For most of us, managing passwords on the Internet is a problem. real hell. And as internet services experience rapid growth and online fraud explodes, the need for strong authentication is more urgent than ever. The French start-up Unissey has developed a unique facial biometric system, based on face recognition and the detection of living things. Its founder, Yassine Mountacif, and Sébastien Brangoulo, the development director, explain their technology to us.
Pay your purchases onor open your bank from a simple look: the
Today, we already use passwords to use services on the Internet. Why change?
The password is an excellent security system but on four conditions: it must be long, complex, changed frequently and noted nowhere! Which is rarely the case, especially when you have 80 or 100as is the case today. Because they are often reused across multiple sites, passwords stolen in an attack can then be used by hackers to log into all of your other accounts. Therefore, to secure access to your account, some companies require two-step authentication systems where you receive a code over the phone or a link in your email, then return to the site, etc. It is very laborious.
What are the existing means of biometrics on the Internet?
Physics … specific to human, measurable and unique to each individual can be used as biometrics. You can even see birth certificates with footprints! But, of course, it is not very practical to reuse Also developed are the , the iris plant
Its … or facial recognition, but these technologies require In the space field , a star sensor is an optoelectronic device used for the orientation of satellites. S : a fingerprint reader, a camera for the iris or a 3D camera for for example. Our facial recognition solution, on the other hand, uses a simple 2D camera that can be found on any or . It is therefore universal.
Facial recognition, is it really safe?
Facial recognition systems are developing more and more in our daily lives, fraudsters are interested in them and try to outsmart the system by various techniques. For example, hackers grab a video or photo of you from the internet and play it back on camera. You can also print the photo on paper, create a mask or a “deep fake”, with your face modeled on a video. To make sure that you are a real person, most facial recognition systems require you to replay an exercise in front of the camera, such as turning your head left or right, following a point with your gaze, etc. However, not only can these exercises be “learned” and replayed by hackers, but they are very restrictive for the user. Often, you have to start over several times and you waste a lot of time. Conversely, our solution is called “passive”, that is to say that it does not require any effort from the user. Just stare at the camera for half a second and we’ll take care of the rest!
How do you go about making sure I’m a genuine person?
We use a
To take … heart versus blood flow to the face, or some lighting. All these criteria combined make our solution 100% safe. We are the world’s most rated facial recognition solution (Nist, iBeta, CLR Labs). Our system has been tested on thousands of attacks of all kinds, very sophisticated, and none have been able to deceive it.
In which areas facial recognition can it be used?
Any remote service that requires authentication or identity confirmation: electronic signature of documents,
The very first automobile: steam powered
The first draft of a mobile vehicle takes place in 1796 when the Frenchman Joseph Cugnot placed a … for example, the rental company will want to make sure that you are the real person appearing on the driving license, and not a usurper who is going to escape with the car! In total, the facial recognition market is growing by 18% and the potential is enormous.
100% secure internet, is it possible?
Internet security is always a compromise between maximum security and ease of use. If the process is too long or too complex, users will become discouraged and give up, resulting in a big loss of value for the services. It is therefore necessary to be both simple and safe. Of course, it’s still a cat-and-mouse game. We are updating our algorithm and we are constantly adding bricks as soon as we identify new types of attacks.
- Date of creation : 2018
- Founder : Yassine Mountacif
- Location : Paris
- Staff : 21 people (3 recruiting)