Video security surveillance learns autonomously with very few errors

The video surveillance system developed by BRS Labs, called Behavioral Analytics, is able to perform cognitive reasoning similar to human brain and learn autonomously with fewer errors than a usual object classification and video motion based software.

The system analyzes and learns human behaviour patterns in a totally autonomous manner, including the surrounding scenes without needing any human calibration or programming. It does this by building memories of its own. The product, which is called “AI Sight Cognitive Video Analytics Software”, is currently installed by various hotels, ports, banks, airports and so on…

The software differs from a rules based video analytics software in the sense that it can identify threats that were not previously defined, by also coming up with fewer erros.

The main components of the software are:

Video Analysis Engine, which creats and analyses a visual model of the scene,
Machine Learning Engine, which is a learning technology based on human cognition,
System Manager, which analyses multiple Aisight servers and controls alerts,
Operator Console which provides operators with views and notification of alerts.
See a more detailed description of the product at BRSLABS websitehere.

Reference: net-security.org/secworld.php?id=8144