Competition on robust deep learning

Yinpeng Dong; Chang Liu; Wenzhao Xiang; Hang Su; Jun Zhu

doi:10.1093/nsr/nwad087

Competition on robust deep learning

Natl Sci Rev. 2023 Apr 7;10(6):nwad087. doi: 10.1093/nsr/nwad087. eCollection 2023 Jun.

Authors

Yinpeng Dong¹, Chang Liu², Wenzhao Xiang^{3

4}, Hang Su^{1

4

5}, Jun Zhu^{1

4

5}

Affiliations

¹ Department of Computer Science and Technology, Institute for AI, Tsinghua-Bosch Joint ML Center, THBI Lab, BNRist Center, Tsinghua University, China.
² Institute of Image Communication and Networks Engineering in the Department of Electronic Engineering (EE), Shanghai Jiao Tong University, China.
³ Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS), Institute of Computing Technology, CAS, China.
⁴ Peng Cheng Laboratory, China.
⁵ Pazhou Laboratory (Huangpu), China.

Abstract

This perspective paper proposes a new adversarial training method based on large-scale pre-trained models to achieve state-of-the-art adversarial robustness on ImageNet.

Keywords: adversarial example; adversarial training; deep learning; robustness.