Doubly Robust Instance-Reweighted Adversarial Training.

Daouda Sow, Sen Lin, Zhangyang Wang, Yingbin Liang

Browse the full ICLR paper archive.