The Kepler Vision Engine sees beyond ordinary visuals. It's cloud based software that processes any HD image and understands body language and actions. On display at CES 2019, the Kepler Engine can use that information to process if a person is defensive, aggressive, relaxed, angry or stressed. Better yet, because it understands body language, it can also watch out for signs that a person might be in trouble.
Kepler's not actually a smart cam in and of itself. It's not an actual piece of hardware, just a concept for now and one aimed at the smart cams of the future. Specifically, Kepler Vision, the startup from Amsterdam behind the engine, wants its tech used to improve eldercare.
The screenshots of the tech in action certainly look cool, and supposedly it can use any HD image so it should work with a wide variety of smart cams. It processes the skeletal information of any person in the image in addition to the usual visual info. It understands faces so it can tell a couple apart, but it's meant more to understand what a person is doing.
The engine will process the visual data in the cloud, and in addition to reading basic signs like arms crossed in a defensive posture, Kepler will understand if a person is distressed, if a person is eating, drinking or smoking, and whether a person is sitting, standing, lying down, bending over or walking. It's supposedly even smart enough to tell if a person is lying on the floor versus the couch.
Representatives from the company told me that the software will be ready to go in two months, so while theare already pretty smart, I look forward to seeing what they can do in the near future with the Kepler Engine.
CES 2019: Every story so far: See all of CNET's coverage of the year's biggest tech show.
: It's six days of jam-packed events. Here's what to expect.