To operate in real-world high-stakes environments, deep learning systems have to endure
noises that have been continuously thwarting their robustness. Data-end defense, which …

Randomized smoothing is a general technique for computing sample-dependent robustness
guarantees against adversarial attacks for deep classifiers. Prior works on randomized …

Pretrained models from self-supervision are prevalently used in fine-tuning downstream
tasks faster or for better accuracy. However, gaining robustness from pretraining is left …

Training convolutional neural networks (CNNs) with a strict Lipschitz constraint under the $
l_ {2} $ norm is useful for provable adversarial robustness, interpretable gradients and …

Adversarial training (AT) for robust representation learning and self-supervised learning
(SSL) for unsupervised representation learning are two active research fields. Integrating AT …

We study the problem of training certifiably robust models against adversarial examples.
Certifiable training minimizes an upper bound on the worst-case loss over the allowed …

Adversarial training suffers from the issue of robust overfitting, which seriously impairs its
generalization performance. Data augmentation, which is effective at preventing overfitting …

Adversarial robustness has attracted extensive studies recently by revealing the vulnerability
and intrinsic characteristics of deep networks. However, existing works on adversarial …

Machine learning models are susceptible to adversarial perturbations: small changes to
input that can cause large changes in output. It is also demonstrated that there exist input …

Adversarial examples have pointed out Deep Neural Network's vulnerability to small local
noise. It has been shown that constraining their Lipschitz constant should enhance …