RUS | ENG

Supported by:

Russian face recognition app now available to Twitter users and Turkish police

Better than Google

Dec 22, 2016
Now Twitter users can find a random person’s page in the network after taking a photo of them in the street. This functionality — a sort of Shazam for people — is offered by FindFace, an app that can scan Twitter’s database of 313 million active users in less than a second and find a match, if it exists.

The app was made available to Twitter users yesterday. Its developers integrated it through Twitter’s public API.

NTechLab, the Russian startup behind FindFace, says that its app will "help Twitter users protect their identities, find long-lost friends and relatives, and identify new potential connections."

"FindFace will help Twitter users ensure their photos aren’t being used for fake accounts," NTechLab co-founder Alexander Kabakov explained. "On top of that, users can find people they knew previously, from old relatives to old relationships. When an exact match isn’t found, FindFace will identify the most similar matches instead."

In Russia, where it is integrated with the leading local social network Vkontakte (VK), FindFace has already been used for both positive and negative purposes. While it helped identify pyromaniacs in St. Petersburg, it has also been used to identify and harass young women, as reported earlier this year by East-West Digital News.

Complaints have been few so far, however, and there haven’t been any lawsuits, the NTechLab press service told EWDN.

"We continue to monitor the usage of FindFace and block suspicious activities. In case of complaints, our customer support team treats each of them individually."

Better than Google

The FindFace technology is built on deep learning and a neural network-based architecture. In November 2015, NTechLab won the MegaFace Benchmark, a world championship in face recognition organized by the University of Washington.

The challenge was to recognize the largest number of people in a database of more than a million photos. With a recognition accuracy rate of 73.3%, the Russian startup bypassed more than 100 competitors, including Google with its program FaceNet.

In the fast-growing market of facial recognition software, NTechLab has begun applying its solution far beyond social networks.

"We have a pipeline of clients from different spheres: banking, casinos, security providers. For instance, NTechLab has a strategic agreement with TIAR Technology, a system integrator that works with biometric solution provider Papilon Savunma. This company is a market leader in a range of Middle-Eastern and Asian countries, including Turkey where it serves the police, the military, intelligence services, ministries and large corporations," NTechLab told EWDN.

In Russia, N-Tech Lab’s face recognition solution has been adopted by 1C Bitrix, the local enterprise software giant.