Nicholas Carlini, Guy Katz, Clark Barrett, David L. Dill
TL;DR利用形式验证技术构建对抗样本,证明这些样本是最小扭曲的,从而增加了对抗性训练的鲁棒性。
Abstract
The ability to deploy neural networks in real-world, safety-critical systems
is severely limited by the presence of adversarial examples: slightly perturbed
inputs that are misclassified by the network. In recent