It’s no secret that Facebook is huge on each AR and VR, each for leisure and conversation functions. And new analysis suggests it’s running on AR programs that would no longer simply alter or substitute your face, however your whole frame.
A weblog publish lately has the AI Digicam Staff appearing off a bit of of labor obviously aimed at doing full-body substitute or tracking in a VR or AR context.
“We recently developed a new technology that can accurately detect body poses and segment a person from their background. Our model is still in research phase at the moment, but it is only a few megabytes, and can run on smart phones in real time,” the researchers write.
In fact, this sort of analysis is a long way from extraordinary; skeletal tracking methods are common in lots of industries. And certainly this weblog publish is extra about how this actual gadget and its part neural networks perform than claiming any main advances.
That stated, Facebook is obviously having a look particularly into holding issues environment friendly and simple to deploy on cell. That implies running inside many arduous barriers so far as sensor information, symbol solution and refresh fee, processing energy to be had, and so forth. Masks R-CNN, because the methodology is known as, is a pleasant step on this route.
We will undoubtedly be expecting extra like this from Facebook one day — the corporate is even on the lookout for a analysis intern devoted to this line of study.
Featured Symbol: Facebook