Exploring Magnitude Perturbation in Adversarial Attack & Defense
Keywords:
Magnitude perturbation, Adversarial attacks, Defense mechanisms, Robustness Security, Deep learning models, Success rate, Perturbation strengths, Attack effectiveness, Defense strategy evaluation, Limitations, Strengths, Security enhancementAbstract
Adversarial attacks pose a significant threat to the robustness and security of machine learning models. In recent years, researchers have focused on developing defense mechanisms to mitigate the impact of adversarial attacks. One such avenue of investigation involves examining the effects of magnitude perturbation on the success rate and effectiveness of these attacks, as well as evaluating the performance of various defense strategies in countering them. In this study, present a comprehensive analysis of the impact of magnitude perturbation in adversarial attacks and the effectiveness of defense mechanisms against them. The investigation of the influence of perturbation magnitudes on the success rate and transferability of attacks across different models and datasets. Furthermore, evaluation the performance of state-of-the-art defense mechanisms under varying perturbation strengths. The experimental results reveal intriguing insights into the behavior of adversarial attacks and the efficacy of defense mechanisms. The observation shows that increasing the magnitude of perturbations can significantly amplify the success rate of attacks, rendering models more vulnerable. Additionally, demonstrate that certain defense mechanisms exhibit varying levels of resilience against different perturbation magnitudes, shedding light on their limitations and strengths. The findings of this study contribute to a deeper understanding of the role of magnitude perturbation in adversarial attacks and the effectiveness of defense mechanisms. This knowledge can aid in the development of robust defense strategies and provide valuable insights for enhancing the security of machine learning systems in the face of adversarial threats.
Downloads
References
Akhtar, N., Mian, A., Kardan, N., Shah, M.: Advances in Adversarial Attacks and Defenses in Computer Vision: A Survey. IEEE Access. 9, 155161–155196 (2021). https://doi.org/10.1109/ACCESS.2021.3127960.
Almuflih, A.S., Vyas, D., Kapdia, V. V, Qureshi, M.R.N.M., Qureshi, K.M.R., Makkawi, E.A.: Novel exploit feature-map-based detection of adversarial attacks. Applied Sciences. 12, 5161 (2022).
Bakhti, Y., Fezza, S.A., Hamidouche, W., Deforges, O.: DDSA: A Defense against Adversarial Attacks Using Deep Denoising Sparse Autoencoder. IEEE Access. 7, 160397–160407 (2019). https://doi.org/10.1109/ACCESS.2019.2951526.
Dasgupta, D., Gupta, K.D.: Dual-filtering (DF) schemes for learning systems to prevent adversarial attacks. Complex and Intelligent Systems. (2022). https://doi.org/10.1007/s40747-022-00649-1.
Deldjoo, Y., Noia, T. Di, Merra, F.A.: A Survey on Adversarial Recommender Systems: From Attack/Defense Strategies to Generative Adversarial Networks. ACM Comput. Surv. 54, (2021). https://doi.org/10.1145/3439729.
Despiegel, V., Despiegel, V.: ScienceDirect Dynamic Dynamic Autoencoders Autoencoders Against Against Adversarial Adversarial Attacks Attacks. 00, (2023). https://doi.org/10.1016/j.procs.2023.03.104.
Folz, J., Palacio, S., Hees, J., Dengel, A.: Adversarial defense based on structure-to-signal autoencoders. Proceedings - 2020 IEEE Winter Conference on Applications of Computer Vision, WACV 2020. 3568–3577 (2020). https://doi.org/10.1109/WACV45572.2020.9093310.
Han, S., Lin, C., Shen, C., Wang, Q., Guan, X.: Interpreting Adversarial Examples in Deep Learning: A Review. (2023). https://doi.org/10.1145/3594869.
Hu, W., Tan, Y.: Generating Adversarial Malware Examples for Black-Box Attacks Based on GAN BT - Data Mining and Big Data. Presented at the (2022).
Jin, L., Tan, F., Jiang, S.: Generative Adversarial Network Technologies and Applications in Computer Vision. Computational Intelligence and Neuroscience. 2020, (2020). https://doi.org/10.1155/2020/1459107.
Kuribayashi, M.: Defense Against Adversarial Attacks BT - Frontiers in Fake Media Generation and Detection. Presented at the (2022). https://doi.org/10.1007/978-981-19-1524-6_6.
