mlsanityData to 7 Oct 2026

Fight Perturbations with Perturbations: Defending Adversarial Attacks via Neuron Influence

arXiv 2112.13060 ↗First posted 24 Dec 2021cs.CV cs.AI cs.CRcited by 0

Not on the map. Why?

Abstract (arXiv)

The vulnerabilities of deep learning models towards adversarial attacks have attracted increasing attention, especially when models are deployed in security-critical domains. Numerous defense methods, including reactive and proactive ones, have been proposed for model robustness improvement. Reactive defenses, such as conducting transformations to remove perturbations, usually fail to handle large perturbations. The proactive defenses that involve retraining, suffer from the attack dependency and high computation cost. In this paper, we consider defense methods from the general effect of adversarial attacks that take on neurons inside the model. We introduce the concept of neuron influence, which can quantitatively measure neurons' contribution to correct classification. Then, we observe that almost all attacks fool the model by suppressing neurons with larger influence and enhancing those with smaller influence. Based on this, we propose Neuron-level Inverse Perturbation (NIP), a novel defense against general adversarial attacks. It calculates neuron influence from benign examples and then modifies input examples by generating inverse perturbations that can in turn strengthen neurons with larger influence and weaken those with smaller influence.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 36 of its 57 references that are in the map. First posted on arXiv 24 Dec 2021.

Builds on

36 of 57 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  3. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  4. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  5. 2016Densely Connected Convolutional Networkscited by 6,691
  6. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  7. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,827
  8. 2016Adversarial examples in the physical worldcited by 1,762
  9. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  10. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,315
  11. 2016SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model sizecited by 1,054
  12. 2015The Limitations of Deep Learning in Adversarial Settingscited by 784
24 more
  1. 2016Universal adversarial perturbationscited by 774
  2. 2019Square Attack: a query-efficient black-box adversarial attack via random searchcited by 682
  3. 2015Distillation as a Defense to Adversarial Perturbations against Deep Neural Networkscited by 675
  4. 2017One pixel attack for fooling deep neural networkscited by 579
  5. 2018Fine-Pruning: Defending Against Backdooring Attacks on Deep Neural Networkscited by 550
  6. 2017Countering Adversarial Images using Input Transformationscited by 517
  7. 2017Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networkscited by 484
  8. 2017Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Modelscited by 459
  9. 2018Feature Denoising for Improving Adversarial Robustnesscited by 312
  10. 2021Adversarial Neuron Pruning Purifies Backdoored Deep Modelscited by 230
  11. 2016A study of the effect of JPG compression on adversarial imagescited by 196
  12. 2018Stochastic Activation Pruning for Robust Adversarial Defensecited by 171
  13. 2018Towards the first adversarially robust neural network model on MNISTcited by 104
  14. 2018Feature Distillation: DNN-Oriented JPEG Compression Against Adversarial Examplescited by 98
  15. 2021Improving Adversarial Robustness via Channel-wise Activation Suppressingcited by 76
  16. 2019Image Super-Resolution as a Defense Against Adversarial Attackscited by 64
  17. 2017Deep Speaker: an End-to-End Neural Speaker Embedding Systemcited by 56
  18. 2017Detecting Adversarial Image Examples in Deep Networks with Adaptive Noise Reductioncited by 49
  19. 2018Characterizing Audio Adversarial Examples Using Temporal Dependencycited by 34
  20. 2018Adversarial Defense by Stratified Convolutional Sparse Codingcited by 31
  21. 2021Attack as Defense: Characterizing Adversarial Examples using Robustnesscited by 14
  22. 2018Facial Attributes: Accuracy and Adversarial Robustnesscited by 8
  23. 2020MP3 Compression To Diminish Adversarial Noise in End-to-End Speech Recognitioncited by 5
  24. 2020One Neuron to Fool Them Allcited by 2

Built on by

Not cited yet.