Corpus ID: 6706414

Explaining and Harnessing Adversarial Examples

@article{Goodfellow2015ExplainingAH,
  title={Explaining and Harnessing Adversarial Examples},
  author={Ian J. Goodfellow and Jonathon Shlens and Christian Szegedy},
  journal={CoRR},
  year={2015},
  volume={abs/1412.6572}
}
  • Ian J. Goodfellow, Jonathon Shlens, Christian Szegedy
  • Published 2015
  • Computer Science, Mathematics
  • CoRR
  • Several machine learning models, including neural networks, consistently misclassify adversarial examples---inputs formed by applying small but intentionally worst-case perturbations to examples from the dataset, such that the perturbed input results in the model outputting an incorrect answer with high confidence. Early attempts at explaining this phenomenon focused on nonlinearity and overfitting. We argue instead that the primary cause of neural networks' vulnerability to adversarial… CONTINUE READING
    5,767 Citations
    Predicting Adversarial Examples with High Confidence
    • 6
    • Highly Influenced
    • PDF
    Hitting Depth : Investigating Robustness to Adversarial Examples in Deep Convolutional Neural Networks
    • Chris Billovits
    • 2016
    • 5
    • Highly Influenced
    • PDF
    Deep neural rejection against adversarial examples
    • 5
    • PDF
    Are Accuracy and Robustness Correlated
    • 37
    • PDF
    Intriguing Properties of Adversarial Examples
    • 44
    • PDF
    Harnessing Model Uncertainty for Detecting Adversarial Examples
    • 3
    • Highly Influenced
    • PDF
    Adversarial Examples on Object Recognition: A Comprehensive Survey
    • 3
    • Highly Influenced
    • PDF
    Principal Component Adversarial Example
    • 1
    • Highly Influenced
    Vulnerability of classifiers to evolutionary generated adversarial examples
    • 1
    Improving the Adversarial Robustness and Interpretability of Deep Neural Networks by Regularizing their Input Gradients
    • 215
    • Highly Influenced
    • PDF

    References

    SHOWING 1-10 OF 22 REFERENCES
    Towards Deep Neural Network Architectures Robust to Adversarial Examples
    • 468
    • PDF
    Intriguing properties of neural networks
    • 5,481
    • PDF
    Dropout: a simple way to prevent neural networks from overfitting
    • 20,000
    • Highly Influential
    • PDF
    Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
    • 1,732
    • Highly Influential
    • PDF
    Learning Multiple Layers of Features from Tiny Images
    • 9,339
    • Highly Influential
    • PDF
    Maxout Networks
    • 1,571
    • PDF
    Multi-Prediction Deep Boltzmann Machines
    • 105
    • PDF
    Visual Causal Feature Learning
    • 47
    • PDF
    Going deeper with convolutions
    • 20,823
    • PDF
    Large Scale Distributed Deep Networks
    • 2,433
    • PDF