Система биометрической идентификации лица человека на основе модели anti-spoofing

Аннотация

Biometric identification is often referred to as pure or real authentication, since it uses not a virtual, but a biometric attribute that is actually relevant to a person. Passwords can be stolen, spied on, forgotten, keys can be forged. But the unique characteristics of the person himself are much more difficult to fake and lose. This can be fingerprints, voice, drawing of blood vessels of the retina, gait, etc. Face recognition looks like a very promising direction for use in the mobile sector. If everyone includes long been used to using fingerprints, and technologies for working with voice are gradually and rather predictably developing, then with face identification the situation is rather unusual and worthy of a small excursion into the history of the issue. Keywords: recognition systems, technology of biometric identification, person’s face, the anti-spoofing model, authentication 289 Рақамли тeхнологиялар соҳасидаги ҳуқуқбузарликларга қарши курашиш ҳамда ахборот хавфсизлигини таъминлашнинг ташкилий-ҳуқуқий масалалари Today’s detection systems show huge accuracy. With the advent of large datasets and complex architectures, it has become possible to achieve face recognition accuracy of up to 0.000001 (one error in a million!), And they are already suitable for portability to mobile platforms. Their vulnerability became the bottleneck. In order to personify another person in our technical reality, and not in the film, masks are most often used. They also attempt to fool the computer system by presenting someone else instead of their face. Masks come in very different qualities, from a printed photo of another person held in front of their face to highly complex 3D heated masks. Masks can either be presented separately in the form of a sheet or screen, or worn on the head. The availability of such vulnerabilities is really hazardous for banking or government systems of user authentication by face, where an intruder’s penetration entails significant losses. We can to call an attempt to deceive the identification system by presenting it with a fake biometric parameter a spoofing attack. And, we will call the complex of protective measures to resist such deception anti-spoofing. It can be implemented in the form of a variety of technologies and algorithms built into the pipeline of the identification system. The ISO offers a somewhat extended set of terminology, with terms such as presentation attack – attempts to force the system to incorrectly identify a user or to allow him to avoid identification by displaying a picture, recorded video, and so on. Normal (Bona Fide) – corresponds to the normal algorithm of the system, that is, everything that is NOT an attack. Presentation attack instrument means a means of performing an attack, for example, an artificially manufactured part of the body. And finally, Presentation attack detection – automated means of detecting such attacks. However, the standards themselves are still in development, so it is impossible to talk about any well-established concepts. Terminology in Russian is almost completely absent. To determine the quality of the system, the HTER metric (Half-Total Error Rate) is often used, which is calculated as the sum of the coefficients of erroneously allowed identifications (FAR-False Acceptance Rate) and erroneously prohibited identifications (FRR – False Rejection Rate) divided by in half. HTER = (FAR + FRR) / 2. It is worth saying that in biometric systems, the most attention is usually paid to FAR, in order to do everything possible to prevent an attacker from entering the system. And they are making good progress in this (remember one millionth from the beginning of the article?) The flip side is the inevitable increase in FRR – the number of ordinary users mistakenly classified as intruders. If this can be sacrificed for government, defense and other similar systems, then mobile technologies, working with their huge scale, variety of subscriber devices and, in general, user-perspective oriented, are very sensitive to any factors that can force users to refuse services. If you want to reduce the number of phones smashed against the wall after the tenth consecutive denial of identification, the FRR is worth looking into! 290 Рақамли тeхнологиялар соҳасидаги ҳуқуқбузарликларга қарши курашиш ҳамда ахборот хавфсизлигини таъминлашнинг ташкилий-ҳуқуқий масалалари The most popular means of deception are masks. Nothing is more obvious than putting on another person’s mask and presenting the face to an identification system (often referred to as a Mask attack). You can also print a photo of yourself or someone else on a sheet of paper and bring it to the camera (let’s agree to call this type of attack Printed attack). Slightly more complex is the Replay attack, when the system is presented with the screen of another device, on which a previously recorded video with another person is played. The complexity of execution is compensated by the high effici