SIFT Keypoint Removal and Injection via Convex Relaxation

Scale invariant feature transform (SIFT), as one of the most popular local feature extraction algorithms, has been widely employed in many computer vision and multimedia security applications. Although SIFT has been extensively investigated from various perspectives, its security against malicious attacks has rarely been discussed. In this paper, we show that the SIFT keypoints can be effectively removed with minimized distortion on the processed image. The SIFT keypoint removal is formulated as a constrained optimization problem, where the constraints are carefully designed to suppress the existence of local extrema and prevent generating new keypoints within a local cuboid in the scale space. To hide the traces of performing SIFT keypoint removal, we then propose to inject a large number of fake SIFT keypoints into the previously cleaned image with minimized distortion. As demonstrated experimentally, our proposed SIFT removal and injection algorithms significantly outperform the state-of-the-art techniques. Furthermore, it is shown that the combined SIFT keypoint removal and injection attack strategy is capable of defeating the most powerful forensic detector designed for SIFT keypoint removal. Our results suggest that an authorization mechanism is required for SIFT-based systems to verify the validity of the input data, so as to achieve high reliability.

[1]  Yuenan Li.  Image copy-move forgery detection based on polar cosine transform and approximate nearest neighbor searching. , 2013, Forensic science international.

[2]  Mauro Barni,et al.  Forensic Analysis of SIFT Keypoint Removal and Injection , 2014, IEEE Transactions on Information Forensics and Security.

[3]  Laurent Amsaleg,et al.  Deluding image recognition in sift-based cbir systems , 2010, MiFor '10.

[4]  Alberto Del Bimbo,et al.  Ieee Transactions on Information Forensics and Security 1 a Sift-based Forensic Method for Copy-move Attack Detection and Transformation Recovery , 2022 .

[5]  Roberto Caldelli,et al.  SIFT match removal and keypoint preservation through dominant orientation shift , 2015, 2015 23rd European Signal Processing Conference (EUSIPCO).

[6]  Xunyu Pan,et al.  Region Duplication Detection Using Image Feature Matching , 2010, IEEE Transactions on Information Forensics and Security.

[7]  Yin Jianqin,et al.  Human Face Feature Extraction and Recognition Base on SIFT , 2008, 2008 International Symposium on Computer Science and Computational Technology.

[8]  Soo-Chang Pei,et al.  Secure and robust SIFT , 2009, ACM Multimedia.

[9]  David G. Lowe,et al.  What and Where: 3D Object Recognition with Accurate Pose , 2006, Toward Category-Level Object Recognition.

[10]  David G. Lowe,et al.  Scene modelling, recognition and tracking with invariant image features , 2004, Third IEEE and ACM International Symposium on Mixed and Augmented Reality.

[11]  Pascal Fua,et al.  LDAHash: Improved Matching with Smaller Descriptors , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[12]  Justin Zobel,et al.  Detection of near-duplicate images for web search , 2007, CIVR '07.

[13]  Roberto Caldelli,et al.  Exploiting perceptual quality issues in countering SIFT-based Forensic methods , 2014, 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[14]  Mauro Barni,et al.  Removal and injection of keypoints for SIFT-based copy-move counter-forensics , 2013, EURASIP J. Inf. Secur..

[15]  Eero P. Simoncelli,et al.  Image quality assessment: from error visibility to structural similarity , 2004, IEEE Transactions on Image Processing.

[16]  Heung-Kyu Lee,et al.  Rotation Invariant Localization of Duplicated Image Regions Based on Zernike Moments , 2013, IEEE Transactions on Information Forensics and Security.

[17]  Andrea Vedaldi,et al.  Vlfeat: an open and portable library of computer vision algorithms , 2010, ACM Multimedia.

[18]  Davide Cozzolino,et al.  Copy-move forgery detection based on PatchMatch , 2014, 2014 IEEE International Conference on Image Processing (ICIP).

[19]  Mauro Barni,et al.  SIFT keypoint removal and injection for countering matching-based image forensics , 2013, IH&MMSec '13.

[20]  Cong Wang,et al.  Towards Efficient Privacy-preserving Image Feature Extraction in Cloud Computing , 2014, ACM Multimedia.

[21]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[22]  Karel J. Zuiderveld,et al.  Contrast Limited Adaptive Histogram Equalization , 1994, Graphics Gems.

[23]  Chun-Shien Lu,et al.  Constraint-optimized keypoint inhibition/insertion attack: security threat to scale-space image feature extraction , 2012, ACM Multimedia.

[24]  Alberto Del Bimbo,et al.  Copy-move forgery detection and localization by means of robust clustering with J-Linkage , 2013, Signal Process. Image Commun..

[25]  Xingming Sun,et al.  Segmentation-Based Image Copy-Move Forgery Detection Scheme , 2015, IEEE Transactions on Information Forensics and Security.

[26]  Bingbing Ni,et al.  Order Preserving Sparse Coding , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[27]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[28]  Mauro Barni,et al.  Counter-forensics of SIFT-based copy-move detection by means of keypoint classification , 2013, EURASIP J. Image Video Process..

[29]  Hanno Scharr,et al.  A Scheme for Coherence-Enhancing Diffusion Filtering with Optimized Rotation Invariance , 2002, J. Vis. Commun. Image Represent..

[30]  Davide Cozzolino,et al.  Efficient Dense-Field Copy–Move Forgery Detection , 2015, IEEE Transactions on Information Forensics and Security.

[31]  Paolo Bestagini,et al.  Multi-Clue Image Tampering Localization , 2014, 2014 IEEE International Workshop on Information Forensics and Security (WIFS).

[32]  Manjunatha Mahadevappa,et al.  Brightness preserving dynamic fuzzy histogram equalization , 2010, IEEE Transactions on Consumer Electronics.

[33]  Jiantao Zhou,et al.  Sift keypoint removal via convex relaxation , 2015, 2015 IEEE International Conference on Multimedia and Expo (ICME).

[34]  Jun Luo,et al.  Person-Specific SIFT Features for Face Recognition , 2007, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.

[35]  Gerald Schaefer,et al.  UCID: an uncompressed color image database , 2003, IS&T/SPIE Electronic Imaging.

[36]  Laurent Amsaleg,et al.  Enlarging hacker's toolbox: Deluding image recognition by attacking keypoint orientations , 2012, 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[37]  Laurent Amsaleg,et al.  NV-Tree: An Efficient Disk-Based Index for Approximate Search in Very Large High-Dimensional Collections , 2009, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[38]  Yu Zhang,et al.  Detection of Copy-Move Forgery in Digital Images Using SIFT Algorithm , 2008, 2008 IEEE Pacific-Asia Workshop on Computational Intelligence and Industrial Application.

[39]  Yan Ke,et al.  Efficient Near-duplicate Detection and Sub-image Retrieval , 2004 .

[40]  Jessica Fridrich,et al.  Detection of Copy-Move Forgery in Digital Images , 2004 .

[41]  Osamah M. Al-Qershi,et al.  Passive detection of copy-move forgery in digital images: state-of-the-art. , 2013, Forensic science international.

[42]  Laurent Amsaleg,et al.  Challenging the security of Content-Based Image Retrieval systems , 2010, 2010 IEEE International Workshop on Multimedia Signal Processing.

[43]  Laurent Amsaleg,et al.  Understanding the security and robustness of SIFT , 2010, ACM Multimedia.

[44]  Mauro Barni,et al.  On the effectiveness of local warping against SIFT-based copy-move detection , 2012, 2012 5th International Symposium on Communications, Control and Signal Processing.

[45]  Trevor Darrell,et al.  Efficient image matching with distributions of local invariant features , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).