Key Features:

  • Real-time face detection, recognition and
    tracking with HD cameras
  • High-granularity students or employees
    attendance tracking
  • Location reporting and match event history
  • Unknown personnel and weapons alerts with 3rd
    party integrations
  • Access automatization with biometric
    authentication
  • [In research] Guns or concealed weapon
    detection in IR spectrum

Next generation of video analysis software based on a combination of Artificial Intelligence techniques:

  • Deep Neural Networks
  • Machine Learning
  • Computer Vision
  • Advanced Image Processing

Architecture, challenges and solutions:

From a computer vision perspective, the main
challenge was building a system sensitive enough
to identify known individuals from surveillance
cameras, while also correctly flagging unknown
individuals.
Resource-wise, the challenge was to process
the video streams as efficiently as possible. We
developed a variety of heuristics and optimizations
in order to minimize the amount of expensive
computations.