In safety-critical domains like automated driving (AD), errors by the object detector may endanger pedestrians and other vulnerable road users (VRU). As common evaluation metrics are not an adequate safety indicator, recent works employ approaches to identify safety-critical VRU and back-annotate the risk to the object detector. However, those approaches do not consider the safety factor in the deep neural network (DNN) training process. Thus, state-of-the-art DNN penalizes all misdetections equally irrespective of their criticality. Subsequently, to mitigate the occurrence of critical failure cases, i.e., false negatives, a safety-aware training strategy might be required to enhance the detection performance for critical pedestrians. In this paper, we propose a novel safety-aware loss variation that leverages the estimated per-pedestrian criticality scores during training. We exploit the reachability set-based time-to-collision (TTC-RSB) metric from the motion domain along with distance information to account for the worst-case threat quantifying the criticality. Our evaluation results using RetinaNet and FCOS on the nuScenes dataset demonstrate that training the models with our safety-aware loss function mitigates the misdetection of critical pedestrians without sacrificing performance for the general case, i.e., pedestrians outside the safety-critical zone.

翻译：在自动驾驶（AD）等安全关键领域，目标检测器的错误可能危及行人及其他弱势道路使用者（VRU）。由于常用评估指标不足以充分体现安全性，近期研究采用识别安全关键型VRU并将风险反向标注至目标检测器的方法。然而，这些方法未在深度神经网络（DNN）训练过程中考虑安全因子。因此，现有最优DNN对所有误检一视同仁地施加惩罚，而不区分其关键性。为缓解关键性故障（即假阴性）的发生，可能需要采用安全感知训练策略来提升对关键行人的检测性能。本文提出一种新颖的安全感知损失变体，在训练过程中利用每个行人的关键性评分。我们采用运动域中基于可达集碰撞时间（TTC-RSB）指标与距离信息相结合的方法，量化最坏情况下的威胁程度。基于RetinaNet和FCOS在nuScenes数据集上的评估结果表明，采用本文安全感知损失函数训练的模型在缓解关键行人误检的同时，并未牺牲一般场景（即安全关键区域外的行人）的检测性能。