mlsanityData to 14 Oct 2026

Training on Foveated Images Improves Robustness to Adversarial Attacks

arXiv 2308.00854 ↗First posted 1 Aug 2023cs.CV cs.AIcited by 5

Not on the map. Why?

Abstract (arXiv)

Deep neural networks (DNNs) have been shown to be vulnerable to adversarial attacks -- subtle, perceptually indistinguishable perturbations of inputs that change the response of the model. In the context of vision, we hypothesize that an important contributor to the robustness of human visual perception is constant exposure to low-fidelity visual stimuli in our peripheral vision. To investigate this hypothesis, we develop RBlur, an image transform that simulates the loss in fidelity of peripheral vision by blurring the image and reducing its color saturation based on the distance from a given fixation point. We show that compared to DNNs trained on the original images, DNNs trained on images transformed by RBlur are substantially more robust to adversarial attacks, as well as other, non-adversarial, corruptions, achieving up to 25% higher accuracy on perturbed data.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 36 of its 56 references that are in the map. First posted on arXiv 1 Aug 2023.

Builds on

36 of 56 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  3. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  4. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
  5. 2013Intriguing properties of neural networkscited by 5,155
  6. 2016Wide Residual Networkscited by 3,070
  7. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,877
  8. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,517
  9. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,451
  10. 2021MLP-Mixer: An all-MLP Architecture for Visioncited by 1,446
  11. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,340
  12. 2018Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examplescited by 1,268
24 more
  1. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 953
  2. 2018AutoAugment: Learning Augmentation Policies from Datacited by 730
  3. 2020Fast is better than free: Revisiting adversarial trainingcited by 727
  4. 2017Countering Adversarial Images using Input Transformationscited by 520
  5. 2019On Evaluating Adversarial Robustnesscited by 447
  6. 2021Recent Advances in Adversarial Training for Adversarial Robustnesscited by 279
  7. 2018Generalisation in humans and deep neural networkscited by 237
  8. 2021Data Augmentation Can Improve Robustnesscited by 194
  9. 2020fastai: A Layered API for Deep Learningcited by 183
  10. 2018Certified Adversarial Robustness with Additive Noisecited by 180
  11. 2022(Certified!!) Adversarial Robustness for Free!cited by 132
  12. 2017A Study and Comparison of Human and Deep Learning Recognition Performance Under Visual Distortionscited by 108
  13. 2018Towards the first adversarially robust neural network model on MNISTcited by 104
  14. 2020Certified Defense to Image Transformations via Randomized Smoothingcited by 51
  15. 2014Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognitioncited by 50
  16. 2015Foveation-based Mechanisms Alleviate Adversarial Examplescited by 32
  17. 2020Biologically Inspired Mechanisms for Adversarial Robustnesscited by 29
  18. 2021FoveaTer: Foveated Transformer for Image Classificationcited by 28
  19. 2020Emergent Properties of Foveated Perceptual Systemscited by 24
  20. 2017Attacking the Madry Defense Model with L1-based Adversarial Examplescited by 21
  21. 2021Center Smoothing: Certified Robustness for Networks with Structured Outputscited by 18
  22. 2022Finding Biological Plausibility for Adversarially Robust Features via Metameric Taskscited by 11
  23. 2017Central and peripheral vision for scene recognition: A neurocomputational modeling explorationcited by 8
  24. 2018Human peripheral blur is optimal for object recognitioncited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 2; 2025: 2; 2026: 1.

5 papers.

  1. 2024Semantic-Syntactic Discrepancy in Images (SSDI): Learning Meaning and Order of Features from Natural Imagescited by 1
  2. 2024On Inherent Adversarial Robustness of Active Vision Systemscited by 1
  3. 2025Two out of Three (ToT): using self-consistency to make robust predictionscited by 0
  4. 2025SAFER-AiD: Saccade-Assisted Foveal-peripheral vision Enhanced Reconstruction for Adversarial Defensecited by 0
  5. 2026FAVE: Foveated Adaptive Visual Encoding for Efficient Fine-Grained Visual Understandingcited by 0