Abstract:Thispaperdescribes anewwayofthevisual This paper introduces anewconcept ofvision system information processing forsmallintelligent aerial based ontheneural networks andtheapplication of vehicles usingneuralnetworks, and itsVLSI spike-based neural hardware implementation bythenew implementation. Thevision using neural networks isfor bio-inspired conductance withtheadvantages ofup-to-date thenavigation toavoidobstacles, basedonthe64x64 VLSItechnology. image. Theareainthefront offlying ismappedto sixteen sectors, andanappropriate pathisselected by II.NEURAL NETWORKS FOR THE VISUAL thetrained neural networks. Themultilayered network INFORAMTAIONBASEDFLYING withtwohidden layers isemployed, andtheempirical rules areaddedtothetraining process. Pre-processing Thevisual information processing forthenavigation is ofsectored imagedataisprovedtoimprovethe widelyinvestigated in many area,especially for cognition performance remarkably, by itslocaliseduninhabited aerial vehicle orflying robots. Oneofissues is attention. Thesize ofneural networks isminimised for toavoidobstacles, whicharelikely a problem for asingle VLSIdevice implementation, lessthan4,000 autonomous flying incomplex environments. Thevision synaptic connections. based ontheinformation ofcolour contrast wasinspired by Theproposed vision system demonstrates thedesired thebiological vision system, andhasbeenapplied tothe behaviour withrealenvironmental scenes after the flying control. Theflying robot withneuromorphic eyes, training by thesimulated data.The feasibility of mimicking theinsect eye,demonstrated thefeasibility of neuromorphic VLSIimplementation isdemonstratedflowsensing toavoid obstacles andfollow terrains. Forthe bytheneural networkwithreduced synaptic weight navigation intheconfined 3-dimensional space, thevision accuracy, whichrepresents themeasured accuracy of based onsuchprinciple wasapplied tocontrol theflying biologically plausible conductance basedCMOS neural blimp. Theobstacles orflying environments like ground or networkVLSI.The proposedneuralnetworks wallneeds tobepredefined colour contrast, suchasblack implementation demonstrates thefeasibility ofan andwhite strips. (1)(2)Suchcolour contrast information is intelligent vision systemofsmalluninhabited aerial considered helpful forahumaninthree-dimensional

An insect vision-inspired neuromorphic vision systems in low-light obstacle avoidance for intelligent vehicles

A Visual Neural Network for Robust Collision Perception in Vehicle Driving Scenarios

Synthetic Neural Vision System Design for Motion Pattern Recognition in Dynamic Robot Scenes

Collision Selective Visual Neural Network Inspired by LGMD2 Neurons in Juvenile Locusts

Event-Based Neuromorphic Vision for Autonomous Driving: A Paradigm Shift for Bio-Inspired Visual Sensing and Perception.

Robustness of Bio-Inspired Visual Systems for Collision Prediction in Critical Robot Traffic

A neural model of the locust visual system for detection of object approaches with real-world scenes

A Biomimetic Visual Detection Model: Event-Driven LGMDs Implemented With Fractional Spiking Neuron Circuits

Neuromorphic Visual Odometry System for Intelligent Vehicle Application with Bio-inspired Vision Sensor.

Mixed Event-Frame Vision System for Daytime Preceding Vehicle Taillight Signal Measurement Using Event-Based Neuromorphic Vision Sensor

Ground Moving Vehicle Detection and Movement Tracking Based on the Neuromorphic Vision Sensor

Neuromorphic Vision Based Multivehicle Detection and Tracking for Intelligent Transportation System

NeuroIV: Neuromorphic Vision Meets Intelligent Vehicle Towards Safe Driving With a New Database and Baseline Evaluations

Embedded and Real-Time Vehicle Detection System for Challenging On-Road Scenes

LGMD and DSNs Neural Networks Integration for Collision Predication.

Neuralnetworks Implementation Ofthevisual Information Processing Foranintelligent Aerial Vehicle

Temperature-based Collision Detection in Extreme Low Light Condition with Bio-inspired LGMD Neural Network

Parameter Optimization and Learning in a Spiking Neural Network for UAV Obstacle Avoidance targeting Neuromorphic Processors

O-LGMD: An Opponent Colour LGMD-Based Model for Collision Detection with Thermal Images at Night.

Looming Detection in Complex Dynamic Visual Scenes by Interneuronal Coordination of Motion and Feature Pathways

Insect-Inspired Elementary Motion Detection Embracing Resistive Memory and Spiking Neural Networks