Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
175 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
42 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Reversing Deep Face Embeddings with Probable Privacy Protection (2310.03005v1)

Published 4 Oct 2023 in cs.CV

Abstract: Generally, privacy-enhancing face recognition systems are designed to offer permanent protection of face embeddings. Recently, so-called soft-biometric privacy-enhancement approaches have been introduced with the aim of canceling soft-biometric attributes. These methods limit the amount of soft-biometric information (gender or skin-colour) that can be inferred from face embeddings. Previous work has underlined the need for research into rigorous evaluations and standardised evaluation protocols when assessing privacy protection capabilities. Motivated by this fact, this paper explores to what extent the non-invertibility requirement can be met by methods that claim to provide soft-biometric privacy protection. Additionally, a detailed vulnerability assessment of state-of-the-art face embedding extractors is analysed in terms of the transformation complexity used for privacy protection. In this context, a well-known state-of-the-art face image reconstruction approach has been evaluated on protected face embeddings to break soft biometric privacy protection. Experimental results show that biometric privacy-enhanced face embeddings can be reconstructed with an accuracy of up to approximately 98%, depending on the complexity of the protection algorithm.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (24)
  1. G. Mai, K. Cao, P. Yuen, and A. Jain., “On the reconstruction of face images from deep face templates,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 41, no. 5, pp. 1188–1202, 2019.
  2. H. Shahreza, V. Hahn, and S. Marcel, “Face reconstruction from deep facial embeddings using a convolutional neural network,” in 2022 IEEE Intl. Conf. on Image Processing (ICIP), 2022, pp. 1211–1215.
  3. X. Dong, Z. Miao, L. Ma, J. Shen, Z. Jin, Z. Guo, and A. Teoh, “Reconstruct face from features using gan generator as a distribution constraint,” 2022.
  4. P. Terhörst, D. Fährmann, N. Damer, F. Kirchbuchner, and A. Kuijper, “On soft-biometric information stored in biometric face embeddings,” IEEE Transactions on Biometrics, Behavior, and Identity Science, pp. 1–17, 2021.
  5. European Council, “Regulation of the european parliament and of the council on the protection of individuals with regard to the processing of personal data and on the free movement of such data (general data protection regulation),” April 2016.
  6. B. Meden, P. Rot, P. Terhörst, N. Damer, A. Kuijper, W. Scheirer, A. Ross, P. Peer, and V. Štruc, “Privacy–enhancing face biometrics: A comprehensive survey,” IEEE Transactions on Information Forensics and Security, vol. 16, pp. 4147–4183, 2021.
  7. V. Mirjalili, S. Raschka, and A. Ross, “Privacynet: Semi-adversarial networks for multi-attribute face privacy,” IEEE Transactions on Image Processing, vol. 29, pp. 9400–9412, 2020.
  8. P. Terhörst, K. Riehl, N. Damer, P. Rot, B. Bortolato, F. Kirchbuchner, V. Struc, and A. Kuijper, “Pe-miu: A training-free privacy-enhancing face recognition approach based on minimum information units,” IEEE Access, vol. 8, pp. 93 635–93 647, 2020.
  9. C. Rathgeb and A. Uhl, “A survey on biometric cryptosystems and cancelable biometrics,” EURASIP Journal on Information Security, vol. 3, March 2011.
  10. D. Osorio-Roig, C. Rathgeb, P. Drozdowski, P. Terhörst, V. Štruc, and C. Busch, “An attack on facial soft-biometric privacy enhancement,” Trans. on Biometrics, Behavior, and Identity Science (TBIOM), vol. 4, no. 2, pp. 263–275, April 2022.
  11. P. Melzi, C. Rathgeb, R. Tolosana, R. Vera-Rodriguez, and C. Busch, “An overview of privacy-enhancing technologies in biometric recognition,” arXiv preprint arXiv:2206.10465, 2022.
  12. P. Terhörst, M. Huber, N. Damer, F. Kirchbuchner, and A. Kuijper, “Unsupervised enhancement of soft-biometric privacy with negative face recognition,” arXiv preprint arXiv:2002.09181, 2020.
  13. B. Bortolato, M. Ivanovska, P. Rot, J. Križaj, P. Terhörst, N. Damer, P. Peer, and V. Štruc, “Learning privacy-enhancing face representations through feature disentanglement,” in 15th IEEE Intl. Conf. on Automatic Face and Gesture Recognition (FG 2020).   IEEE, 2020, pp. 495–502.
  14. P. Terhörst, N. Damer, F. Kirchbuchner, and A. Kuijper, “Unsupervised privacy-enhancement of face representations using similarity-sensitive noise transformations,” Applied Intelligence, vol. 49, no. 8, pp. 3043–3060, 2019.
  15. P. Terhörst, M. Huber, N. Damer, P. Rot, F. Kirchbuchner, V. Štruc, and A. Kuijper, “Privacy evaluation protocols for the evaluation of soft-biometric privacy-enhancing technologies,” in International Conference of the Biometrics Special Interest Group (BIOSIG), 2020, pp. 215–222.
  16. B. Razeghi, S. Rezaeifar, S. Ferdowsi, T. Holotyak, and S. Voloshynovskiy, “Compressed data sharing based on information bottleneck model,” in ICASSP 2022-2022 IEEE Intl. Conf. on Acoustics, Speech and Signal Processing (ICASSP).   IEEE, 2022, pp. 3009–3013.
  17. S. Rezaeifar, S. Voloshynovskiy, M. A. Jirhandeh, and V. Kinakh, “Privacy-preserving image template sharing using contrastive learning,” Entropy, vol. 24, no. 5, p. 643, 2022.
  18. B. Razeghi, F. Calmon, D. Gunduz, and S. Voloshynovskiy, “Bottlenecks club: Unifying information-theoretic trade-offs among complexity, leakage, and utility,” IEEE Transc. on Information Forensics and Security, vol. 18, pp. 2060–2075, 2023.
  19. J.-J. Howard, Y.-B. Sirotin, and A.-R. Vemury, “The effect of broad and specific demographic homogeneity on the imposter distributions and false match rates in face recognition algorithm performance,” in 2019 IEEE 10th Intl. Conf. on Biometrics Theory, Applications and Systems (BTAS).   IEEE, 2019, pp. 1–8.
  20. J. Deng, J. Guo, J. Yang, N. Xue, I. Kotsia, and S. Zafeiriou, “Arcface: Additive angular margin loss for deep face recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 10, pp. 5962–5979, 2022.
  21. B. Fadi, D. Naser, K. Florian, and K. Arjan, “Elasticface: Elastic margin loss for deep face recognition,” 2022.
  22. T. Karras, S. Laine, and T. Aila, “A style-based generator architecture for generative adversarial networks,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4401–4410.
  23. P. Terhorst, N.-K. Kolf, N. Damer, F. Kirchbuchner, and A. Kuijper, “Ser-fiq: Unsupervised estimation of face image quality based on stochastic embedding robustness,” in Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition, 2020, pp. 5651–5660.
  24. G. Huang, M. Ramesh, T. Berg, and E. Learned-Miller, “Faces in the wild: a database for studying face recognition in unconstrained environments,” Technical Report, pp. 07–49, 2007.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now