Facial recognition technology is utilized in many different systems. Biometric software is used in facial recognition tools for security purposes and other applications such as social media marketing. Algorithms use a statistical approach to identify facial features – and facial recognition is increasingly used as a crime-fighting tool. In the future it could be used to monitor employee attendance at work, to enhance security measures at ATMs and to prevent voter fraud. Many privacy advocates see a problem with this technology because it could quickly turn us into a surveillance society.
University of North Carolina researchers have discovered a way to get around facial recognition security. By using a virtual reality (VR) system to develop 3D models of the face, they were able to trick the biometric security measures. They did this with just a handful of photos found on Facebook and were able to fool the systems more than half the time (Newsweek).
Clearly this is a huge security flaw in the technology which means other types of “verifiable data” would need to be used for authentication in order for facial recognition to be a feasible option. One technique that could be used is the detection of infrared radiation which would be given off by a real face, not a 3D model (Techworm).
For more information on how facial recognition technologies work, check out this video from Brit Lab:
1 Comment
Every input can be faked.
Really secure is only an encryption already inside the human body with algorithms based on the password bits, changing passwords and 70% data transfer used to exchange tiled 3-time-pads.
Ideally, information-storages only serve so much information per time in a defined order and automatically apply a changing encryption to outgoing data.