Puzzling face verification algorithms for privacy protection

This paper presents a new approach for de-identifying face images, i.e. for preventing automatic matching with public face collections. The overall motivation is to offer tools for privacy protection on social networks. We address this question by drawing a parallel between face de-identification and oracle attacks in digital watermarking. In our case, the identity of the face is seen as the watermark to be removed. Inspired by oracle attacks, we forge de-identified faces by superimposing a collection of carefully designed noise patterns onto the original face. The modification of the image is controlled to minimize the probability of good recognition while minimizing the distortion. In addition, these de-identified images are - by construction - made robust to counter attacks such as blurring. We present an experimental validation in which we de-identify LFW faces and show that resulting images are still recognized by human beings while deceiving a state-of-the-art face recognition algorithm.

[1]  2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014, Atlanta, GA, USA, December 3-5, 2014 , 2014, WIFS.

[2]  Bradley Malin,et al.  Preserving privacy by de-identifying face images , 2005, IEEE Transactions on Knowledge and Data Engineering.

[3]  Yana Welinder.  A Face Tells More than a Thousand Posts: Developing Face Recognition Privacy in Social Networks , 2012 .

[4]  Hyeonjoon Moon,et al.  The FERET evaluation methodology for face-recognition algorithms , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[5]  F. Pérez-González,et al.  Blind newton sensitivity attack , 2006 .

[6]  Fernando Pérez-González,et al.  The Return of the Sensitivity Attack , 2005, IWDW.

[7]  Bruce A. Draper,et al.  The CSU Face Identification Evaluation System: Its Purpose, Features, and Structure , 2003, ICVS.

[8]  Andrew Zisserman,et al.  Fisher Vector Faces in the Wild , 2013, BMVC.

[9]  Touradj Ebrahimi,et al.  A framework for the validation of privacy protection solutions in video surveillance , 2010, 2010 IEEE International Conference on Multimedia and Expo.

[10]  John W. Earl.  Tangential sensitivity analysis of watermarks using prior information , 2007, Electronic Imaging.

[11]  Frédéric Jurie,et al.  Face Recognition using Local Quantized Patterns , 2012, BMVC.

[12]  Martin Jaggi,et al.  Sparse Convex Optimization Methods for Machine Learning , 2011 .

[13]  Frédéric Jurie,et al.  PCCA: A new approach for distance learning from sparse pairwise constraints , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[14]  Martin Jaggi,et al.  Revisiting Frank-Wolfe: Projection-Free Sparse Convex Optimization , 2013, ICML.

[15]  Markus Dürmuth,et al.  Achieving Anonymity against Major Face Recognition Algorithms , 2013, Communications and Multimedia Security.

[16]  Ken Perlin,et al.  Improving noise , 2002, SIGGRAPH.

[17]  Jian Sun,et al.  Blessing of Dimensionality: High-Dimensional Feature and Its Efficient Compression for Face Verification , 2013, 2013 IEEE Conference on Computer Vision and Pattern Recognition.

[18]  Marwan Mattar,et al.  Labeled Faces in the Wild: A Database forStudying Face Recognition in Unconstrained Environments , 2008 .

[19]  Ralph Gross,et al.  Semi-supervised learning of multi-factor models for face de-identification , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[20]  R. Gross,et al.  Towards Real-World Face De-Identification , 2007, 2007 First IEEE International Conference on Biometrics: Theory, Applications, and Systems.