Archive News

July, 2019

FOR IMMEDIATE RELEASE

ARCTURUS DEMONSTRATES EDGE-BASED AI AND VISION USING NXP i.MX 8M MINI

Demonstration Illustrates Capability of Edge Detection for Public Safety Applications

TORONTO – JUNE 12, 2019 – Arcturus Networks Inc. demonstrated today at NXP Connects in Santa Clara the application of edge-based AI and machine vision as a method of detecting public safety events. The demo makes use of NXP’s i.MX 8M Mini applications processor combined with Arcturus AI and machine vision detection software. Together, this forms a cost-effect, embedded-edge solution that transforms video surveillance from passive observation to active detection. The demonstration analyzes video footage of a subway platform, providing analytics including how crowded the platform is, how close to the edge of the platform people are standing and if any abandoned packages or luggage are present.

The demonstration was developed by utilizing the NXP eIQ machine learning software development environment to enable AI and vision processing on the i.MX 8M Mini CPU. This tool supported the quick migration of existing Arcturus software to the i.MX 8M environment and supported the integration of Arm®NN, achieving a 2x performance improvement over existing detection methods. All AI and vision processing was performed in realtime, directly on the quad Cortex®-A53 processor without the use of a GPU or dedicated hardware ML acceleration. A video of the demonstration can be viewed at the NXP website, along with additional information on the NXP eIQ machine learning software development environment.

Pricing and Availability

To request a demo or for more information contact Arcturus.

More Information
Additional videos can be viewed at the Arcturus YouTube Channel.
Preliminary information on the vision product offering can be obtained from the product landing page. Vision Middleware product landing page.
For more information or to set up a demo, contact Arcturus.

Press Contacts:
Arcturus Networks Inc.
media@arcturusnetworks.com
416.621.0125 x233


# # #