In the world of today, facial recognition is quite a common feature and is being used by almost one and all. From banks to visitors at home, start-ups to large organizations, this technology is now also being used to gain access to mobile devices and social media platforms as well.
The basic proposition of this technology is very simple; the user’s facial features are recorded in a database, and every time a user accesses his/her account, the image is checked with that in the records. If both the images match, the user has the permission to log in.
Enhancing Facial Recognition with 3-D Technology
However, most institutions use 2-D images of the users and this is where some serious problems emerge. For example, a cybercriminal can take a photo of the user furtively, scan it and later present it at the banks for fake transactions. Modern face recognition technologies use 3-D images of users. This means that these pictures carry depth and are very difficult to counterfeit. The efficacy of a 3-D technology is increased by skin and iris scans.
Adding layers to 3-D Technology
3-D face technologies have various layers of protection and are very complex. However, even 3-D face tech runs into some challenges. For starters, scans though complex, are not 100 percent perfect. Therefore to make such scans secure, organizations like Biometrid have an extra layer of security – The humble password.
The Twin Syndrome
There is an even bigger issue that facial recognition technologists face – the problem of twin brothers and sisters. Faces are similar, and unless the scans are perfect, there will still be a significant risk of identity theft. To solve this identity problem, engineers at Biometrid have come out with a unique solution. At the time of logging in, the user will be asked whether he/she has a twin. If the answer is yes, then the other twin will be required to submit his/her password, or another form of identity. This solution will be the order of the day until scans are perfected to differentiate twins.
To make face recognition more secure, the Biometrid platform uses nodes. These are complex numbers and represent the 3-D pictures. The pictures are never stored per se. In order to unlock these nodes, Master Keys are required. And what are these Master Keys? They’re nothing but the users face. When the correct face is presented, these nodes unravel just like the pieces of a puzzle.
Decentralizing Facial Recognition
However, what makes this entire process decentralized is that the Biometrid community runs them. The process of unraveling these nodes ultimately is determined by the various users of this platform. It should also be noted that in order that this method of identification to be successful, the participants should be incentivized. The Biometrid platform indeed rewards all those who participate in this identity verification process.
From a technical perspective, every time a facial image is registered in a computer, it is stored as a reference image. Every time a user logs in, his face is matched with that in the records. The system then gives out a score between 0 and 1. If the score is closer to 0, then access is delayed, else the user is allowed access.
To summarize, facial recognition technologies are advancing every passing day, with companies like Biometrids offering innovative solutions that are decentralized as well as anonymous.
Images courtesy of Biometrids