Face recognition or Facial Recognition has become more and more an issue in recent months. While some countries try to control their criminality through such applications, services like Facebook want to recognize their users faster and more effectively. But how does such a face recognition actually work?
Facial recognition starts
Before the advent of artificial intelligence, facial recognition worked in a similar way to other “biometric” services for identifying voice, iris or fingerprint. A computer analyzed an entered image, identified so-called markers and compared these specific points with aspects of comparative photos. If the program identified many similarities, the results were output.
Facial recognition functionality
Simply explained, the software replaces the image source with a version that highlights specific features relevant to face recognition. In more complex color systems, all pixels are compared to the surrounding details to discover a commonality.
However, the simple analysis programs soon reached their limits. The real breakthrough came with the advent of artificial intelligence and deep learning in particular.
A labelled machine learning data set promotes the learning of the machine and the results become more precise. Today, these exercises enable machines not only to recognize people, but also to interpret special features. In this way, you can see what gender the person in question is, how old he is or what mood he has.
Companies in Japan therefore use these applications, for example, to convince those for a special target group of special products. For example, an energy drink is more likely to be recommended to a tired person, while a soft drink is more likely to be offered to a teenager.
Facial recognition is currently a topic that is absolutely on the rise. Apart from applications in industry and marketing, it also offers a lot of potential in the smart home sector, where the intelligent devices recognise the respective occupant and issue individual offers.