ORNL'S Deep Learning Detects Security Risks In Traffic Patterns. скачать в хорошем качестве

ORNL'S Deep Learning Detects Security Risks In Traffic Patterns.

Researchers at the Department of Energy’s Oak Ridge National Laboratory have developed a powerful deep learning algorithm that spots unusual vehicle activity before it becomes a security risk. By analyzing drone and security camera footage, the software establishes normal “patterns of life” in traffic and detects deviations that may signal illicit activity like the movement of nuclear materials. Unlike traditional vehicle recognition systems that require specific camera angles, ORNL’s model can match vehicles across vastly different perspectives, in different lighting, and when views are partially obscured. Driven by advanced neural networks, the technology is over 97% accurate at identifying whether two images show the same vehicle. This unprecedented accuracy with all kinds of vehicle makes and models provides a new layer of U.S. nuclear security. ORNL is using #BigScience to make a big impact. Learn more: https://www.ornl.gov/bigscience