TY - GEN
T1 - The Quantitative Relationship between Adversarial Training and Robustness of CNN Model
AU - Wang, Jie
AU - Lu, Minyan
AU - Ai, Jun
AU - Sun, Xueyuan
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/11
Y1 - 2020/11
N2 - With the increasing application of deep neural networks in security-critical systems, robustness becomes an important property for deep learning. However, deep neural networks are very vulnerable to perturbations, especially adversarial attacks. The adversarial training by adding adversarial examples to the training set has become a common method to improve the performance of deep neural networks. In this paper, based on the existing robustness metrics and indicators, we study the quantitative influence of adversarial training on the robustness of network-in-network and residual network, and the differences of indicators are compared. The experimental results show that adversarial training can affect the accuracy and robustness of CNN models, and the variation trends of testing accuracy and robustness are opposite.
AB - With the increasing application of deep neural networks in security-critical systems, robustness becomes an important property for deep learning. However, deep neural networks are very vulnerable to perturbations, especially adversarial attacks. The adversarial training by adding adversarial examples to the training set has become a common method to improve the performance of deep neural networks. In this paper, based on the existing robustness metrics and indicators, we study the quantitative influence of adversarial training on the robustness of network-in-network and residual network, and the differences of indicators are compared. The experimental results show that adversarial training can affect the accuracy and robustness of CNN models, and the variation trends of testing accuracy and robustness are opposite.
KW - adversarial examples
KW - deep neural network
KW - global adversarial robustness
KW - local adversarial robustness
KW - robustness
UR - https://www.scopus.com/pages/publications/85100528022
U2 - 10.1109/DSA51864.2020.00092
DO - 10.1109/DSA51864.2020.00092
M3 - 会议稿件
AN - SCOPUS:85100528022
T3 - Proceedings - 2020 7th International Conference on Dependable Systems and Their Applications, DSA 2020
SP - 543
EP - 549
BT - Proceedings - 2020 7th International Conference on Dependable Systems and Their Applications, DSA 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 7th International Conference on Dependable Systems and Their Applications, DSA 2020
Y2 - 28 November 2020 through 29 November 2020
ER -