图像分类中out-of-distribution检测训练目标的系统比较
A Systematic Comparison of Training Objectives for Out-of-Distribution Detection in Image Classification
AI总结：
本文系统比较了四种图像分类中out-of-distribution检测的训练目标，发现交叉熵损失在近端和远端out-of-distribution表现最一致，其他目标在特定设置中也可竞争。
摘要：
out-of-distribution检测在安全敏感应用中至关重要。尽管这一挑战已从各种视角得到解决，但训练目标对out-of-distribution行为的影响仍相对研究不足。本文提出了一种系统比较四种广泛使用的训练目标：交叉熵损失、原型损失、三元组损失和平均精度（AP）损失，涵盖概率、原型基、度量学习和排序监督，用于图像分类中的out-of-distribution检测，遵循标准化的OpenOOD协议。在CIFAR-10/100和ImageNet-200上，我们发现交叉熵损失、原型损失和AP损失在in-distribution准确性上相当，而交叉熵损失在近端和远端out-of-distribution表现上最为一致；其他目标在特定设置中可以具有竞争力。
英文摘要：
Out-of-distribution (OOD) detection is critical in safety-sensitive applications. While this challenge has been addressed from various perspectives, the influence of training objectives on OOD behavior remains comparatively underexplored. In this paper, we present a systematic comparison of four widely used training objectives: Cross-Entropy Loss, Prototype Loss, Triplet Loss, and Average Precision (AP) Loss, spanning probabilistic, prototype-based, metric-learning, and ranking-based supervision, for OOD detection in image classification under standardized OpenOOD protocols. Within the evaluated ResNet-18/OpenOOD setting and objective-specific OOD scoring rules, Cross-Entropy Loss, Prototype Loss, and AP Loss achieve comparable in-distribution accuracy, while Cross-Entropy Loss provides the most consistent near- and far-OOD AUROC overall; the other objectives can be competitive in specific settings.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
