AI Guardsman Employs Computer Visualization To Mark Shoplifters

Have an irksome feeling that something or somebody is keeping an eye on you? Chances are it is a camera. Cities all over the world use vision-enabled computer CCTV to observe pedestrian and vehicle traffic. And in the years to come, brick-and-mortar shops may employ them to catch shoplifters.

NTT East, the Japanese telecom company, joined hands with Earth Eyes (a tech startup) to make AI Guardsman. It is an ML system that tries to catch shoplifters in the act. Depending on open source tech designed by Carnegie Mellon University, it scrutinizes cameras’ live video streams in supermarkets and convenience stores, thereby tracking each customer within the shop. When it recognized suspicious activity (for instance, when a possible thief begins looking for blind spots or starts anxiously checking their ambiance) the system transfers an alert to the smartphone of store clerk with the location and mugshot of the possible thief.

Differentiating between indecisive shoppers and thieves is not simple, although, and NTT East makes no statements that its algorithm is ideal. It sometimes marks salesclerks restocking shelves store and users who pick & put back goods as thieves, a spokesperson of the firm claimed to the media in an interview.

But the company states that AI Guardsman educates itself from its errors eventually. In the companion handset app, clerks register whether or not they caught the suspects recognized by the system, and the outcomes are employed to filter out false data.

Speaking of AI, China’s Baidu carries on to make path in the automotive sector after it signed a deal with Ford China. This deal will witness the 2 firms operate together to make the driving experience in China smarter.

The 2 firms have worked together before, most remarkably by jointly spending $150 Million into Velodyne (the LiDAR sensors startup), and this initiative by China will convey technologies such as artificial intelligence, connectivity, and digital marketing into the vehicles.

You might like