ODExAI: A Comprehensive Object Detection Explainable AI Evaluation
This work addresses the problem of inconsistent evaluation for XAI in object detection, which hinders method comparison and selection for researchers and practitioners, though it is incremental as it builds on existing XAI techniques.
The paper tackled the lack of standards for evaluating explainable AI (XAI) methods in object detection by introducing ODExAI, a framework that benchmarks methods on localization, faithfulness, and complexity, revealing trade-offs such as region-based methods achieving high faithfulness (OA = 0.863) but high runtime (71.42s) and CAM-based methods offering superior localization (PG = 96.13%) with low runtime (0.54s) but lower faithfulness.
Explainable Artificial Intelligence (XAI) techniques for interpreting object detection models remain in an early stage, with no established standards for systematic evaluation. This absence of consensus hinders both the comparative analysis of methods and the informed selection of suitable approaches. To address this gap, we introduce the Object Detection Explainable AI Evaluation (ODExAI), a comprehensive framework designed to assess XAI methods in object detection based on three core dimensions: localization accuracy, faithfulness to model behavior, and computational complexity. We benchmark a set of XAI methods across two widely used object detectors (YOLOX and Faster R-CNN) and standard datasets (MS-COCO and PASCAL VOC). Empirical results demonstrate that region-based methods (e.g., D-CLOSE) achieve strong localization (PG = 88.49%) and high model faithfulness (OA = 0.863), though with substantial computational overhead (Time = 71.42s). On the other hand, CAM-based methods (e.g., G-CAME) achieve superior localization (PG = 96.13%) and significantly lower runtime (Time = 0.54s), but at the expense of reduced faithfulness (OA = 0.549). These findings demonstrate critical trade-offs among existing XAI approaches and reinforce the need for task-specific evaluation when deploying them in object detection pipelines. Our implementation and evaluation benchmarks are publicly available at: https://github.com/Analytics-Everywhere-Lab/odexai.