Abstract
Deep neural networks (DNNs) have been found to be easily mislead by adversarial examples that add small perturbations to inputs to produce false results. Different attack and defense strategies have been proposed to better study the security of deep neural networks. But these works only focus on an aspect such as attack or defense. In this work, we propose a robust GAN based on the attention mechanism, which uses the deep latent features of the original image as prior knowledge to generate adversarial examples, and it can jointly optimize the generator and discriminator in the case of adversarial attacks. The generator generates fake images based on the attention mechanism to deceive discriminator, the adversarial attacker perturbs the real images to deceive discriminator, and the discriminator wants to minimize the loss between fake images and adversarial images. Through this training, we can not only improve the quality of adversarial images generated by GAN, but also enhance the robustness of the discriminator under strong adversarial attacks. Experimental results show that our classifier is more robust than Rob-GAN [14], and the generator outperforms Rob-GAN on CIFAR-10.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Krizhevsky, I.S., Hinton, G.E.: ImageNet classification with deep convolutional neural networks. Commun. ACM 60(6), 84–90 (2017)
Bahdanau, D., Cho, K., Bengio, Y.: Neural machine translation by jointly learning to align and translate. In: 3rd International Conference on Learning Representations, pp. 11–17. ICLR, San Diego (2015)
Hinton, G., Deng, L., Yu, D., et al.: Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Sig. Process. Mag. 29(6), 82–97 (2012)
Yuan, Z., Lu, Y.Q., Wang, Z.G., Xue, Y.B.: Droid-Sec: deep learning in android malware detection. In: ACM SIGCOMM 2014 Conference, pp. 371–372. ACM, Chicago (2014)
Eykholt, K., Evtimov, I., Fernandes, E., et al.: Robust physical world attacks on deep learning models. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 1625–1634. IEEE Computer Society, Salt Lake City (2018)
Majumdar, R., Kunčak, V. (eds.): CAV 2017. LNCS, vol. 10426. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-63387-9
Cubuk, E.D., Zoph, B., Schoenholz, S.S., Le, Q.V.: Intriguing properties of adversarial examples. In: 6th International Conference on Learning Representations (ICLR), pp. 106–118. ICLR, Vancouver (2018)
Goodfellow, I.J., Shlens, J., Szegedy, C.: Explaining and harnessing adversarial examples. In: 3rd International Conference on Learning Representations, pp. 65–78. ICLR, San Diego (2015)
Tramèr, F., Kurakin, A., Papernot, N., GoodfellowI, B.D., McDaniel, P.: Ensemble adversarial training: attacks and defences. In: 5th International Conference on Learning Representations, pp. 123–142. ICLR, Toulon (2017)
Carlini, N., Wagner, D.: Towards evaluating the robustness of neural networks. In: IEEE Symposium on Security and Privacy, vol. 0, pp. 39–57. IEEE, San Jose (2017)
Zhao, Z., Dua, D., Singh, S.: Generating natural adversarial examples. In: 6th International Conference on Learning Representations, pp. 108–115. ICLR, Vancouver (2018)
Papernot, N., Mc Daniel, P., Wu, X., Jha, S., Swami, A.: Distillation as a defense to adversarial perturbations against deep neural networks. In: 2016 IEEE Symposium on Security and Privacy (SP), pp. 582–597. IEEE, San Jose (2016)
Wu, Y., Bamman, D., Russell, S.: Adversarial training for relation extraction. In: Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pp. 1778–1783. ACL (2017)
Liu, X., Hsieh, C.: Rob-GAN: generator, discriminator, and adversarial attacker. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 11226–11235. IEEE Computer Society, Long Beach (2019)
Madry, A., Makelov, A., Schmidt, L., Tsipras, D., Vladu, A.: Towards deep learning models resistant to adversarial attacks. In: 5th International Conference on Learning Representations, pp. 1538–1549. ICLR, Toulon (2017)
Kurakin, A., Goodfellow, I., Bengio, S.: Adversarial examples in the physical world. In: 5th International Conference on Learning Representations, pp. 995–1012. ICLR, Toulon (2017)
Hu, J., Li, S., Albanie, S., Sun, G., Wu, E.: Squeeze-and-excitation networks. IEEE Trans. Pattern Anal. Mach. Intell. 42(8), 2011–2023 (2018)
LeCun, Y., Cortes, C.: MNIST handwritten digit database. Proc. IEEE 86(11), 2278–2324 (1989)
Krizhevsky, A., Nair, V., Hinton, G.: CIFAR-10 (Canadian institute for advanced research). http://www.cs.toronto.edu/~kriz/cifar.html
Athalye, A., Carlini, N., Wagner, D.: Obfuscated gradients give a false sense of security: circumventing defenses to adversarial examples. In: 35th International Conference on Machine Learning, ICML 2018, vol. 1, pp. 436–448. IMLS, Stockholm (2018)
Acknowledgements
We acknowledge the support by the National Natural Science Foundation of China (No. 66162019); National Natural Science Foundation of China Enterprise Innovation and Development Joint Fund (No. U19B2044).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Wu, Q., Cao, C., Mai, J., Tao, F. (2021). Robust GAN Based on Attention Mechanism. In: Cheng, J., Tang, X., Liu, X. (eds) Cyberspace Safety and Security. CSS 2020. Lecture Notes in Computer Science(), vol 12653. Springer, Cham. https://doi.org/10.1007/978-3-030-73671-2_8
Download citation
DOI: https://doi.org/10.1007/978-3-030-73671-2_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-73670-5
Online ISBN: 978-3-030-73671-2
eBook Packages: Computer ScienceComputer Science (R0)