Wednesday, September 23, 2009

"Video surveillance system that reasons like a human brain" (sic)


The Next Generation of Video Analytics Software

The AISight™ Cognitive Video Analytics™ software takes visual input from either a live camera or recorded video, learns what activities and behaviors are normal, and generates real-time alerts that notify appropriate security personnel of potentially threatening behaviors. Unlike rules-based video-analytics software, the intelligent AISight solution identifies threats and behaviors that were not previously defined or anticipated, and it does so with fewer false positives. In test after test, AISight has achieved a false-positive rate and overall accuracy previously unseen in the video-analytics industry.

If you dig a little bit more, you'll find these precise explanations on how the program works:

Cognitive Video Analytics Learns Just Like You and Me

As mentioned, the use of vision analytics algorithms in the area of video surveillance is not new and using the concepts around the cognitive sciences (the study of how the human brain functions) in the development of technology has been applied in many different applications and industries. However, creating an interconnection between vision analytics and a system that emulates the cognitive process, utilizing various machine intelligence and machine-learning technologies, represents a breakthrough for the video surveillance industry. A cognitive-based video analytics system is not only equipped with the ability to “see” better but also to “learn,” “remember,” and “make observations” much like a human brain. A cognitive-based video analytics system constructs its own understanding of the world it is observing by evaluating the patterns of activity for any given environment over time. A "mental model" is then created for each scene to make sense of observed activities. Learning is achieved by continuing to adjust the mental models to interpret and alert on new activities as they occur, within the context of previous activities. Thus, a cognitive-based system creates an understanding of what is seen through a camera’s field of view and establishes what it determines to be normal for any given environment. It is therefore able to alert on activity it determines to be abnormal.

Question: what will happen to the program if you give it the following videos as input? Will it start "laughing" and die (crash)?

Posted via email from Past news & new news, thus updated often

No comments:

Post a Comment