域偏移下行人检测的置信度受控XAI审计
Confidence-Controlled XAI Auditing for Pedestrian Detection under Domain Shift
浏览论文内容
中文总结 AI 辅助
本研究审计域偏移下YOLOv8s行人检测器的可解释性，发现D-RISE忠实度与检测强度强耦合，控制后仍存在域差异，表明需置信度受控的XAI审计。
中文摘要 AI 辅助
可解释性在智能车辆感知模型中的需求日益增长，然而在驾驶域偏移下解释是否仍然忠实仍鲜为人知。本研究对固定的YOLOv8s行人检测器在PIE和JAAD数据集上的事后解释进行了审计，采用基于ROI的D-Deletion、冻结置信度三分位数、秩次检验、bootstrap区间和Holm校正。审计显示，基于删除的忠实度与解释时的检测强度强耦合，D-RISE的Spearman相关系数在0.70至0.82之间，使得朴素的置信度分层比较不可靠。在固定f0分箱内控制检测强度后，D-RISE的忠实度在中心f0范围内仍具有域依赖性，PIE的D-Deletion高于JAAD，且经Holm调整后显著。非扰动性EigenCAM基线比D-RISE更不忠实，但也表现出分数耦合，表明该效应并非D-RISE特有，而是与基于删除的评估设置相关。这些结果促使在域偏移下对安全关键感知进行置信度受控的XAI审计。
英文摘要
Explainability is increasingly required for perception models in intelligent vehicles, yet whether explanations remain faithful under driving domain shift is still poorly understood. This work audits post-hoc explanations of a fixed YOLOv8s pedestrian detector across PIE and JAAD using ROI-based D-Deletion, frozen confidence terciles, rank-based tests, bootstrap intervals, and Holm correction. The audit shows that deletion-based faithfulness is strongly coupled to detection strength at explanation time, with Spearman correlations between 0.70 and 0.82 for D-RISE, making naive confidence-stratified comparisons unreliable. After controlling for detection strength within fixed f0 bins, D-RISE faithfulness remains domain-dependent in the central f0 range, with PIE showing higher D-Deletion than JAAD and Holm-adjusted significance. A non-perturbative EigenCAM baseline is less faithful than D-RISE but also exhibits score coupling, suggesting that the effect is not specific to D-RISE and is related to the deletion-based evaluation setup. These results motivate confidence-controlled XAI audits for safety-critical perception under domain shift.
发表机构
- Universidad Nacional de San Agustín de Arequipa（阿雷基帕圣奥古斯丁国立大学）
机构由 AI 辅助整理，请以论文原文为准。
