The problem with most CCTV cameras used for surveillance is that it is not 'smart'. It merely provides a recording or live feed of the surroundings, something that proves helpful when you have to carry out the post-mortem of an event or keep an eye-out for what is going on. However, a startup now wants to shake things up by making these dumb CCTVs more proactive.

New York-based Ctrl Vision enables connected surveillance cameras to "see" and "understand" their environment instantly using AIand deep learning algorithms. The startup's product, Ctrl Surveillance, specifically can detect and understand multiple detailed attributes about people, vehicles and objects seen by surveillance cameras.

"Our value proposition is to do advanced AI algorithms and run the vision software at the edge and only send important information or alerts. We have optimised vision software to run on SoC Devices that power IoT Devices enabling smarter vision," says Nemanja Bajic, co-founder of Ctrl Vision.

Ctrl solution

Ctrl Vision focuses on providing AI-based computer vision to IoT devices like surveillance cameras. Most of the 300 million plus surveillance cameras worldwide only record videos for "post-mortem" analysis and are "blind". Ctrl Vision enables these cameras to "see" and "understand" their environment (specifically people, vehicles and objects) providing real-time alerts or analytics from these camera feeds.