[PDF][PDF] Early layers are more important for adversarial robustness

C Bakiskan, M Cekic, U Madhow - … 2022 Workshop on New Frontiers in …, 2022 - par.nsf.gov
Adversarial training and its variants have become the de facto standard for combatting
against adversarial attacks in machine learning models. In this paper, we seek insight into …

[PDF][PDF] Broad adversarial training with data augmentation in the output space

N Worzyk, S Yu - The AAAI-22 Workshop on Adversarial Machine …, 2021 - eecs.umich.edu
In image classification, data augmentation and the usage of additional data has been shown
to increase the efficiency of clean training and the accuracy of the resulting model. However …