Deciphering the Definition of Adversarial Robustness for post-hoc OOD Detectors

arXiv:2406.15104v1 Announce Type: cross
Abstract: Detecting out-of-distribution (OOD) inputs is critical for safely deploying deep learning models in real-world scenarios. In recent years, many OOD detectors have been developed, and even the benchmarking has been standardized, i.e. OpenOOD. The number of post-hoc detectors is growing fast and showing an option to protect a pre-trained classifier against natural distribution shifts, claiming to be ready for real-world scenarios. However, its efficacy in handling adversarial examples has been neglected in the majority of …

abstract adversarial arxiv benchmarking cs.cr cs.cv deep learning definition deploying detectors distribution hoc inputs robustness type world