Laykaviriyakul, P., Phaisangittisagul, E.: Collaborative Defense-GAN for protecting adversarial attacks on classification system. Expert Systems with Applications. 214, 118957 (2023). https://doi.org/https://doi.org/10.1016/j.eswa.2022.118957.
Li, F., Du, X., Zhang, L.: Adversarial Attacks Defense Method Based on Multiple Filtering and Image Rotation. Discrete Dynamics in Nature and Society. 2022, (2022). https://doi.org/10.1155/2022/6124895.
Liang, H., He, E., Zhao, Y., Jia, Z., Li, H.: Adversarial Attack and Defense: A Survey. Electronics (Switzerland). 11, 1–19 (2022). https://doi.org/10.3390/electronics11081283.
Liu, S., Zhuang, Y., Ma, X., Wang, H., Cao, D.: An Adversarial Sample Defense Method Based on Saliency Information BT - Ubiquitous Security. Presented at the (2023).
Ren, K., Zheng, T., Qin, Z., Liu, X.: Adversarial Attacks and Defenses in Deep Learning. Engineering. 6, 346–360 (2020). https://doi.org/10.1016/j.eng.2019.12.012.
Ryu, G., Choi, D.: A hybrid adversarial training for deep learning model and denoising network resistant to adversarial examples. Applied Intelligence. 9174–9187 (2022). https://doi.org/10.1007/s10489-022-03991-6.
Samangouei, P., Kabkab, M., Chellappa, R.: Defense-Gan: Protecting classifiers against adversarial attacks using generative models. 6th International Conference on Learning Representations, ICLR 2018 - Conference Track Proceedings. (2018).
Singh, A.B., Awasthi, L.K., Urvashi: Defense Against Adversarial Attacks Using Chained Dual-GAN Approach BT - Smart Data Intelligence. Presented at the (2022).
Taheri, S., Khormali, A., Salem, M., Yuan, J.S.: Developing a robust defensive system against adversarial examples using generative adversarial networks. Big Data and Cognitive Computing. 4, 1–15 (2020). https://doi.org/10.3390/bdcc4020011.
Wang, J., Wu, H., Wang, H., Zhang, J., Luo, X., Ma, B.: Immune Defense: A Novel Adversarial Defense Mechanism for Preventing the Generation of Adversarial Examples. (2023).
Wang, Y., Sun, T., Li, S., Yuan, X., Ni, W., Hossain, E., Poor, H.V.: Adversarial Attacks and Defenses in Machine Learning-Powered Networks: A Contemporary Survey. 1–46 (2023).
Yadav, A., Upadhyay, A., Sharanya, S.: An integrated Auto Encoder-Block Switching defense approach to prevent adversarial attacks. (2022).
Yang, J., Shao, M., Liu, H., Zhuang, X.: Generating adversarial samples by manipulating image features with auto-encoder. International Journal of Machine Learning and Cybernetics. 14, 2499–2509 (2023). https://doi.org/10.1007/s13042-023-01778-w.
Yu, F., Wang, L., Fang, X., Zhang, Y.: The defense of adversarial example with conditional generative adversarial networks. Security and Communication Networks. 2020, (2020). https://doi.org/10.1155/2020/3932584.
Zhao, W., Mahmoud, Q.H., Alwidian, S.: Evaluation of GAN-Based Model for Adversarial Training. Sensors. 23, (2023). https://doi.org/10.3390/s23052697.
Zhou, M., Niu, Z., Wang, L., Zhang, Q., Hua, G.: Adversarial Ranking Attack and Defense. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 12359 LNCS, 781–799 (2020). https://doi.org/10.1007/978-3-030-58568-6_46.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
All papers should be submitted electronically. All submitted manuscripts must be original work that is not under submission at another journal or under consideration for publication in another form, such as a monograph or chapter of a book. Authors of submitted papers are obligated not to submit their paper for publication elsewhere until an editorial decision is rendered on their submission. Further, authors of accepted papers are prohibited from publishing the results in other publications that appear before the paper is published in the Journal unless they receive approval for doing so from the Editor-In-Chief.
IJISAE open access articles are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. This license lets the audience to give appropriate credit, provide a link to the license, and indicate if changes were made and if they remix, transform, or build upon the material, they must distribute contributions under the same license as the original.
