Mixup Inference: Better Exploiting Mixup to Defend Adversarial Attacks.

Tianyu Pang, Kun Xu, Jun Zhu

Browse the full ICLR paper archive.