Abstract: Many studies have demonstrated the vulnerability of Deep Neural Networks (DNNs) to adversarial attacks. While numerous research efforts have proposed high-performance adversarial attacks and ...