Corpus ID: 3488815

Towards Deep Learning Models Resistant to Adversarial Attacks

  title={Towards Deep Learning Models Resistant to Adversarial Attacks},
  author={A. Madry and Aleksandar Makelov and L. Schmidt and D. Tsipras and Adrian Vladu},
  • A. Madry, Aleksandar Makelov, +2 authors Adrian Vladu
  • Published 2018
  • Computer Science, Mathematics
  • ArXiv
  • Recent work has demonstrated that neural networks are vulnerable to adversarial examples, i.e., inputs that are almost indistinguishable from natural data and yet classified incorrectly by the network. [...] Key Method Its principled nature also enables us to identify methods for both training and attacking neural networks that are reliable and, in a certain sense, universal. In particular, they specify a concrete security guarantee that would protect against any adversary.Expand Abstract
    2,905 Citations
    Adversarial Robustness Against the Union of Multiple Perturbation Models
    • 19
    • Highly Influenced
    • PDF
    Divide-and-Conquer Adversarial Detection
    • 3
    • Highly Influenced
    Hardening Deep Neural Networks via Adversarial Model Cascades
    • 7
    • PDF
    Towards Natural Robustness Against Adversarial Examples
    • Highly Influenced
    • PDF
    On the Connection between Differential Privacy and Adversarial Robustness in Machine Learning
    • 13
    • Highly Influenced
    Defending Against Adversarial Attacks Using Random Forests
    • PDF
    Learning to Disentangle Robust and Vulnerable Features for Adversarial Detection
    • Highly Influenced
    • PDF
    Defending Against Adversarial Attacks Using Random Forest
    • 1
    • PDF
    Defending Against Adversarial Samples Without Security through Obscurity
    • 2
    • PDF


    Towards Evaluating the Robustness of Neural Networks
    • 3,044
    • PDF
    Distillation as a Defense to Adversarial Perturbations Against Deep Neural Networks
    • 1,566
    • PDF
    Ground-Truth Adversarial Examples
    • 60
    • PDF
    The Limitations of Deep Learning in Adversarial Settings
    • 1,908
    • PDF
    Towards Deep Neural Network Architectures Robust to Adversarial Examples
    • 497
    • PDF
    Towards Robust Deep Neural Networks with BANG
    • 48
    • PDF
    Ensemble Adversarial Training: Attacks and Defenses
    • 1,133
    • PDF
    Adversarial Machine Learning at Scale
    • 1,266
    • Highly Influential
    • PDF
    Towards the first adversarially robust neural network model on MNIST
    • 173
    • PDF