Title :
Distinguishable de-identified faces
Author :
Zongji Sun;Li Meng;Aladdin Ariyaeeinia
Author_Institution :
School of Engineering and Technology, University of Hertfordshire, Hatfield, AL10 9AB, UK
fDate :
5/1/2015 12:00:00 AM
Abstract :
The k-anonymity approach adopted by k-Same face de-identification methods enables these methods to serve their purpose of privacy protection. However, it also forces every k original faces to share the same de-identified face, making it impossible to track individuals in a k-Same de-identified video. To address this issue, this paper presents an approach to the creation of distinguishable de-identified faces. This new approach can serve privacy protection perfectly whilst producing de-identified faces that are as distinguishable as their original faces.
Keywords :
"Privacy","Face recognition","Active appearance model","Clustering algorithms","Testing","Brightness","Facial features"
Conference_Titel :
Automatic Face and Gesture Recognition (FG), 2015 11th IEEE International Conference and Workshops on
DOI :
10.1109/FG.2015.7285019