A biomimetic collision-detection visual neural model coordinating self-and-lateral inhibitions
The 14th International Conference on Biomimetic and Biohybrid Systems (Living Machines 2025), 2025
Abstract
Biologically inspired visual systems provide elegant and efficient solutions to collision-detection challenges encountered by intelligent robots. Lobula Giant Movement Detectors (LGMD1 and LGMD2), neurons located in the locust optic lobe, are specialized in detecting approaching objects (looming perception) and have been widely modeled for integration into mobile robots. In bio-inspired robotic implementations of LGMD, inhibitory processes are crucial, as they help maintain selective responses to looming stimuli, enabling reliable collision avoidance. However, current robotic implementations of LGMD models often struggle with nearby translating movements, frequently generating false positive collision alerts. Recent biological studies have identified transmedulla afferent (TmA) neurons within the LGMD dendritic region, which act as a form of self-inhibition (SI). These neurons rapidly suppress intermediate neuronal activities in situ within the LGMD structure, effectively complementing lateral inhibition (LI). Together, SI and LI enhance the speci city of looming responses, reducing interference from translating motions. Despite their biological signi cance, these mechanisms have yet to be e ectively modeled and tested within arti cial robotic vision systems. In response, this study introduces a biohybrid visual neural model that incorporates SI and coordinates it with LI during looming perception. The proposed neural computation explicitly activates SI during initial looming events and during translating movements by leveraging spatial correlations within segmented, localized image areas, defined as the local visual eld (LVF). This innovative model has been integrated into a bio-inspired micro-robot, named Colias, serving as its sole collision-detection mechanism. Both o ine evaluations and real-world robotic tests demonstrate the e cacy of the biohybrid model in distinguishing looming from translating motions. Consequently, the robot exhibits signi cantly enhanced collision detection selectivity, closely resembling the capabilities observed in biological organisms.
Keywards
Biohybrid visual model, LGMD, Self-inhibition, Lateral inhibition, Collision detection