APE-GAN: Adversarial Perturbation Elimination with GAN

By Guoqing Jin, Shiwei Shen, Dongming Zhang, Feng Dai, Yongdong Zhang. 2017

245 citations recorded by the source corpus.

Although Deep Neural Networks could achieve state-of-the-art performance while recongnizing images, they often suffer a tremendous defeat from adversarial examples–inputs generated by utilizing imperceptible but intentional perturbations to samples from the datasets. So far, very few methods have provided a significant defense to adversarial examples. In this paper, an effective framework based Generative Adversarial Nets(GAN) is proposed to defense against the adversarial examples. The essense of the model is to eliminate the adversarial perturbations being highly aligned with the weight vectors of nueral models. Extensive experiments on benchmark datasets MNIST, CIFAR10 and ImageNet indicate that our framework is able to defense against adversarial examples effectively.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Recent Progress on Generative Adversarial Networks (GANs): A Survey (2019, background): “To against adversarial attacks, [80] and [81] use GANs to make the right defense.”
Towards the first adversarially robust neural network model on MNIST (2018, methodology): “• Manifold projections: An input sample is projected onto a learned data manifold (Samangouei et al., 2018; Ilyas et al., 2017; Shen et al., 2017; Song et al., 2018).”
Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey (2018, background): “[58] use the generator part of the network to rectify a perturbed image.”
MagNet and "Efficient Defenses Against Adversarial Attacks" are Not Robust to Adversarial Examples (2017, methodology): “Investigating APE-GAN’s Failure.”
Are adversarial examples inevitable? (2018, methodology): “This can be done via an auto-encoder (Meng & Chen, 2017; Shen et al., 2017), JPEG encoding (Das et al., 2018), or quantization (Xu et al., 2017).”
Disentangling Adversarial Robustness and Generalization (2018, methodology): “Defenses: Proposed defenses include detection and rejection methods [31, 25, 54, 60, 3, 63], pre-processing, quantization and dimensionality reduction methods [12, 73, 7], manifold-projection methods [39, 72, 82, 86], methods based on stochasticity/regularization or adapted architectures [109, 7, 68”
Label Smoothing and Logit Squeezing: A Replacement for Adversarial Training? (2019, background): “Approaches such as feature squeezing, denoising, and encoding (Xu et al., 2017; Samangouei et al., 2018; Shen et al., 2017; Meng & Chen, 2017) have had some success at pre-processing images to remove adversarial perturbations.”
A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications (2020, background): “To be against adversarial attacks [118], [119], references [120], [121] utilize GANs to conduct the right defense.”
Adversarial Attacks and Defenses in Deep Learning (2020, methodology): “Defense-GAN and adversarial perturbation elimination GAN (APE-GAN) are two typical algorithms among all these works.”
Evaluation and Design of Robust Neural Network Defenses (2018, methodology): “39 3.1.2 MagNet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 3.1.3 “Efficient Defenses against Adversarial Attack” . . . . . . . . . 48 3.1.4 APE-GAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 3.2 Detection Defenses . . . . . . . . . . . . . . . . . . . . . . . . ”
A Self-supervised Approach for Adversarial Robustness (2020, background): “[41] proposed to recover adversaries using GAN and [35] super-resolve images to minimize adversarial effect.”
Defending Against Adversarial Attacks by Leveraging an Entire GAN (2018, background): “APE-GAN (Shen et al., 2017) modiﬁes a GAN such that the generator learns to clean the adversarial samples, whereas the discriminator tries to discriminate be-tween real and adversarial input.”
Robust Perception through Analysis by Synthesis (2018, background): “This includes in particular DefenseGAN [Samangouei et al., 2018], Adversarial Perturbation Elimination GAN [Shen et al., 2017] and Robust Manifold Defense [Ilyas et al., 2017b], all of which project an image onto the manifold deﬁned by a generator network G .”
Defense without Forgetting: Continual Adversarial Defense with Anisotropic & Isotropic Pseudo Replay (2024, background): “To maintain adversarial robustness under attacks, early adversarial defense are often heuristic, including input transformation [48], model ensemble [4], adversarial denoiser [40].”
DeepFense: Online Accelerated Defense Against Adversarial Deep Learning (2017, background): “Defenses Against Adversarial Attacks [9], and APE-GAN [10] in the white-box setting.”
Secure and Robust Machine Learning for Healthcare: A Survey (2020, background): “In a similar study [181], generative adversarial networks (GANs) are used for cleaning of adversarial noise.”
Ensembling with Deep Generative Views (2021, methodology): “[49] demonstrate projecting adversarial images through a generator on MNIST and Fashion-MNIST, while other works [51, 25, 19] train models using the generator to improve robustness.”