New Apple patents - advanced face recognition technology and intelligent call hold
Makradar Technologies / / December 19, 2019
Apple Inc. has long had a reputation as an innovative company and to date there is no reason to stop there. Recently it became known about new patents relating to the face detection and intelligent call containment systems.
The first patent relating to facial recognition technology we are talking about "face print» (faceprint), created as a result of analyzing a plurality of images and identify unique facial parts of a person. A curious fact is that in most cases in the text of the patent we are talking about well-known and or logos instead of the normal user or their friends and families.
The invention provides a set of vector analysis features of the object, which indicates that the technology can be used not only for individuals, but also for a variety of architectural structures.
The logic technology is not complicated: the user makes a photo before it is analyzed and compared with the existing base of the faceprint-s, which can be stored locally or remotely. If the system can not locate a match, it adds to the picture label, which can store basic information about the person and, for example, his page on Twitter and Facebook. To avoid false positives in the system match threshold set coefficient. If it turned out to be below the specified level, then further processing is terminated. Also, to increase the accuracy of recognition can be used system of feedback from users who can leave comments on the quality of recognition.
database visualization faceprint-s
The second patent describes a technology that is the next stage is already available in the iOS 6, wherein when an incoming call, you can immediately answer one of the text options.
The technology can be divided into two main parts: manual and automatic. With regards to manual technology is simple enough. If the user receives a call on the second line and can not answer right now, he can choose one of the answers, which are synthesized into speech and will be played in the waiting line. The user can specify a set of rules and conditions given that the system decides how to act.
The user can specify the call waiting time, and then the caller will be known about the remaining waiting time.
Also interesting feature is the ability to send a companion waiting on the line to voice mail without interrupting the current call.
A very interesting feature is that the user can select not only one of the predefined answers, but also to enter your while remaining on the line.
With specific regard to the intellectual response of the system, it is all a bit more interesting. The device analyzes the performance with GPS, accelerometer, and calendar, to determine where the user is located, be it home, work, an important meeting or a party. This feature can be useful for drivers, as is claimed in the patent the possibility of determining when a user is in the way and performing a specified action for this situation.
[via Apple Insider]