Neuromorphic engineering : taking AI to the edge

Tools

Kirkland, Paul and Di Caterina, Gaetano and Soraghan, John and Thomas, Kevin and Matich, George (2020) Neuromorphic engineering : taking AI to the edge. Polaris Innovation Journal.

Full text not available in this repository.Request a copy

Abstract

Technology is an emerging field that presents many practical applications and benefits. While introducing the concepts of the Sensing and Processing, with sparse asynchronous encoding on the sensor side, coupled with asynchronous processing, able to return up to two orders of magnitude computational reduction, this article presents a Neuromorphic approach for the challenging problem of UAV detection and tracking. The small cross section of a paired with the expansive search space, highlight the key advantages of an approach. Together with the ability to deliver a significantly higher temporal resolution without the computational overhead, makes the feasibility of a low profile with microsecond accurate tracking updates within the electrico-optical domain attainable. In this context, the Dynamic Vision Sensor is used to detect a within a scene and return the location of the target, achieving a 91% detection rate while only utilising 4% of the sensor.

ORCID iDs

Kirkland, Paul ORCID logoORCID: https://orcid.org/0000-0001-5905-6816, Di Caterina, Gaetano ORCID logoORCID: https://orcid.org/0000-0002-7256-0897, Soraghan, John ORCID logoORCID: https://orcid.org/0000-0003-4418-7391, Thomas, Kevin and Matich, George;
CORE (COnnecting REpositories)