Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Morphing Detection Using a General-Purpose Face Recognition System

: Wandzik, Lukasz; Kaeding, Gerald; Vicente-Garcia, Raul

Postprint urn:nbn:de:0011-n-5343367 (359 KByte PDF)
MD5 Fingerprint: 9cf5017c7b6b626704caac8eca1741ff
© IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.
Erstellt am: 13.2.2019

Institute of Electrical and Electronics Engineers -IEEE-; IEEE Signal Processing Society; European Association for Speech, Signal and Image Processing -EURASIP-:
26th European Signal Processing Conference, EUSIPCO 2018 : 3-7 September 2018, Roma, Italy
Piscataway, NJ: IEEE, 2018
ISBN: 978-9-0827-9701-5
ISBN: 978-90-827970-0-8
ISBN: 978-1-5386-3736-4
European Signal Processing Conference (EUSIPCO) <26, 2018, Roma>
Bundesministerium für Bildung und Forschung BMBF (Deutschland)
147184; ANANAS
Konferenzbeitrag, Elektronische Publikation
Fraunhofer IPK ()
face tracking; feature extraction; face recognition; Task analysis; support vector machine

Image morphing has proven to be very successful at deceiving facial recognition systems. Such a vulnerability can be critical when exploited in an automatic border control scenario. Recent works on this topic rely on dedicated algorithms which require additional software modules deployed alongside an existing facial recognition system. In this work, we address the problem of morphing detection by using state-of-the-art facial recognition algorithms based on hand-crafted features and deep convolutional neural networks. We show that a general-purpose face recognition system combined with a simple linear classifier can be successfully used as a morphing detector. The proposed method reuses an existing feature extraction pipeline instead of introducing additional modules. It requires neither fine-tuning nor modifications to the existing recognition system and can be trained using only a small dataset. The proposed approach achieves state-of-the-art performance on our morphing datasets using a 5-fold cross-validation.