mlsanityData to 7 Oct 2026

Real-time Detection of Practical Universal Adversarial Perturbations

arXiv 2105.07334 ↗First posted 16 May 2021cs.LG cs.AI cs.CRcited by 6

Not on the map. Why?

Abstract (arXiv)

Universal Adversarial Perturbations (UAPs) are a prominent class of adversarial examples that exploit the systemic vulnerabilities and enable physically realizable and robust attacks against Deep Neural Networks (DNNs). UAPs generalize across many different inputs; this leads to realistic and effective attacks that can be applied at scale. In this paper we propose HyperNeuron, an efficient and scalable algorithm that allows for the real-time detection of UAPs by identifying suspicious neuron hyper-activations. Our results show the effectiveness of HyperNeuron on multiple tasks (image classification, object detection), against a wide variety of universal attacks, and in realistic scenarios, like perceptual ad-blocking and adversarial patches. HyperNeuron is able to simultaneously detect both adversarial mask and patch UAPs with comparable or better performance than existing UAP defenses whilst introducing a significantly reduced latency of only 0.86 milliseconds per image. This suggests that many realistic and practical universal attacks can be reliably mitigated in real-time, which shows promise for the robust deployment of machine learning systems.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 36 of its 45 references that are in the map. First posted on arXiv 16 May 2021.

Builds on

36 of 45 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. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  6. 2013Intriguing properties of neural networkscited by 5,114
  7. 2015You Only Look Once: Unified, Real-Time Object Detectioncited by 4,829
  8. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,430
  9. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 950
  10. 2016Practical Black-Box Attacks against Machine Learningcited by 893
  11. 2016Universal adversarial perturbationscited by 774
  12. 2020Fast is better than free: Revisiting adversarial trainingcited by 723
24 more
  1. 2017Evasion Attacks against Machine Learning at Test Timecited by 686
  2. 2017Adversarial Patchcited by 514
  3. 2016Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samplescited by 439
  4. 2019Fooling automated surveillance cameras: adversarial patches to attack person detectioncited by 268
  5. 2019Adversarial Training and Robustness for Multiple Perturbationscited by 182
  6. 2018Physical Adversarial Examples for Object Detectorscited by 163
  7. 2018SentiNet: Detecting Localized Universal Attacks Against Deep Learning Systemscited by 148
  8. 2018DPatch: An Adversarial Patch Attack on Object Detectorscited by 146
  9. 2020Physically Realizable Adversarial Examples for LiDAR Object Detectioncited by 137
  10. 2018Generalizable Data-free Objective for Crafting Universal Adversarial Perturbationscited by 88
  11. 2018Universal Adversarial Trainingcited by 85
  12. 2017Fast Feature Fool: A data independent approach to universal adversarial perturbationscited by 65
  13. 2019Adversarial Objects Against LiDAR-Based Autonomous Driving Systemscited by 62
  14. 2017Art of singular vectors and universal adversarial perturbationscited by 49
  15. 2017Defense against Universal Adversarial Perturbationscited by 48
  16. 2018On the Structural Sensitivity of Deep Convolutional Networks to the Directions of Fourier Basis Functionscited by 31
  17. 2018Defending Against Universal Perturbations With Shared Adversarial Trainingcited by 25
  18. 2019Transfer of Adversarial Robustness Between Perturbation Typescited by 25
  19. 2018Procedural Noise Adversarial Examples for Black-Box Attacks on Deep Convolutional Networkscited by 21
  20. 2021Object Removal Attacks on LiDAR-based 3D Object Detectorscited by 15
  21. 2018AdVersarial: Perceptual Ad Blocking meets Adversarial Machine Learningcited by 14
  22. 2019Universal Adversarial Robustness of Texture and Shape-Biased Modelscited by 8
  23. 2021Jacobian Regularization for Mitigating Universal Adversarial Perturbationscited by 4
  24. 2020Shadow-Catcher: Looking Into Shadows to Detect Ghost Objects in Autonomous Vehicle 3D Sensingcited by 2

Built on by

Cited by year

Citing papers per year: 2022: 4; 2023: 1; 2024: 0; 2025: 1.

6 papers.

  1. 2022On the Real-World Adversarial Robustness of Real-Time Semantic Segmentation Models for Autonomous Drivingcited by 30
  2. 2022Adversarial Patch Attacks and Defences in Vision-Based Tasks: A Surveycited by 17
  3. 2022Defending From Physically-Realizable Adversarial Attacks Through Internal Over-Activation Analysiscited by 10
  4. 2022CARLA-GeAR: a Dataset Generator for a Systematic Evaluation of Adversarial Robustness of Vision Modelscited by 8
  5. 2023Attention-Based Real-Time Defenses for Physical Adversarial Attacks in Vision Applicationscited by 3
  6. 2025SpaNN: Detecting Multiple Adversarial Patches on CNNs by Spanning Saliency Thresholdscited by 0