Scientific Visualization, 2019, volume 11, number 2, pages 99 - 113, DOI: 10.26583/sv.11.2.08
Use of methods of visual analytics to obtain integrated indicator of the quality of biometric authentication using the mechanism gesture manipulation
Author: Yu.E. Kozlov1
Financial University under the Government of the Russian Federation (Financial University), Moscow, Russia
1 ORCID: 0000-0002-4448-0232, kozlovye@yandex.ru
Abstract
Employing multimodal authentication techniques in the mobile applications which use several interconnected parameters in their operation demands studying them from the point of view of reliability. One of widespread ways of defining characteristics of such techniques is defining types I and II errors, and receiving a integrated quality indicator which can be defined upon visualizing type I errors (FRR) and type II errors (FAR).
Joint construction of dependency charts of FRR and FAR on the threshold value is possible only when using computer modeling and tools of visual analytics allowing to be convinced of correctness of the drawn conclusions visually. Traditionally for biometric authentication systems having probability indicators of reliability, the parameter which characterizes the integrated quality indicator is the equal error rate (EER) which is at the intersection of FRR and FAR curves.
The given article offers the research which has allowed to compare visually efficiency of a number of algorithms when carrying out the authentication procedure and to define the most favorable ones for biometric authentication in mobile applications by means of the in-air signature mechanism in the context of reliability.
Keywords: in-air signature, MTDS, signature, wearable device, authentication.
