To obtain access to full text of journal and articles you must register!
- Article name
- Improving the security of the biometric face authentication procedure based on neural network converters "biometrics-code"
- Authors
- Lozhnikov P. S., , lozhnikov@gmail.com, Omsk State Technical University, Omsk, Russia
Sulavko A. E., , sulavich@mail.ru, Omsk State Technical University, Omsk, Russia
Panfilova I. E., , panfilova_2015@bk.ru, Samara State Technical University, Samara, Russia
- Keywords
- neural network biometrics-to-code converter / secure biometric authentication / face recognition / deep learning / spoofing attacks
- Year
- 2024 Issue 3 Pages 3 - 11
- Code EDN
- VAPPGY
- Code DOI
- 10.52190/2073-2600_2024_3_3
- Abstract
- The concept of secure biometric authentication by face based on neural network biometric-to-code converters (NNBC) is proposed, providing resistance to destructive influences in the form of attacks on NNBC knowledge extraction, biometric data compromise and attacks on biometric presentation (spoofing attacks). By using a classical neural network converter trained to classify real and fake images, the problem of counteracting spoofing attacks is solved, and the parameters of the user NNBC are protected by using the neural network container protection mechanism (NPC). As NNBC for authentication, a model of a neural network converter based on a new type of neurons based on a trigonometric measure of estimating the distance between biometric images of subjects in the subspace of feature pairs of the original face feature vector is used. The proposed proximity measure does not use the parameters of distributions and / or characteristics of the "Own" images, which ensures the protection of open biometric data of faces and NNBC knowledge from compromise.
- Text
- BUY for read the full text of article
- Buy
- 500.00 rub
