Detecting and rejecting unknown out-of-distribution (OOD) samples is critical for deployed neural networks to void unreliable predictions. In real-world scenarios, however, the efficacy of existing OOD detection methods is often impeded by the inherent imbalance of in-distribution (ID) data, which causes significant performance decline. Through statistical observations, we have identified two common challenges faced by different OOD detectors: misidentifying tail class ID samples as OOD, while erroneously predicting OOD samples as head class from ID. To explain this phenomenon, we introduce a generalized statistical framework, termed ImOOD, to formulate the OOD detection problem on imbalanced data distribution. Consequently, the theoretical analysis reveals that there exists a class-aware bias item between balanced and imbalanced OOD detection, which contributes to the performance gap. Building upon this finding, we present a unified training-time regularization technique to mitigate the bias and boost imbalanced OOD detectors across architecture designs. Our theoretically grounded method translates into consistent improvements on the representative CIFAR10-LT, CIFAR100-LT, and ImageNet-LT benchmarks against several state-of-the-art OOD detection approaches. Code is available at https://github.com/alibaba/imood.

翻译：检测并拒识未知的分布外（OOD）样本对于已部署的神经网络避免不可靠预测至关重要。然而，在实际场景中，现有OOD检测方法的有效性常受到分布内（ID）数据固有非平衡性的阻碍，导致性能显著下降。通过统计观察，我们识别了不同OOD检测器面临的两个常见挑战：将尾部类别的ID样本误判为OOD，同时错误地将OOD样本预测为ID中的头部类别。为解释此现象，我们引入了一个广义统计框架（称为ImOOD），以形式化非平衡数据分布上的OOD检测问题。相应的理论分析表明，在平衡与非平衡OOD检测之间存在一个类别感知的偏差项，该偏差导致了性能差距。基于这一发现，我们提出了一种统一的训练时正则化技术，以减轻该偏差并提升跨架构设计的非平衡OOD检测器性能。我们这种具有理论依据的方法，在具有代表性的CIFAR10-LT、CIFAR100-LT和ImageNet-LT基准测试中，相较于多种先进的OOD检测方法，均实现了持续的性能提升。代码发布于https://github.com/alibaba/imood。