Neural Processing and Smart Cameras

Neural Processing and Smart Cameras

April 09, 2019, 8:30 AM - 10:00 AM

Sands 302

Deep learning is revolutionizing visual intelligence for physical security. Modern cameras find and identify faces and vehicles, analyze behavior and organize and control assets. Deep learning capable hardware is evolving at a frantic pace, GPU and NPU (neural processing unit) co-processors are commonly embedded in cameras and video management system. As artificial intelligence penetrates all levels of visual data acquisition, analysis and management, it’s essential that security professionals at all levels understand the capabilities and challenges inherent in this technology and that they prepare for the rapid changes in hardware assets that will be required to support it. This presentation reviews the history of artificial neural processing, analyzes examples of its use in current products from a variety of vendors. Determine how neural processing may be deployed in on-camera and VMS-based assets and how emerging GPU and NPU platforms can be used to support scale-able application environments. Lastly, consider information security and privacy/legal issues arising in the application of this technology.

Learning Objectives:
1. Define "deep learning" and describe its application in smart cameras for physical security.
2. Describe the computational hardware that enables artificial intelligence in security camera networks and analyze its impact on system specification and deployment.
3. Discuss policy concerns arising from deployment of smart camera networks.


  • David Brady



    Duke University

    David J. Brady is the Fitzpatrick Family Professor of Photonics at Duke University. Professor Brady graduated from Macalester College in Physics and...


  1. Track
    End-User Physical Security Track

We use cookies to operate this website and to improve its usability. Full details of what cookies are, why we use them and how you can manage them can be found by reading our Privacy & Cookies page. Please note that by using this site you are consenting to the use of cookies.