In line with the new update to Google Play Services coming out, we take a look at a new API that app developers can play around with when planning out their new camera-based apps. It’s a new Face Detection API, and it’s embedded into the new Google Play Services 7.8 release. Let’s dig in.
Up front, Google is pretty adamant that it reminds us that this is NOT a facial recognition API. That can be used for security-related implementations and apps – this is not that. Instead, the Android FaceDetector.Face API detects areas in an image or video file that are human faces. From there, it can also “remember” that face in frame 1 and mark the same face in the next frames, even if it changes position. The limitation is that if a face leaves the frame and then appears again, the API can’t mark it as a previously detected face.
How does the API do that? It specifically remembers “landmarks” in a face – that is, it recognizes one of the: the bottom of mouth, left cheek, left ear, left ear tip, left eye, left mouth, base of nose, right cheek, right ear, right ear tip, right eye or right mouth. And because it is able to take note of multiple data from multiple faces, the API is then capable of detecting a face even half of the facial data is missing, just because it has a record of those facial landmarks.
So how does one implement this API in an app? Well, you can develop a camera app that automatically “clicks” the shutter when all of the detected faces are smiling. You can define a smile by giving your app landmarks to watch out for. Pretty cool eh?
SOURCE: Google
Lazy bastards – this has been there forever through OEMs. Had they put together facial recognition that sycned with your contacts and automatically gave options to tags faces locally and on other services, it would have been something.