Сценарії JavaScript вимкнено для Вашого браузера. Деякі функції цього сайту не будуть працювати без них.
| dc.contributor.author | Senkivskyy, Arsen
|
|
| dc.date.accessioned | 2024-02-19T10:42:15Z | |
| dc.date.available | 2024-02-19T10:42:15Z | |
| dc.date.issued | 2019 | |
| dc.identifier.citation | Senkivskyy, Arsen. Anti-spoofing system for facial recognition / Arsen Senkivskyy; Supervisor: Oles Dobosevych; Ukrainian Catholic University, Department of Computer Sciences. – Lviv: 2019. – 32 p. | uk |
| dc.identifier.uri | https://er.ucu.edu.ua/handle/1/4561 | |
| dc.language.iso | en | uk |
| dc.title | Anti-spoofing system for facial recognition | uk |
| dc.type | Preprint | uk |
| dc.status | Публікується вперше | uk |
| dc.description.abstracten | The biometric recognition systems had massive success in recent years. Since webcameras are incorporated in many different devices(cell phones, tablets, laptops, entrance doors in some facilities, etc.), facial recognition systems become highly popular. Hence, the more people use these systems, the more people try to trick them to get unauthorized access. There are three types of attack on the facial recognition system: picture-based attack, when an attacker is presenting a picture of another user’s face; Video-based attack where an attacker is showing a prerecorded video of another user; Maskbased attack when attacker uses a mask of authorized user in order to spoof the facial recognition system. In this work, I tackle picture-based and video-based attacks. For this reason, I develop a challenge-response system. The idea an approach is to detect where a user can do what system has challenged him to do. This way, we know that the face that is presented to the camera is alive. The user is required to watch a moving dot on the screen. The dot starts from the center of the screen and goes to the randomly chosen side of the screen, so this way user cannot present a prerecorded video. As the user follows the dot, the system estimates the direction where the user’s eyes are moving. For these purposes, I implemented three different approaches. The custom neural network that takes as an input projections of three consecutive frames of an eye movement and classifies which the direction of the movement. In the third approach, I hypothesized then when the user is watching at collinear points on a vertical line, the x coordinates of the user’s pupil will be approximately the same, having small variance. The same applies to y coordinates on a horizontal line. Thus by analyzing the variance of the coordinates, we can detect whether an attacker is not presenting some else’s picture. | uk |
Files in this item
This item appears in the following Collection(s)
-
2019 [23]
