We just circle our fingers: it was proved that when recognizing a person’s face, we are mistaken about 10 times more often than cars. Studies were conducted at the University of Massachusetts based on VisionLabs algorithms.
VisionLabs is a platform that allows you to recognize faces with high accuracy. The main drawback of modern face recognition technologies is the deterioration of the quality of work in low light and changing the position of the head or angle.
Therefore, most of these technologies are not precise enough to incorporate them into the business. But the VisionLabs algorithm was recognized by the Massachusetts University as one of the best among the existing ones.
More information about its platform and its capabilities representatives of the company will tell on AI Conference. In the meantime, they presented three real cases with solutions already implemented.
VisionLabs: What’s so special about the platform and how does it work?
The VisionLabs LUNA platform analyzes not a picture, but a set of characteristics derived from it. First, the technology detects the person in the frame and starts the tracking algorithm in the video stream – the program determines which of the 25 frames per second captured the person in the best quality and perspective. The portrait, cleared from the background and rotated to the desired position, is sent to the recognition service. From the standard JPEG format, it is converted into a descriptor – a set of immutable face parameters, which is used for later comparison of the picture with another image. At the same time, such factors as the level of illumination of the room, the age-related changes in a person, the hairstyle and make-up, the presence or absence of a beard and a mustache are eliminated.
Then the program compares the two descriptors and gives the answer, whether the person who was in the frame is entered in the database. The similarity is determined in percentages: for example, the system can produce the result of the coincidence of descriptors by 65 or 99%.
Banking: the amount of prevented fraud is estimated at hundreds of millions of rubles
The task: to prevent possible fraud on the part of employees in the “Post Bank”
The most common threat for banks is the leakage of personal data and their falsification, which leads to both financial losses due to internal fraud, and to serious reputational risks and, as a result, a reduction in the value of shares.
According to the experience of VisionLabs, up to 90% of fraud in banks is committed by employees – alone or in collusion with external fraudsters. Sectors in which fraud is most evident is consumer lending, ATMs and all remote banking services. In “Post Bank” this problem was solved by authorization when accessing personal data in an electronic database – using face recognition technology.
Implementation process and results
The LUNA Identification and Verification Platform daily processes several hundred thousand photographs.
It compares the biometric parameters of new customers of the bank with the parameters of existing clients in its database, and also compares them with the database of scammers.
The platform was implemented for 50,000 jobs in the bank and points of sale of partners. The cameras used in the workplaces of agents are not subject to special requirements. The bank assures that the quality of the image of almost any camera is sufficient for effective customer recognition.
The economic effect of the implementation of the system is estimated by the volume of prevented fraud: taking into account the growth dynamics of the retail network and the client base, the bank estimates it to hundreds of millions of rubles. At the same time, the number of attempts to commit fraudulent transactions decreased, because potential fraudsters already knew about the face recognition system.
In addition, face recognition allowed each transaction of employees in the bank to be personalized. As a result, it became easier to work with clients and the level of information security increased. The photo of the logged in user is saved, and the employee of the bank or merchant-partner can not access the information of the clients. The labor discipline of employees has increased: the situation with the transfer of their data to other employees is completely excluded. Also, the bank has a tool for objective and accurate accounting of employees’ working hours.
The task: to create the possibility of money transfers through the client’s photo in the bank “Otkrytie”
Traditionally, to transfer money to the client, you need to specify the beneficiary’s card number or his phone number if he is served in the same bank.
Bank Otkrytie was the first bank in the world to launch money transfers through a customer’s photo in December 2017.
Implementation process and results The
service is implemented using a face recognition system that allows you to identify with a high degree of accuracy the customer by his biometric data.
In the first quarter of 2017, the bank has already introduced a system for client authentication in three Moscow branches, which simplifies their maintenance and shortens waiting time in the queue. Already at the beginning of the second quarter the bank used VisionLabs developments in the mobile application of the bank for the prototype of the face authentication solution for iOS. The solution was tested by the working group of the bank, and in May 2017 it was presented at the international exhibition Connect: ID.
You can translate a photo from a map of any Russian bank to the customers of Otkrytie Bank, which were photographed in the offices or when the card was delivered by courier. Next year, “Otkrytie” plans to launch translations for photographs and for users who are not customers of the bank – users of the “Opening. Translations will be able to upload their own photos directly through the application.
Education: students become more responsible for learning
The task: to transfer the process of taking exams to the online format at the Moscow Institute of Psychoanalysis
The Institute developed a training portal for students, teachers and administration, but full-fledged work with it was impossible, since a large number of students tried to use third parties to pass the exams. This could not be controlled in any way, because the system used a standard password access mechanism.
Implementation process and results
At the moment, more than 5,000 students access the course materials every week with biometric identification. Representatives of the institute not only identify students, but also analyze their activity when working with materials. About 5% of students try to use third parties to take exams, but the system prevents all cases of fraud.
The institute confirms that biometric identification contributes to improving the quality of education for students who no longer rely on illegal methods of passing tests and exams, but are more responsible for preparation. Also, students began to more consciously and responsibly approach the work with materials, as they understand that their activity is fixed.
The Institute conducted a questionnaire before and after the introduction of biometric identification. The survey showed that initially students reacted to such an introduction with distrust, because they were worried about the increased control over their activities on the training portal – but after the introduction the institute received feedback from students who believe that they are now more consciously approaching their education and working with materials.