English

Built-in Vulnerabilities to Imperceptible Adversarial Perturbations

Computer Vision and Pattern Recognition 2019-05-09 v2 Machine Learning Machine Learning

Abstract

Designing models that are robust to small adversarial perturbations of their inputs has proven remarkably difficult. In this work we show that the reverse problem---making models more vulnerable---is surprisingly easy. After presenting some proofs of concept on MNIST, we introduce a generic tilting attack that injects vulnerabilities into the linear layers of pre-trained networks by increasing their sensitivity to components of low variance in the training data without affecting their performance on test data. We illustrate this attack on a multilayer perceptron trained on SVHN and use it to design a stand-alone adversarial module which we call a steganogram decoder. Finally, we show on CIFAR-10 that a poisoning attack with a poisoning rate as low as 0.1% can induce vulnerabilities to chosen imperceptible backdoor signals in state-of-the-art networks. Beyond their practical implications, these different results shed new light on the nature of the adversarial example phenomenon.

Keywords

Cite

@article{arxiv.1806.07409,
  title  = {Built-in Vulnerabilities to Imperceptible Adversarial Perturbations},
  author = {Thomas Tanay and Jerone T. A. Andrews and Lewis D. Griffin},
  journal= {arXiv preprint arXiv:1806.07409},
  year   = {2019}
}
R2 v1 2026-06-23T02:35:09.654Z