Apple has just obtained a new patent concerning a method of detecting faces in a digital video stream, thanks to the use of information related to depth. the United States Patent and Trademark Office, the document describes how detection algorithms can identify the presence of faces in live video when people are located at different distances from the camera.
As noted AppleInsider, the depth mapping system described in the patent is based on motion tracking technology developed by the Israeli firm PrimeSense, bought by Apple in 2013.
Although the system is able to recognize faces in general, it does not have the capacity to identify individual differences between them. It is therefore not a facial recognition solution per se, but rather a crucial step towards a wider authentication system.
As a reminder, Apple is said to be developing a “revolutionary” camera system on the next iPhone 8. The technology would consist of three modules allowing the smartphone to offer 3D detection functionality. Although there is no way to know for sure if this patent describes one of these modules, the camera system is well powered by PrimeSense algorithms.