Global explainability in aligned image modalities

Justin Engelmann, Amos Storkey, Miguel O. Bernabeu

Deep learning (DL) models are very effective on many computer vision problems and increasingly used in critical applications. They are also inherently black box. A number of method…