基于训练数据的图像离群点合成
Image-based Outlier Synthesis With Training Data
提出一种无需外部数据的虚拟离群点合成方法：通过向ID输入添加梯度归因值来破坏不变特征并放大真实类logit，生成近流形离群点，可用于提升细粒度分类和虚假相关场景下的OOD检测鲁棒性。
原文依据 we synthesize virtual outliers from ID data by approximating the destruction of invariant features
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论分布外（OOD）检测对于确保深度学习模型在关键应用中的安全部署至关重要。深度学习模型常常会将OOD样本误判为分布内（ID）样本。在训练集存在虚假相关时，这种脆弱性会进一步加剧。同样，在细粒度分类场景中，由于细粒度OOD样本与ID样本高度相似，检测它们本身就具有挑战性。然而，当前OOD检测研究主要关注相对容易的（常规）情况。即使近期少数针对这些挑战性场景的工作也依赖于精心整理或合成的离群点，最终需要外部数据。这促使我们提出核心研究问题：“我们能否创新OOD检测训练框架，以应对细粒度和虚假相关场景，且完全不需要任何外部数据？”在本工作中，我们提出了一种统一方法ASCOOD，用于虚假相关、细粒度和常规OOD检测，消除了对外部数据的依赖。首先，我们通过近似破坏不变特征，从ID数据中合成虚拟离群点。具体而言，我们提出向ID输入添加梯度归因值，以破坏不变特征同时放大真实类logit，从而合成具有挑战性的近流形虚拟离群点。然后，我们同时激励ID分类和对虚拟离群点的预测不确定性。为此，我们进一步提出利用z-score归一化的标准化特征。ASCOOD有效缓解了虚假相关的影响，并鼓励捕获细粒度属性。在7个数据集上的大量实验以及与30多种方法的比较表明，ASCOOD在虚假相关、细粒度和常规场景中均具有优势。
英文原文摘要 1841 字符
Out-of-distribution (OOD) detection is critical to ensure the safe deployment of deep learning models in critical applications. Deep learning models can often misidentify OOD samples as in-distribution (ID) samples. This vulnerability worsens in the presence of spurious correlation in the training set. Likewise, in fine-grained classification settings, detection of fine-grained OOD samples becomes inherently challenging due to their high similarity to ID samples. However, current research on OOD detection has focused instead largely on relatively easier (conventional) cases. Even the few recent works addressing these challenging cases rely on carefully curated or synthesized outliers, ultimately requiring external data. This motivates our central research question: "Can we innovate OOD detection training framework for fine-grained and spurious settings without requiring any external data at all?" In this work, we present a unified Approach to Spurious, fine-grained, and Conventional OOD Detection (\ASCOOD) that eliminates the reliance on external data. First, we synthesize virtual outliers from ID data by approximating the destruction of invariant features. Specifically, we propose to add gradient attribution values to ID inputs to disrupt invariant features while amplifying true-class logit, thereby synthesizing challenging near-manifold virtual outliers. Then, we simultaneously incentivize ID classification and predictive uncertainty towards virtual outliers. For this, we further propose to leverage standardized features with z-score normalization. ASCOOD effectively mitigates impact of spurious correlations and encourages capturing fine-grained attributes. Extensive experiments across 7 datasets and comparisons with 30+ methods demonstrate merit of ASCOOD in spurious, fine-grained and conventional settings.