DExT: Detector Explanation Toolkit
This addresses the need for interpretability in object detectors for safety-critical applications, though it is incremental as it builds on existing explanation methods.
The paper tackles the problem of explaining object detector decisions by proposing DExT, an open-source toolkit that generates holistic explanations for bounding box and classification outputs using gradient-based methods, with evaluations showing SSD is more faithfully explained and SmoothGrad with Guided Backpropagation provides the most trustworthy explanations.
State-of-the-art object detectors are treated as black boxes due to their highly non-linear internal computations. Even with unprecedented advancements in detector performance, the inability to explain how their outputs are generated limits their use in safety-critical applications. Previous work fails to produce explanations for both bounding box and classification decisions, and generally make individual explanations for various detectors. In this paper, we propose an open-source Detector Explanation Toolkit (DExT) which implements the proposed approach to generate a holistic explanation for all detector decisions using certain gradient-based explanation methods. We suggests various multi-object visualization methods to merge the explanations of multiple objects detected in an image as well as the corresponding detections in a single image. The quantitative evaluation show that the Single Shot MultiBox Detector (SSD) is more faithfully explained compared to other detectors regardless of the explanation methods. Both quantitative and human-centric evaluations identify that SmoothGrad with Guided Backpropagation (GBP) provides more trustworthy explanations among selected methods across all detectors. We expect that DExT will motivate practitioners to evaluate object detectors from the interpretability perspective by explaining both bounding box and classification decisions.