TY - GEN
T1 - Towards Evaluating the Security of Human Computable Passwords Using Neural Networks
AU - Murata, Issei
AU - He, Pengju
AU - Gu, Yujie
AU - Sakurai, Kouichi
N1 - Funding Information:
Acknowledgement. The last author, Kouichi Sakurai, is grateful to Support Center for Advanced Telecommunications Technology Research (SCAT) for their academic support on this research.
Publisher Copyright:
© 2023, Springer Nature Switzerland AG.
PY - 2023
Y1 - 2023
N2 - Passwords are playing a major role for authentication in our daily life. However contemporary passwords are typically either difficult to remember or vulnerable to various attacks. In 2017, Blocki, Blum, Datta and Vempala introduced the concept of human computable passwords as a promising authentication method. The fundamental concerns for designing human computable passwords are their usability and security. So far, the security evaluation on human computable passwords authentication schemes is mainly based on complexity-theoretic analysis. In this paper, we initially investigate the security of human computable passwords against neural network-based adversarial attacks. Specifically, we employ the typical multilayer perceptron (MLP) model to attempt to attack the human computable passwords authentication scheme proposed by Blocki-Blum-Datta-Vempala. We present implementation results and the corresponding analysis as well. Our results imply that it is possible for an MLP to learn a simple function, but is difficult for an MLP to learn piecewise functions well.
AB - Passwords are playing a major role for authentication in our daily life. However contemporary passwords are typically either difficult to remember or vulnerable to various attacks. In 2017, Blocki, Blum, Datta and Vempala introduced the concept of human computable passwords as a promising authentication method. The fundamental concerns for designing human computable passwords are their usability and security. So far, the security evaluation on human computable passwords authentication schemes is mainly based on complexity-theoretic analysis. In this paper, we initially investigate the security of human computable passwords against neural network-based adversarial attacks. Specifically, we employ the typical multilayer perceptron (MLP) model to attempt to attack the human computable passwords authentication scheme proposed by Blocki-Blum-Datta-Vempala. We present implementation results and the corresponding analysis as well. Our results imply that it is possible for an MLP to learn a simple function, but is difficult for an MLP to learn piecewise functions well.
UR - http://www.scopus.com/inward/record.url?scp=85148021389&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85148021389&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-25659-2_22
DO - 10.1007/978-3-031-25659-2_22
M3 - Conference contribution
AN - SCOPUS:85148021389
SN - 9783031256585
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 303
EP - 312
BT - Information Security Applications - 23rd International Conference, WISA 2022, Revised Selected Papers
A2 - You, Ilsun
A2 - Youn, Taek-Young
PB - Springer Science and Business Media Deutschland GmbH
T2 - 23rd International Conference on Information Security Applications, WISA 2022
Y2 - 24 August 2022 through 26 August 2022
ER -