Squeeze Training for Adversarial Robustness
This addresses a critical security issue in machine learning for applications like autonomous systems, though it appears incremental as it builds on existing adversarial training methods.
The paper tackles the vulnerability of deep neural networks to adversarial examples by introducing collaborative examples to enhance adversarial training, achieving new state-of-the-art results in adversarial robustness.
The vulnerability of deep neural networks (DNNs) to adversarial examples has attracted great attention in the machine learning community. The problem is related to non-flatness and non-smoothness of normally obtained loss landscapes. Training augmented with adversarial examples (a.k.a., adversarial training) is considered as an effective remedy. In this paper, we highlight that some collaborative examples, nearly perceptually indistinguishable from both adversarial and benign examples yet show extremely lower prediction loss, can be utilized to enhance adversarial training. A novel method is therefore proposed to achieve new state-of-the-arts in adversarial robustness. Code: https://github.com/qizhangli/ST-AT.