Neuromorphic Engineering

Neuromorphic Visual Perception

Overview

The Neuromorphic Engineering team develops bio-inspired systems to tackle sensory and motor tasks commonly encountered in real world situations. The inspiration for using bio-inspired principles comes from the efficiency and robustness with which biological organisms are able to sense and navigate through their surroundings. By learning from biology, we move towards a better understand how biology achieves such efficiency, while simultaneously improving the efficiency and robustness of the robotic sensory and motor systems we are developing.

Neuromorphic engineering adopts a complete approach to tackling sensory motor problems, from custom design of bio-inspired front-end sensors in silicon, to design of back-end computational hardware for implementing neural networks on which to implement algorithms. The approach also involves the development and optimization of neural network algorithms to tackle the task at hand, whether it be visual motion estimation, tactile texture discrimination, or control of legged locomotion.