VeriFace: Defending against Adversarial Attacks in Face Verification Systems
Source abstract
Abstract
Face verification systems are critical in a wide range of applications, such as security systems and biometric authentication. However, these systems are vulnerable to adversarial attacks, which can significantly compromise their accuracy and reliability. Adversarial attacks are designed to deceive the face verification system by adding subtle perturbations to the input images. These perturbations can be imperceptible to the human eye but can cause the system to misclassify or fail to recognize the person in the image. To address this issue, we propose a novel system called VeriFace that comprises two defense mechanisms, adversarial detection, and adversarial removal. The first mechanism, adversarial detection, is designed to identify whether an input image has been subjected to adversarial perturbations. The second mechanism, adversarial removal, is designed to remove these perturbations from the input image to ensure the face verification system can accurately recognize the person in the image. To evaluate the effectiveness of the VeriFace system, we conducted experiments on different types of adversarial attacks using the Labelled Faces in the Wild (LFW) dataset. Our results show that the VeriFace adversarial detector can accurately identify adversarial images with a high detection accuracy of 100%. Additionally, our proposed VeriFace adversarial removal method has a significantly lower attack success rate of 6.5% compared to state-of-the-art removal methods.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Face recognition and analysis · Biometric Identification and Security · Adversarial Robustness in Machine Learning
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (41)
Show source-provided references
1. Towards deep learning models resistant to adversarial attacks
2. I. J. Goodfellow, J. Shlens and C. Szegedy, “Explaining and harnessing adversarial examples,” arXiv:1412.6572, 2014.
3. Advfaces: Adversarial face synthesis
4. Deep tree learning for zero-shot face anti-spoofing
5. On disentangling spoof trace for generic face anti-spoofing
6. On the detection of digital face manipulation
7. Efficient decision-based black-box adversarial attacks on face recognition
8. Fast geometrically-perturbed adversarial faces
9. Fawkes: Protecting privacy against unauthorized deep learning models
10. Nonlinear total variation-based noise removal algorithms10.1016/0167-2789(92)90242-F ↗
11. Countering adversarial images using input transformations
12. Adaptive wavelet thresholding for image denoising and compression
13. U. Shaham, J. Garritano, Y. Yamada, E. Weinberger, A. Cloninger et al., “Defending against Adversarial Images using Basis Functions Transformations,” ArXiv, vol. abs/1803.10840, 2018.
14. A. G. Howard, M. Zhu, B. Chen, D. Kalenichenko, W. Wang et al., “Mobilenets: Efficient convolutional neural networks for mobile vision applications,” ArXiv, vol. abs/1704.04861, 2017. http://arxiv.org/abs/1704.04861
15. Adversarial examples are not easily detected: Bypassing ten detection methods
16. Detection of face recognition adversarial attacks10.1016/j.cviu.2020.103103 ↗
17. Are image-agnostic universal adversarial perturbations for face recognition difficult to detect?
18. K. Grosse, P. Manoharan, N. Papernot, M. Backes and P. Mcdaniel, “On the (Statistical) detection of adversarial examples,” arXiv:1702.06280, 2017. http://arxiv.org/abs/1702.06280
19. Z. Gong, W. Wang and W. S. Ku, “Adversarial and clean data are not twins,” arXiv:abs/1704.04960, 2017.
20. W. Xu, D. Evans and Y. Qi, “Feature squeezing: Detecting adversarial examples in deep neural networks,” arXiv:1704.01155, 2017. http://arxiv.org/abs/1704.0115510.14722/ndss.2018.23198 ↗
21. On detecting adversarial perturbations
22. Image quilting for texture synthesis and transfer
23. DeepFool: A simple and accurate method to fool deep neural networks
24. Towards evaluating the robustness of neural networks
25. Feature distillation: DNN-oriented jpeg compression against adversarial examples
26. Imagenet: A large-scale hierarchical image database
27. Analysis of adversarial attacks on face verification systems
28. Ensemble adversarial training: Attacks and defenses
29. Early methods for detecting adversarial images
30. Going deeper with convolutions
31. Very deep convolutional networks for large-scale image recognition
32. Labeled faces in the wild: A database for studying face recognition in unconstrained environments
33. Joint face detection and alignment using multitask cascaded convolutional networks10.1109/LSP.2016.2603342 ↗
34. D. Yi, Z. Lei, S. Liao and S. Li, “Learning face representation from scratch,” arXiv:1411.7923, 2014. https://arxiv.org/abs/1411.7923
35. FaceGuard: A self-supervised defense against adversarial face images
36. Adaptive principal components and image denoising
37. B. Arjun, C. Daniel and M. Prateek, “Dimensionality reduction as a defense against evasion attacks on machine learning classifiers,” arXiv:1704.02654, 2017. http://arxiv.org/abs/1704.02654
38. Manifold projection for adversarial defense on face recognition10.1007/978-3-030-58577-8_18 ↗
39. Identity mappings in deep residual networks
40. An introduction to wavelets10.1109/99.388960 ↗
41. Obfuscated gradients give a false sense of security: Circumventing defenses to adversarial examples