Published on February 21st, 2017 | by Diogo Costa

Ntechlab Provides a Glimpse on the Power of Facial Recognition

Facial recognition is a technique of biometrics based on people’s facial features. It is no secret that humans can, easily and accurately, recognize other people’s faces — even if two people are alike, we are able to distinguish them. This technique can be transmitted to computers but, for this, it is necessary to define unique traits of human faces, which must be mapped in binary codes.

The use of this technique is almost endless — one of the most recent examples comes from the Schiphol Airport, in the Netherlands, where KLM is testing boarding using facial recognition. But it has several alternative applications, as the company Ntechlab clearly shows.

This deep learning face recognition company specializes in machine learning and neural networks, focusing on applying these technologies to solve real-world problems, thus being helpful for security and law enforcement authorities.

Ntechlab has a truly powerful algorithm, winner of MegaFaceBenchmark 2015, a global facial recognition contest — outperforming Google and China’s best programs in the subject. Based on this algorithm they built FindFace.pro, a SaaS facial recognition platform that any company can apply to their products via its API.

Besides, Ntechlab’s facial recognition algorithm was used to release several interesting and “playful” services. Just before the 58th quadrennial U.S. presidential election, where Donald Trump faced Hillary Clinton, they released a (now unavailable) web app for people to upload their pictures and find out if they look most like one or the other.

It was also used to build FindFace.ru and FindFace.me, services allowing to find users via their looks both on VK, the biggest social network of Eastern Europe, and Twitter, respectively. While these applications are mostly recreational, they can actually be useful.

For example, people can upload their own pictures to check if anyone else is using them. Finding missing people is also possible, as well as finding people with whom users lost contact.

Ntechlab’s research and development is showing us the power and potential that facial recognition really has. It will just be a matter of time before the Schiphol Airport’s trial becomes the widespread norm and, when that happens, Ntechlab will surely play an important role.

Tags: algorithms, deep learning, facial recognition, findface, machine learning, ntechlab