Artificial Intelligence and facial recognition technologies are two major things that I have predicted would increase dramatically in the coming years with the new advanced in computers and the desire to revive the ‘surveillance state’ concept of the past but with the tools of the modern age. China and London have been two examples of places where surveillance is used to monitor the population, and now the Russian capitol of Moscow, at a time when she is dreaming of the days of Stalin and Lenin, is using the tools that both people would have only dreamed about and begged for the chance to use that is now supposedly the largest of such networks in the world.
The Moscow video analytics system has been introduced “on a mass scale” since January 1, Moscow Mayor Sergei Sobyanin announced on January 23 at an expanded meeting of the board of the Main Directorate of the Ministry of Internal Affairs for Moscow. An important element of the urban video surveillance system with face recognition is the technology for their search in the video stream. Its Moscow Department of Information Technology (DIT) acquired from Ntechlab (12.5% is controlled by Rostec, 25% is owned by Ruben Vardanyan’s fund, the rest is from five founders and a minority shareholder of the first round of investments). Information on the purchase of a video detector is on the government procurement website, the parties confirmed to Vedomosti the fact of the transaction.
The exclusive rights to the Ntechlab detector were purchased by the Moscow Electronic City company, which belonged to the mayor’s office, and it forwarded the request to DIT. The initial competitive price for development was 200 million rubles. The parties did not name the amount of the contract.
The face recognition process in video consists of two parts, explains the representative of Ntechlab: face search in the frame (detection) and its recognition directly – reconciliation with databases. The camera generates a video stream from a huge number of frames, the detector in each one searches and captures images of faces, passes them to another algorithm – it converts the image into a unique digital code and compares it with image codes from the necessary databases. For example, if the algorithm establishes a sufficient degree of similarity between two images – detected and from the search base, a law enforcement officer receives a notification where to look for a possible offender, says a DIT representative. The detector identifies depersonized images, and the identity of citizens is established only by law enforcement officials in accordance with the law, he emphasizes.
Ntechlab developed the detector for Moscow from scratch as a separate technical solution, the company did not sell the source code for its main FindFace Security product, an Ntechlab spokesman said. The company developed the detector, taking into account the experience of launching complex high-load projects in the interests of Moscow, he continues, the product is adapted for use simultaneously on hundreds of thousands of cameras and already successfully operates in the face recognition system. In the recognition process itself – comparisons of images with databases – the city can use the algorithms of different suppliers. (source)