Егор Цветков (Egor Tsvetkov), a photographer in Russia, took pictures of random people on the subway and portraits connected to social media and complete profiles using the face matching technology. This is a game changer.
It used to be that technology was good enough to tell if two images appear to be the same person. We have now reached a point where an input picture can (mostly) be used to find the person matching among tens of millions of inflection, where the processing power used is low enough that the Service to be free. This is a complete game changer.
The severity of this does not really hit you until you see the examples, where the photos are taken under lighting and angles radically different from that of portrait photos, and sometimes with different facial hair too. In addition, this photographer has used a service available corresponding photo - FindFace.ru - which already imported a large quantity of photos on vKontakte , which is (all?) the equivalent of the Russian Facebook, and leave a neural network study all these photos.
accordingly, the photographer claims a success rate of 70% corresponding to random people photographed in the subway, in completely different lighting conditions, with full profile social media.
You can not stop the mere existence of the technology. Within five years there will be CCTV cameras which lists all persons currently in phase with their name and portrait of social media. Soon after, the implementation of the law will be used for automated spotting mandate, RoboCop style -. Or at least be eager
In ten years, the service to be available for personal use -. superimposed portraits and social names of people directly to your field of vision, using glasses or projection contacts already prototype
and maybe fifteen, traders will cry and weep for the ability to provide "targeted messages" of the kind that we saw in Minority Report -. where the billboards change in front of you to meet your particular interests, or what marketing think are your interests
With the technology available as a free service, increased power treatment is just a matter of Moore's law and throwing money at it. Who will provide the neural network and photo databases? There are obviously giants like Facebook and Google, but it would also be a potential new monetization for a service which many people have sent pictures - like tinder. . Real-time image matching for monitoring use and convenience
Privacy remains your own responsibility
UPDATE: .. See also Hacker News discussion on the subject
0 Komentar