mlsanityData to 14 Oct 2026

Salient Feature Extractor for Adversarial Defense on Deep Neural Networks

arXiv 2105.06807 ↗First posted 14 May 2021cs.CV cs.AI cs.CRcited by 0

Not on the map. Why?

Abstract (arXiv)

Recent years have witnessed unprecedented success achieved by deep learning models in the field of computer vision. However, their vulnerability towards carefully crafted adversarial examples has also attracted the increasing attention of researchers. Motivated by the observation that adversarial examples are due to the non-robust feature learned from the original dataset by models, we propose the concepts of salient feature(SF) and trivial feature(TF). The former represents the class-related feature, while the latter is usually adopted to mislead the model. We extract these two features with coupled generative adversarial network model and put forward a novel detection and defense method named salient feature extractor (SFE) to defend against adversarial attacks. Concretely, detection is realized by separating and comparing the difference between SF and TF of the input. At the same time, correct labels are obtained by re-identifying SF to reach the purpose of defense. Extensive experiments are carried out on MNIST, CIFAR-10, and ImageNet datasets where SFE shows state-of-the-art results in effectiveness and efficiency compared with baselines. Furthermore, we provide an interpretable understanding of the defense and detection process.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 35 of its 54 references that are in the map. First posted on arXiv 14 May 2021.

Builds on

35 of 54 references.

References · 35

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,364
  2. 2014Explaining and Harnessing Adversarial Examplescited by 7,364
  3. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
  6. 2013Intriguing properties of neural networkscited by 5,155
  7. 2016Adversarial examples in the physical worldcited by 1,769
  8. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,318
  9. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 953
  10. 2017Ensemble Adversarial Training: Attacks and Defensescited by 840
  11. 2015The Limitations of Deep Learning in Adversarial Settingscited by 790
  12. 2016Universal adversarial perturbationscited by 784
23 more
  1. 2015Distillation as a Defense to Adversarial Perturbations against Deep Neural Networkscited by 678
  2. 2017ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Modelscited by 666
  3. 2017Mitigating Adversarial Effects Through Randomizationcited by 383
  4. 2018Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Modelscited by 362
  5. 2017DualGAN: Unsupervised Dual Learning for Image-to-Image Translationcited by 310
  6. 2017MagNet: a Two-Pronged Defense against Adversarial Examplescited by 269
  7. 2018Generating Adversarial Examples with Adversarial Networkscited by 258
  8. 2017On Detecting Adversarial Perturbationscited by 232
  9. 2016Coupled Generative Adversarial Networkscited by 216
  10. 2020A Review on Generative Adversarial Networks: Algorithms, Theory, and Applicationscited by 209
  11. 2018Stochastic Activation Pruning for Robust Adversarial Defensecited by 171
  12. 2018Towards the first adversarially robust neural network model on MNISTcited by 104
  13. 2018Deflecting Adversarial Attacks with Pixel Deflectioncited by 71
  14. 2015Learning with a Strong Adversarycited by 68
  15. 2017APE-GAN: Adversarial Perturbation Elimination with GANcited by 57
  16. 2017The Relative Performance of Ensemble Methods with Deep Convolutional Neural Networks for Image Classificationcited by 50
  17. 2016Simple Black-Box Adversarial Perturbations for Deep Networkscited by 42
  18. 2018Defending Against Universal Perturbations With Shared Adversarial Trainingcited by 25
  19. 2019GAT: Generative Adversarial Training for Adversarial Example Detection and Robust Classificationcited by 20
  20. 2019POBA-GA: Perturbation Optimized Black-Box Adversarial Attacks via Genetic Algorithmcited by 14
  21. 2018Facial Attributes: Accuracy and Adversarial Robustnesscited by 8
  22. 2020ROBY: Evaluating the Robustness of a Deep Model by its Decision Boundariescited by 4
  23. 2019Spot Evasion Attacks: Adversarial Examples for License Plate Recognition Systems with Convolutional Neural Networkscited by 3

Built on by

Not cited yet.