The patent request for "Automatic
video stream selection" by Apple Inc., announced by the U.S. Patent and
Trademark Office on Thursday, reveals a system that can wisely determine
whether to make use of the video feed from a front or rear facing camera of
Smartphone. The technology can be utilized in video calls such as those made
through Apple's FaceTime or for content locally stored on a device. Several handsets
have two cameras for picture and video taking, as noted in the document. The
camers are usually facing in opposite directions with the rear-facing module
utilized for higher-resolution shots whereas the front-facing unit manages the duties
of video call.
With enhancements in mobile
processing power, few of the smart phones are able to take video streams from
both cameras at the same time; however the utilization for such a feature has
been constrained. Currently the iPhone uses just one camera at a time,
providing users the choice to toggle between the two as preferred. It has been
noted by Apple that current bandwidth restrictions would make it hard to stream
both video feeds simultaneously, thus a system is required to intuitively
choose one while keeping audio in sync.
Apple is developing a system that
takes and monitors two video streams for voice or visual signals, then outputs
one of the two relying on a set of determined criteria. One instance would be a
FaceTime call which has used a rear-facing camera video stream before the feed robotically
shifts to the front-facing camera when a person starts to talk.
In order to select which camera to
use, the technology recognizes the speech by utilizing data from microphones
facing in the identical location as each camera or the facial expressions such
as the movements of lip. When no speech or lip movement is recognized, the
video shifts back to the scene.

















