臺灣博碩士論文加值系統

臉部辨識是廣為人知的影像識別應用，並且在當代社會很常被應用來做身份的辨別。然而，許多系統卻忽視了保護臉部影像的重要性。若是臉部的影像未被保護起來，惡意人士便能竊取、複製影像，並偽裝成為其他人。為了解決這個問題，我們設計了一個能同時保護臉部影像的臉部辨識的系統。我們使用 DeepID2 卷積類神經網路來提取臉部特徵、使用最大期望演算法來來協助進行臉部辨識。為了確保臉部影像的安全，我們使用同態加密來將臉部特徵加密起來，使最大期望演算法能在密文域中做計算。基於不同的安全性考量，我們針對社區設計了三套不同的門禁系統，並在最後對三個系統做辨識準確率與執行速度等實驗，比較各優缺點。

Face Verification is a well known image analysis application and wildly used for recognizing individuals in contemporary society. However, in real applications, some of recognition systems ignore the importance of protecting the facial images that are used for verification. If the facial images are not protected, malicious people can steal and copy the images to disguise as someone else. To conquer this problem, we design a secure face verification system that can also protect the facial images to be imitated. In our work, we use the DeepID2 convolutional neural network to extract the feature of a facial image and use the EM algorithm to do the facial verification problem. In order to keep the facial images privacy, we use the homomorphic encryption scheme to encrypt the facial data and compute the EM algorithm in the ciphertext domain. Based on difference privacy concerns, we build up three face recognition systems for surveillance or entry and exist control of a local community. Lastly, we conduct experiments of accuracy and time consuming and compare pros can cons between these systems.

口試委員審定書 i
致謝 ii
中文摘要 iii
Abstract iv
Contents v
List of Figures vi
List of Tables vii
1. Introduction 1
1.1. Problem Definition 1
1.2. Contribution 1
2. Background 3
3. Preliminaries 6
3.1. DeepID2 (Facial Feature Extraction) 6
3.2. EM Algorithm Model (Face Verification Algorithm) 8
3.2.1. Reformulation of the Problem 10
3.2.2. Joint Formulation with Prior Information 11
3.2.3. EM Algorithm 12
3.3. Encryption Mechanisms 16
3.3.1. Homomorphic Encryption 16
3.3.2. The Paillier Cryptosystems 17
3.3.3. The Fan and Vercauteren Scheme 19
3.4. Some Secure Computational Protocols 21
3.4.1. The Secure Comparison Protocol 21
3.4.2. The Secure Matrix Multiplication Protocol 22
3.5. Dataset 23
4. The Proposed Scheme 25
4.1. Overview 25
4.2. Application Scenarios 26
4.2.1. The First Scenario 29
4.2.2. The Second Scenario 31
4.2.3. The Third Scenario 34
4.3. A Comparison among the three Scenarios 35
5. Performance Evaluation 36
5.1. Accuracy 36
5.2. Timing Performance 37
6. Conclusion 39
7. References 40

1. RAMKUMAR, G.; MANIKANDAN, M. Face recognition-survey. Sciencepublication. org, 2013, 260-268.
2. SADEGHI, Ahmad-Reza; SCHNEIDER, Thomas; WEHRENBERG, Immo. Efficient Privacy-Preserving Face Recognition. In: ICISC. 2009. p. 229-244.
3. LAGENDIJK, Reginald L.; ERKIN, Zekeriya; BARNI, Mauro. Encrypted signal processing for privacy protection: Conveying the utility of homomorphic encryption and multiparty computation. IEEE Signal Processing Magazine, 2013, 30.1: 82-105.
4. ERKIN, Zekeriya, et al. Privacy-preserving face recognition. In: International Symposium on Privacy Enhancing Technologies Symposium. Springer, Berlin, Heidelberg, 2009. p. 235-253.
5. SUN, Yi, et al. Deep learning face representation by joint identification-verification. In: Advances in neural information processing systems. 2014. p. 1988-1996.
6. SUN, Yi; WANG, Xiaogang; TANG, Xiaoou. Deep convolutional network cascade for facial point detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2013. p. 3476-3483.
7. NAIR, Vinod; HINTON, Geoffrey E. Rectified linear units improve restricted boltzmann machines. In: Proceedings of the 27th international conference on machine learning (ICML-10). 2010. p. 807-814.
8. MOGHADDAM, Baback; JEBARA, Tony; PENTLAND, Alex. Bayesian face recognition. Pattern Recognition, 2000, 33.11: 1771-1782.
9. CHEN, Dong, et al. Bayesian face revisited: A joint formulation. Computer Vision–ECCV 2012, 2012, 566-579.
10. DEMPSTER, Arthur P.; LAIRD, Nan M.; RUBIN, Donald B. Maximum likelihood from incomplete data via the EM algorithm. Journal of the royal statistical society. Series B (methodological), 1977, 1-38.
11. ASLETT, Louis JM; ESPERANÇA, Pedro M.; HOLMES, Chris C. A review of homomorphic encryption and software tools for encrypted statistical machine learning. arXiv preprint arXiv:1508.06574, 2015.
12. PAILLIER, Pascal, et al. Public-key cryptosystems based on composite degree residuosity classes. In: Eurocrypt. 1999. p. 223-238.
13. FAN, Junfeng; VERCAUTEREN, Frederik. Somewhat Practical Fully Homomorphic Encryption. IACR Cryptology ePrint Archive, 2012, 2012: 144.
14. REGEV, Oded. On lattices, learning with errors, random linear codes, and cryptography. Journal of the ACM (JACM), 2009, 56.6: 34.
15. LYUBASHEVSKY, Vadim; PEIKERT, Chris; REGEV, Oded. On ideal lattices and learning with errors over rings. In: Annual International Conference on the Theory and Applications of Cryptographic Techniques. Springer, Berlin, Heidelberg, 2010. p. 1-23.
16. VEUGEN, Thijs. Comparing encrypted data. Multimedia Signal Processing Group, Delft University of Technology, The Netherlands, and TNO Information and Communication Technology, Delft, The Netherlands, Tech. Rep, 2011.
17. BOST, Raphael, et al. Machine Learning Classification over Encrypted Data. In: NDSS. 2015.
18. HUANG, Gary B., et al. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. Technical Report 07-49, University of Massachusetts, Amherst, 2007.
19. YI, Dong, et al. Learning face representation from scratch. arXiv preprint arXiv:1411.7923, 2014.
20. CHEN, Dong, et al. Blessing of dimensionality: High-dimensional feature and its efficient compression for face verification. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. 2013. p. 3025-3032.
21. DUNTEMAN, George H. Principal components analysis. Sage, 1989.
22. https://github.com/rbost/ciphermed
23. https://github.com/CryptoExperts/FV-NFLlib