mlsanityData to 7 Oct 2026

Get Fooled for the Right Reason: Improving Adversarial Robustness through a Teacher-guided Curriculum Learning Approach

arXiv 2111.00295 ↗First posted 30 Oct 2021cs.LG cs.CR cs.CVcited by 4

Not on the map. Why?

Abstract (arXiv)

Current SOTA adversarially robust models are mostly based on adversarial training (AT) and differ only by some regularizers either at inner maximization or outer minimization steps. Being repetitive in nature during the inner maximization step, they take a huge time to train. We propose a non-iterative method that enforces the following ideas during training. Attribution maps are more aligned to the actual object in the image for adversarially robust models compared to naturally trained models. Also, the allowed set of pixels to perturb an image (that changes model decision) should be restricted to the object pixels only, which reduces the attack strength by limiting the attack space. Our method achieves significant performance gains with a little extra effort (10-20%) over existing AT models and outperforms all other methods in terms of adversarial as well as natural accuracy. We have performed extensive experimentation with CIFAR-10, CIFAR-100, and TinyImageNet datasets and reported results against many popular strong adversarial attacks to prove the effectiveness of our method.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 27 of its 33 references that are in the map. First posted on arXiv 30 Oct 2021.

Builds on

27 of 33 references.

References · 27

  1. 2014Explaining and Harnessing Adversarial Examplescited by 7,287
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  3. 2016Wide Residual Networkscited by 3,057
  4. 2016Towards Evaluating the Robustness of Neural Networkscited by 2,976
  5. 2019Theoretically Principled Trade-off between Robustness and Accuracycited by 1,510
  6. 2020Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attackscited by 1,442
  7. 2015DeepFool: a simple and accurate method to fool deep neural networkscited by 1,315
  8. 2016Adversarial Machine Learning at Scalecited by 934
  9. 2018Robustness May Be at Odds with Accuracycited by 873
  10. 2017Ensemble Adversarial Training: Attacks and Defensescited by 833
  11. 2020Fast is better than free: Revisiting adversarial trainingcited by 723
  12. 2019Adversarial Training for Free!cited by 638
15 more
  1. 2020On Adaptive Attacks to Adversarial Example Defensescited by 547
  2. 2018Feature Denoising for Improving Adversarial Robustnesscited by 312
  3. 2020Attacks Which Do Not Kill Training Make Adversarial Learning Strongercited by 242
  4. 2018Adversarial Logit Pairingcited by 224
  5. 2019Improving Adversarial Robustness via Promoting Ensemble Diversitycited by 213
  6. 2020Geometry-aware Instance-reweighted Adversarial Trainingcited by 198
  7. 2019Adversarial Robustness through Local Linearizationcited by 146
  8. 2018Curriculum Adversarial Trainingcited by 89
  9. 2019On the Connection Between Adversarial Robustness and Saliency Map Interpretabilitycited by 80
  10. 2019Instance adaptive adversarial training: Improved accuracy tradeoffs in neural netscited by 55
  11. 2019Jacobian Adversarially Regularized Networks for Robustnesscited by 40
  12. 2020Single-step Adversarial training with Dropout Schedulingcited by 40
  13. 2019What it Thinks is Important is Important: Robustness Transfers through Input Gradientscited by 34
  14. 2019Harnessing the Vulnerability of Latent Layers in Adversarially Trained Modelscited by 26
  15. 2019Enforcing Linearity in DNN succours Robustness and Adversarial Image Generationcited by 3

Built on by

Cited by year

Citing papers per year: 2022: 3; 2023: 1.

4 papers.

  1. 2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 126
  2. 2022Increasing Confidence in Adversarial Robustness Evaluationscited by 12
  3. 2022Learning Robust Kernel Ensembles with Kernel Average Poolingcited by 1
  4. 2023Topology-preserving Adversarial Training for Alleviating Natural Accuracy Degradationcited by 1