Privacy Protection Face Verification System Using Homomorphic Encryption and Siamese Network 


Vol. 47,  No. 3, pp. 551-558, Mar.  2022
10.7840/kics.2022.47.3.551


PDF
  Abstract

Deep learning, which has shown good performance in image recognition, has recently become an essential element in various fields, and as the importance of biometric recognition increases, deep learning models suitable for face verification are being studied. However, deep learning model uses the user’s face image to learn. In this case, it implies that the Inversion attack, the vulnerability of the deep learning model may extract the user’s face image and not protect privacy. Therefore, there is a need for a technology that enables normal verification of the deep learning model while protecting the privacy of the face image, which is the user’s personal information. In this paper, to protect users’ privacy, encryption using Paillier isomorphic encryption is applied to facial images used for deep learning and reasoning. In addition, by using a Siamese network that infers using similarities between feature vectors, high accuracy in the learning and verification process was also shown for encrypted data that is difficult to distinguish.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Cite this article

[IEEE Style]

H. Cho, H. Kang, J. Sim, Y. Hong, H. Kim, "Privacy Protection Face Verification System Using Homomorphic Encryption and Siamese Network," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 3, pp. 551-558, 2022. DOI: 10.7840/kics.2022.47.3.551.

[ACM Style]

Hyun-jin Cho, Hyo-eun Kang, Jun-seok Sim, Yoon-young Hong, and Ho-won Kim. 2022. Privacy Protection Face Verification System Using Homomorphic Encryption and Siamese Network. The Journal of Korean Institute of Communications and Information Sciences, 47, 3, (2022), 551-558. DOI: 10.7840/kics.2022.47.3.551.

[KICS Style]

Hyun-jin Cho, Hyo-eun Kang, Jun-seok Sim, Yoon-young Hong, Ho-won Kim, "Privacy Protection Face Verification System Using Homomorphic Encryption and Siamese Network," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 3, pp. 551-558, 3. 2022. (https://doi.org/10.7840/kics.2022.47.3.551)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880