A modified neural network model for Lobula Giant Movement Detector with additional depth movement feature

Abstract

Abstract-The Lobula Giant Movement Detector (LGMD) is a wide-field visual neuron that is located in the Lobula layer of the Locust nervous system. The LGMD increases its firing rate in response to both the velocity of the approaching object and its proximity. It has been found that it can respond to looming stimuli very quickly and can trigger avoidance reactions whenever a rapidly approaching object is detected. It has been successfully applied in visual collision avoidance systems for vehicles and robots. This paper proposes a modified LGMD model that provides additional movement depth direction information. The proposed model retains the simplicity of the previous neural network model, adding only a few new cells. It has been tested on both simulated and recorded video data sets. The experimental results shows that the modified model can very efficiently provide stable information on the depth direction of movement.

Towards Computational Models and Applications of Insect Visual Systems for Motion Perception: A Review

“…The experimental data is adapted from [101]. model encoding onset and offset responses by luminance increments and decrements, adapted from [114], (b) a modified LGMD1 model for multiple looming objects detection, adapted from [233], (c) a simplified LGMD1 model for collision avoidance of an UAV, adapted from [188], (d) a modified LGMD1 model with enhancement of collision selectivity, adapted from [133,132], (e) a modified LGMD1 model with a new layer for noise reduction and spikingthreshold mediation, adapted from [198,197], (f) an LGMD1 neural network based on the modelling of elementary motion detectors for collision detection in ground vehicle scenarios, adapted from [91]. Based on this LGMD1 modelling theory, a good number of models have been produced during the past two decades; these works have not only been extending and consolidating the LGMD1's original functionality for looming perception, but also investigating the possible applications to mobile machines like robots and vehicles.…”

Computational model of the LGMD neuron for automatic collision detection

“…The biological inspired neural network here proposed (figure 1) is based on previous models described on [6], [8].…”

“…The A (Approaching) and R (Receding) cells (modified from [8]) are two grouping cells for depth movement direction recognition. The D cell or Direction cell (∈ {−1, 0, 1} in case of receding, no movement and approaching object, respectively) is used to calculate the direction of movement (for further details, see our previous paper [9]).…”

“…The model continued to evolve [5], [7], [8], [6] and it was used in many applications for collision detection. However, further work is needed to develop more robust models that can account for complex aspects of visual motion.…”

“…In this article, we extend our previous LGMD model [9] which was able to achieve noise immunity [6] and direction sensitivity [8]. We propose to improve over the existing LGMD model by introducing a novel pixel mapping on the captured image that feeds the neural network.…”