NtechLab is one of the world leaders in the field of biometric technologies, reported on the results of testing face recognition algorithms by the National Institute of Standards and Technology of the US Department of Commerce (NIST). The results of the Face Recognition Vendor Test (FRVT) contest demonstrated high results of NtechLab algorithm on the most complex photo databases.
FRVT testing 1:1 corresponds to the scenario of confirming a person’s identity by verifying with a photo. This function is used in a wide range of civil, law enforcement and national security programs, including photos examination on visa documents and for issuing passports, identifying individuals using images from the media. Within FRVT framework, more than 100 algorithms from various suppliers were evaluated. The accuracy and speed of the search, as well as the adaptability of the algorithm to subsequent changes were taken into summarizing account. The study also aims to help the US Department of Commerce to identify the world's best suppliers of such software solutions.
FRVT testing was conducted on several photo bases - Wild and Child Exp, which contain photos from social networks, as well as base of people suspected of child abuse. The leading algorithms in such tests continue to work effectively in so-called wild conditions and are used to ensure public safety. The NtechLab algorithm, on which Find Face Security's solution is based, ranked second when working with the Child Exp base and ranked third with the Wild base, having improved by doubling accuracy rates as compared with the previous test.
“A number of new companies took part in this NIST contest, many of which are from China. It can be said that the contest was highly competitive as never before. Based on its results, we can gladly assert that regular improvements of our algorithm allow it to remain among the world leaders in the accuracy of identifying individuals in a variety of conditions”, said Artem Kukharenko, founder of NtechLab, Head of the neural networks laboratory.
NtechLab algorithms have repeatedly proven their unmatched superiority in international competitions. The development of NtechLab ranked first in 2017 according to the results of the American Agency of Advanced Studies in Intelligence competition in categories of the most accurate and the fastest algorithms. NtechLab entered the top three winners of the Wider Pedestrian Challenge in 2018 on detecting pedestrians based on their silhouettes. The algorithm is successfully piloted in ten major cities of Russia and CIS countries, including Moscow, Ryazan, Almetyevsk, as well as in municipal video surveillance systems of cities in India, South America and Southeast Asia.
