Existing explainable AI (XAI) methods for computer vision models often fail to capture the importance of spatial context, which can be crucial for accurate predictions in real-world applications.
