TU Darmstadt / ULB / TUbiblio

The effect of face morphing on face image quality

Fu, Biying ; Spiller, Noemie ; Chen, Cong ; Damer, Naser (2021):
The effect of face morphing on face image quality.
In: BIOSIG 2021: Proceedings of the 20th International Conference of the Biometrics Special Interest Group,
IEEE, 20th International Conference of the Biometrics Special Interest Group, virtual Conference, 15.-17.09.2021, ISBN 978-1-6654-2693-0,
DOI: 10.1109/BIOSIG52210.2021.9548302,
[Conference or Workshop Item]

Abstract

Face morphing poses high security risk, which moti-vates the work on detection algorithms, as well as on anticipating novel morphing approaches. Using the statistical and perceptual image quality of morphed images in previous works has shown no clear correlation between the image quality and the realistic appearance. This motivated our study on the effect of face morphing on image quality and utility, we, therefore, applied eight general image quality metrics and four face-specific image utility metrics. We showed that MagFace (face utility metric) shows a clear difference between the bona fide and the morph images, regardless if they were digital or re-digitized. While most quality and utility metrics do not capture the artifacts introduced by the morphing process. Acknowledged that morphing artifacts are more apparent in certain areas of the face, we further investigated only these areas, for instance, tightly cropped face, nose, eyes, and mouth regions. We found that especially close to the eyes and the nose regions, using general image quality metrics as MEON and dipIQ can capture the image quality deterioration introduced by the morphing process.

Item Type: Conference or Workshop Item
Erschienen: 2021
Creators: Fu, Biying ; Spiller, Noemie ; Chen, Cong ; Damer, Naser
Title: The effect of face morphing on face image quality
Language: English
Abstract:

Face morphing poses high security risk, which moti-vates the work on detection algorithms, as well as on anticipating novel morphing approaches. Using the statistical and perceptual image quality of morphed images in previous works has shown no clear correlation between the image quality and the realistic appearance. This motivated our study on the effect of face morphing on image quality and utility, we, therefore, applied eight general image quality metrics and four face-specific image utility metrics. We showed that MagFace (face utility metric) shows a clear difference between the bona fide and the morph images, regardless if they were digital or re-digitized. While most quality and utility metrics do not capture the artifacts introduced by the morphing process. Acknowledged that morphing artifacts are more apparent in certain areas of the face, we further investigated only these areas, for instance, tightly cropped face, nose, eyes, and mouth regions. We found that especially close to the eyes and the nose regions, using general image quality metrics as MEON and dipIQ can capture the image quality deterioration introduced by the morphing process.

Title of Book: BIOSIG 2021: Proceedings of the 20th International Conference of the Biometrics Special Interest Group
Publisher: IEEE
ISBN: 978-1-6654-2693-0
Uncontrolled Keywords: Biometrics, Face recognition, Deep learning, Morphing Attack, Quality estimation
Divisions: 20 Department of Computer Science
20 Department of Computer Science > Interactive Graphics Systems
Event Title: 20th International Conference of the Biometrics Special Interest Group
Event Location: virtual Conference
Event Dates: 15.-17.09.2021
Date Deposited: 29 Sep 2021 13:14
DOI: 10.1109/BIOSIG52210.2021.9548302
Export:
Suche nach Titel in: TUfind oder in Google
Send an inquiry Send an inquiry

Options (only for editors)
Show editorial Details Show editorial Details