1 paper · 1 filter
Yi Zhang, Zheng Wang, Zhen Chen +5
Deep learning models are notoriously vulnerable to imperceptible perturbations. Most existing research centers on adversarial robustness (AR), which evaluates models under worst-ca…