Authors: N. Ferrante, L. Fanucci, F. Rossi, F. Terrosi, A. Bondavalli ABSTRACT To achieve confidence in safety-critical systems, requires among others to meet high requirements on online testing of computer systems, as dictated by safety standards…
Deep learning image classifiers often struggle with domain shift, leading to significant performance degradation in real-world applications. In this paper, we introduce our CROCODILE framework, showing how tools from causality can foster a model’s robustness to domain…
Deep learning image classifiers often struggle with domain shift, leading to significant performance degradation in real-world applications. In this paper, we introduce our CROCODILE framework, showing how tools from causality can foster a model’s robustness to domain…