Документ взят из кэша поисковой машины. Адрес оригинального документа : http://graphics.cs.msu.ru/en/node/577
Дата изменения: Sun Apr 10 00:13:27 2016
Дата индексирования: Sun Apr 10 00:13:27 2016
Кодировка: UTF-8
Face Liveness Verification based on Feature Tracking in Face Recognition Systems | Graphics and Media Lab

Face Liveness Verification based on Feature Tracking in Face Recognition Systems

TitleFace Liveness Verification based on Feature Tracking in Face Recognition Systems
Publication TypeConference Paper
Year of Publication2010
AuthorsPotapov D, Konushin V
Refereed DesignationUnknown
Conference NamePRIA-10-2010
Pagination113-116
Publication LanguageEnglish
ISBN 978-5-7325-0972-4
KeywordsFace Recognition, Facial Features Localization, Impostor Detection, Landmarks Detection, Spoofing, Tracking
Abstract

In this paper, we present a new approach for liveness verification in facial recognition systems. While face recognition techniques are developing rapidly, a threat of cheating becomes more dangerous. It was shown that industrial systems can be easily deceived using a photo of a person?s face, if they don?t have impostor detection mechanism.
We propose a method to recognize this type of attacks based on facial features tracking. Using 2D facial features tracking it is able to distinguish motions of a real 3-dimensional object from a planar analogue (photo).
Our system prototype has shown the ability to distinguish a photo from a live person?s face in a video sequence. Our method adds very little computational overhead because facial features tracking is an inherent part of facial recognition systems.

Citation Key577