Improved Surprise Adequacy Tools for Corner Case Data Description and Detection

Facing the increasing quantity of AI models applications, especially in life- and property-related fields, it is crucial for designers to construct safety- security-critical systems. As a major factor affecting safety models, corner case data its related description/detection techniques are important design phase quality assurance. In this paper, inspired by surprise adequacy (SA), tool having advantages on capture behaviors, we developed three modified versions distance-based-SA (DSA) detecting cases classification problems. Through experiment analysis MNIST, CIFAR, industrial example data, feasibility usefulness proposed tools detection verified. Moreover, Qualitative quantitative experiments validated that DSA can achieve improved performance describing cases’ behaviors.