This project is an innovative undertaking, combining insights of computational neuroscience, neuroethology, computer vision and robotics.
It aims to develop a biologically-inspired computationally simple model of complex motion detection that could be used to drive generic actuators for robotics and other applications. Specifically, we will investigate the extent to which a computational model of biological constrained neural systems of insects, with an emphasis in the locust and bees, is able to effectively and robustly detect collisions and trigger evasion maneuvres in an efficient manner, in any light condition.
Our neuronal model will include components for collision avoidance. This system use a well-described insect collision detection system that consists of the locust Lobula Giant Movement Detector (LGMD) and its postsynaptic partner, the Descending Contralateral Movement Detector (DCMD).
In order to develop a very realistic model of the LGMD neuron, we are collecting physiological recordings from the LGMD pathway to complex visual stimuli and the the corresponding neuronal responses obtained.
The experiments are being conducted in robots (Pioneer 3DX) to show the effictiveness, feasibility and robustness of the proposed model.