1st place
in the Megaface challenge

Augmenting Intelligence

NTechLab develops and implements artificial intelligence algorithms. We are a team of experts in machine and deep learning, focused on building software that makes the world a safer and more comfortable place.

FaceN project allows to solve any tasks related to face recognition.
Product & Technology Findface – first project using the FaceN algorithm
About

NTechLab was founded in 2015 by Artem Kuharenko to create algorithms as intelligent as humans and as efficient as machines. Our team uses the most advanced techniques in the field of artificial neural networks and machine learning to develop smart and innovative software products.

A face recognition system already developed by our lab, has proved to be among the most accurate ones throughout the world. As tests have shown, it surpasses a human capability to identify faces.

We aim to offer intelligent solutions to our customers, so that their challenges are solved in a fundamentally new and efficient way.

Quote
In recent years artificial neural networks (ANN) have emereged as a mature and viable framework with many applications in various areas.
R. M. Hristev,
"Artifical Neural Networks"
FaceN
Face Recognition System
FaceN allows to solve any tasks in face recognition. Due to the specific technology based on the artificial neural network, the algorithm can extract the characteristic features of a person’s face from a picture. The system is already applied to identify people by the photographs and used in access control.
1st place in the Megaface challenge
NTechLab won the world challenge in face recognition The MegaFace Benchmark organized by the University of Washington that took place in November 2015. 100 teams during the 20 days of competition, using its unique recognizing algorithms, should have developed a way to recognize the largest number of people from 1 million of photos.
The competition represented world`s leading digital labs, employees of Google and several Russian teams. FaceN algorithm, created by Ntechlab, outperformed all the competitors and took №1 position.
FaceN
Algorithm
To identify a person by his photo correctly, the algorithm must be able to extract the inherent facial features. These features are to be distinguished from the ones that can vary depending on the different temporal factors (such as age, facial expression, make-up, glasses) and the particular camerawork (angle, lights, background). The unchanging characteristic features are called invariant, and due to them it is possible to tell people apart.

Produced by the training of a deep learning model, our algorithm can retrieve invariant features specific to the individual face structure. People can notice only some of them (eye size, eyebrow structure, lip shape etc.), while others are not perceptible for the human brain. Therefore the algorithm is capable of recognizing faces in a more precise and advanced way than people are.
FaceN
People Search by photographs
Provided with just one picture, FaceN can identify a person in a split second and find his photographs among billions of other portraits in the database. The system can be used at planetary scale due to the efficient search algorithm and to the specially developed technology that allows to store the extracted features in a compact way.
Contacts
Moscow, 123056,
Bol'shaya Gruzinskaya street, 61-2

Tel.:
+7 (499) 110-22-54

E-mail:
info@ntechlab.ru
Contact form