAO

State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, China

Abstract:

How to transform a mixed flow of sensory and motor information into a memory state of self-location and to build map-like representations of the environment are central questions in navigation research. Studies in neuroscience have shown that place cells in the hippocampus of rodent brains form dynamic cognitive representations of locations in the environment. In the talk, I shall introduce a neural network model (SeMINet), which learns cognitive map representations by integrating sensory and motor information while an agent is exploring a virtual environment. The biologically inspired SeMINet provides a unified model of perception and memory for the navigation task.