English

Learning with Multiplicative Perturbations

Machine Learning 2020-06-24 v2 Machine Learning

Abstract

Adversarial Training (AT) and Virtual Adversarial Training (VAT) are the regularization techniques that train Deep Neural Networks (DNNs) with adversarial examples generated by adding small but worst-case perturbations to input examples. In this paper, we propose xAT and xVAT, new adversarial training algorithms, that generate \textbf{multiplicative} perturbations to input examples for robust training of DNNs. Such perturbations are much more perceptible and interpretable than their \textbf{additive} counterparts exploited by AT and VAT. Furthermore, the multiplicative perturbations can be generated transductively or inductively while the standard AT and VAT only support a transductive implementation. We conduct a series of experiments that analyze the behavior of the multiplicative perturbations and demonstrate that xAT and xVAT match or outperform state-of-the-art classification accuracies across multiple established benchmarks while being about 30\% faster than their additive counterparts. Furthermore, the resulting DNNs also demonstrate distinct weight distributions.

Keywords

Cite

@article{arxiv.1912.01810,
  title  = {Learning with Multiplicative Perturbations},
  author = {Xiulong Yang and Shihao Ji},
  journal= {arXiv preprint arXiv:1912.01810},
  year   = {2020}
}

Comments

Accepted as a conference paper at ICPR 2020

R2 v1 2026-06-23T12:35:13.388Z