Explainable AI (XAI) for Object Detection and Application to Satellite Imagery

Explainability is critical to the wider deployment of deep learning models. And especially so in view of the worldwide effort to regulate AI with laws that impose a requirement for transparency in AI systems. In this paper, we present a new explainable modeling approach for object detection that would generally apply to classification problems as well. In this modeling approach, the prediction of an object in an image depends on explicit verification of some of its parts as defined by users. For example, the fuselage, tail, wings, and engines could be airplane parts that need to be verified. Identification of meaningful parts of objects by users is at the core of this form of explainability. User involvement in parts identification implies that these parts have meaning to them. In this approach, we reverse the process of creating explainable models. This approach is similar to an architect first consulting with a client to determine the needs before designing and building a house. Thus, the definition of explanation comes first, then comes model creation based on that definition. Defense Advanced Research Projects Agency (DARPA) originally proposed this type of explicit, symbolic form of explainability when it started the Explainable AI (XAI) program in 2015. We apply this DARPA idea of XAI to detect objects in satellite imagery. We tested our XAI object detection models on some publicly available satellite imagery to find ships and airplanes. Our experiments show that these explainable models can provide higher accuracy and a higher level of protection against adversarial attacks.