Corpus ID: 220525387

Fast Differentiable Clipping-Aware Normalization and Rescaling

@article{Rauber2020FastDC,
  title={Fast Differentiable Clipping-Aware Normalization and Rescaling},
  author={Jonas Rauber and M. Bethge},
  journal={ArXiv},
  year={2020},
  volume={abs/2007.07677}
}
  • Jonas Rauber, M. Bethge
  • Published 2020
  • Computer Science, Mathematics
  • ArXiv
  • Rescaling a vector ~δ ∈ Rn to a desired length is a common operation in many areas such as data science and machine learning. When the rescaled perturbation η~δ is added to a starting point ~x ∈ D (where D is the data domain, e.g. D = [0, 1]n), the resulting vector ~v = ~x + η~δ will in general not be in D. To enforce that the perturbed vector v is in D, the values of ~v can be clipped to D. This subsequent elementwise clipping to the data domain does however reduce the effective perturbation… CONTINUE READING
    EagerPy: Writing Code That Works Natively with PyTorch, TensorFlow, JAX, and NumPy
    1

    References

    Publications referenced by this paper.
    SHOWING 1-8 OF 8 REFERENCES
    Increasing the robustness of DNNs against image corruptions by playing the Game of Noise
    12
    TensorFlow Eager: A Multi-Stage, Python-Embedded DSL for Machine Learning
    23
    PyTorch: An Imperative Style, High-Performance Deep Learning Library
    1818
    EagerPy: Writing Code That Works Natively with PyTorch, TensorFlow, JAX, and NumPy
    1
    Foolbox Native: A Python toolbox to create adversarial examples that fool neural networks in PyTorch, TensorFlow, and JAX
    • 2020
    JAX: composable transformations of Python+NumPy programs
    • 2018
    NumPy: A guide to NumPy
    • 2006