Unsupervised Adaptive Multi-Object Tracking-by-Clustering Algorithm With a Bio-Inspired System

Julio Guillen-Garcia,Cristina Conde,Enrique Cabello,Daniel Palacios-Alonso

doi:10.1109/access.2022.3154895

Julio Guillen-Garcia, Cristina Conde + Show 2 more

Open Access

https://doi.org/10.1109/access.2022.3154895

Copy DOI

Abstract

A problem of current interest is how to emulate nature by acquiring information in a neuromorphic-like fashion; namely, by using configurable hardware and electronic systems to emulate the information gathering and processing strategies of biological systems. In this paper, we introduce BioCAMSHIFT, an algorithm for a bio-inspired system that acquires information via a neuromorphic process and uses it to track multiple objects. The system consists of a silicon retina that simulates the behavior of the human eye together with a communication system that uses an Address-Event Representation protocol to transmit information in a way analogous to that of biological neural systems. An unsupervised procedure, based on the CAMSHIFT algorithm, is then used for multi-object tracking. It takes advantage of the retina’s high event rate to adapt to the changing sizes of the objects in its field of view. The proposed system has been experimentally validated using a data set from Freeway 210 in Pasadena, California, demonstrating a significantly better improvement in terms of multi-vehicle detection and tracking performance over the current state of the art.

Highlights

Neuromorphic systems [1], [2] attempt to emulate very specific biological functions, usually of a sensory type, whose structure and functionality have been analyzed in great detail
Their aim is to produce bio-inspired systems by using configurable hardware and electronic systems to emulate the ways of acting, information processing, and problemresolution strategies of biological systems
To validate the performance of BioCAMSHIFT, multiple tests were carried out in different scenarios, in order to test the various aspects of the algorithm

Summary

Introduction

Neuromorphic systems [1], [2] attempt to emulate very specific biological functions, usually of a sensory type, whose structure and functionality have been analyzed in great detail Their aim is to produce bio-inspired systems by using configurable hardware and electronic systems to emulate the ways of acting, information processing, and problemresolution strategies of biological systems. The typical approaches found in the literature [3] use computers to extract information from images of the physical world Such images are usually presented as functions associating to every point in the image, a value relative to some property of the pixel or voxel it represents (e.g., brightness, hue, intensity, etc.). One may say that current AV approaches attempt to “fit the problem to the tool”, a tool that operates in a way that differs vastly from the biological principles AV strives to emulate

Methods

Results

Conclusion