Apple has applied for a patent (number 20190042833) for “face detection, pose estimation, and distance from a camera estimation using a single network.” It involves improving FaceID in upcoming iPhones, iPads, and — let’s hope — Macs.
In the patent filing, Apple notes that facial recognition processes may generally be used to identify individuals in an image. For example, it may be used to detect faces in an image for use in the facial recognition process. Head pose is often estimated in addition to face detection.
The traditional approach for achieving both face detection and pose estimation is to identify (detect) a face and put a boundary box around the face and, after the boundary box is identified, operate on the face in the boundary box to determine a pose estimation. However, Apple thinks it can improve the process.
Here’s the summary of the patent filing: “A single network encodes and decodes an image captured using a camera on a device. The single network detects if a face is in the image. If a face is detected in the image, the single network determines properties of the face in the image and outputs the properties along with the face detection output.
“Properties of the face may be determined by sharing the task for face detection. Properties of the face that are output along with the face detection output include the location of the face, the pose of the face, and/or the distance of the face from the camera.”
Of course, Apple files for — and is granted — lots of patents by the U.S. Patent & Trademark Office. Many are for inventions that never see the light of day. However, you never can tell which ones will materialize in a real product.
