Abstract: Deep learning (DL) has received a lot of attention in hyperspectral anomaly detection (HAD) in recent years. While some progress has been made in boosting the generalization of DL-based HAD ...