Corpus ID: 204402740

Verification of Neural Networks: Specifying Global Robustness using Generative Models

@article{Fijalkow2019VerificationON,
  title={Verification of Neural Networks: Specifying Global Robustness using Generative Models},
  author={Nathanael Fijalkow and M. Gupta},
  journal={ArXiv},
  year={2019},
  volume={abs/1910.05018}
}
  • Nathanael Fijalkow, M. Gupta
  • Published 2019
  • Mathematics, Computer Science
  • ArXiv
  • The success of neural networks across most machine learning tasks and the persistence of adversarial examples have made the verification of such models an important quest. Several techniques have been successfully developed to verify robustness, and are now able to evaluate neural networks with thousands of nodes. The main weakness of this approach is in the specification: robustness is asserted on a validation set consisting of a finite set of examples, i.e. locally. We propose a notion of… CONTINUE READING

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 11 REFERENCES
    Safety Verification of Deep Neural Networks
    399
    Towards Deep Learning Models Resistant to Adversarial Attacks
    2233
    Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks
    630
    A Style-Based Generator Architecture for Generative Adversarial Networks
    1072
    Towards Fast Computation of Certified Robustness for ReLU Networks
    210
    EMNIST: an extension of MNIST to handwritten letters
    260
    Multi-column deep neural networks for image classification
    2690