mlsanityData to 7 Oct 2026

On the (Un-)Avoidability of Adversarial Examples

arXiv 2106.13326 ↗First posted 24 Jun 2021cs.LG stat.MLcited by 0

Not on the map. Why?

Abstract (arXiv)

The phenomenon of adversarial examples in deep learning models has caused substantial concern over their reliability. While many deep neural networks have shown impressive performance in terms of predictive accuracy, it has been shown that in many instances an imperceptible perturbation can falsely flip the network's prediction. Most research has then focused on developing defenses against adversarial attacks or learning under a worst-case adversarial loss. In this work, we take a step back and aim to provide a framework for determining whether a model's label change under small perturbation is justified (and when it is not). We carefully argue that adversarial robustness should be defined as a locally adaptive measure complying with the underlying distribution. We then suggest a definition for an adaptive robust loss, derive an empirical version of it, and develop a resulting data-augmentation framework. We prove that our adaptive data-augmentation maintains consistency of 1-nearest neighbor classification under deterministic labels and provide illustrative empirical evaluations.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 25 of its 42 references that are in the map. First posted on arXiv 24 Jun 2021.

Builds on

25 of 42 references.

References · 25

  1. 2013Intriguing properties of neural networkscited by 5,114
  2. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  3. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,324
  4. 2019Adversarial Examples Are Not Bugs, They Are Featurescited by 950
  5. 2018Robustness May Be at Odds with Accuracycited by 873
  6. 2018Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Surveycited by 459
  7. 2019On Evaluating Adversarial Robustnesscited by 442
  8. 2019Provably Robust Deep Learning via Adversarially Trained Smoothed Classifierscited by 328
  9. 2018Adversarially Robust Generalization Requires More Datacited by 318
  10. 2018Adversarial Attacks and Defences: A Surveycited by 259
  11. 2018Rademacher Complexity for Adversarially Robust Generalizationcited by 132
  12. 2019VC Classes are Adversarially Robustly Learnable, but Only Improperlycited by 83
13 more
  1. 2018Adversarial examples from computational constraintscited by 79
  2. 2014Rates of Convergence for Nearest Neighbor Classificationcited by 45
  3. 2017Analyzing the Robustness of Nearest Neighbors to Adversarial Examplescited by 36
  4. 2018Adversarial Risk and Robustness: General Definitions and Implications for the Uniform Distributioncited by 32
  5. 2019On the Hardness of Robust Classificationcited by 23
  6. 2020Efficiently Learning Adversarially Robust Halfspaces with Noisecited by 22
  7. 2020When are Non-Parametric Methods Robust?cited by 15
  8. 2019On Robustness to Adversarial Examples and Polynomial Optimizationcited by 14
  9. 2020Black-box Certification and Learning under Adversarial Perturbationscited by 13
  10. 2019Robustness for Non-Parametric Classification: A Generic Attack and Defensecited by 12
  11. 2018Sufficient Conditions for Idealised Models to Have No Adversarial Examples: a Theoretical and Empirical Study with Bayesian Neural Networkscited by 9
  12. 2019PDA: Progressive Data Augmentation for General Robustness of Deep Neural Networkscited by 5
  13. 2019Adversarial Training with Voronoi Constraintscited by 4

Built on by

Not cited yet.