没有免费午餐：异常检测中过度表达表征的危害
No Free Lunch: The Hazards of Over-Expressive Representations in Anomaly Detection
AI总结：
该研究针对“表征表达力越强异常检测性能越好”的假设，通过理论玩具模型和实证研究证明异常检测存在“没有免费午餐”效应，过度表达的表征会导致性能下降，并指出需引导表征聚焦相关属性，给出了未来缓解方向。
摘要：
近年来，深度学习驱动的异常检测方法取得了显著进展，这主要得益于表征能力的提升。人们很容易提出这样的假设：只要扩大网络规模、让表征更具表达力，异常检测的性能就能无限提升。在本文中，我们从理论和实证两方面提供了相反的证据。实际上，我们通过实验证明，当在研究充分的以物体为中心的数据集之外进行评估时，即使是简单的异常，表达力极强的表征也可能无法检测到。为研究这一现象，我们首先提出了一种用于分析异常检测性能的新型理论玩具模型。该模型揭示了表征充分性与过度表达性之间的根本权衡关系，为异常检测领域的“没有免费午餐”定理提供了依据：提升表征表达力最终会导致性能下降。相反，必须提供引导，让表征聚焦于与目标异常相关的属性。我们开展了广泛的实证研究，结果表明，当前最先进的表征往往存在过度表达问题，无法检测多种类型的异常。我们的研究证实了这种过度表达性在实际场景中如何损害图像异常检测性能。最后，我们提出了缓解该问题的未来研究方向。
英文摘要：
Anomaly detection methods, powered by deep learning, have recently been making significant progress, mostly due to improved representations. It is tempting to hypothesize that anomaly detection can improve indefinitely by increasing the scale of our networks, making their representations more expressive. In this paper, we provide theoretical and empirical evidence to the contrary. In fact, we empirically show cases where very expressive representations fail to detect even simple anomalies when evaluated beyond the well-studied object-centric datasets. To investigate this phenomenon, we begin by introducing a novel theoretical toy model for anomaly detection performance. The model uncovers a fundamental trade-off between representation sufficiency and over-expressivity. It provides evidence for a no-free-lunch theorem in anomaly detection stating that increasing representation expressivity will eventually result in performance degradation. Instead, guidance must be provided to focus the representation on the attributes relevant to the anomalies of interest. We conduct an extensive empirical investigation demonstrating that state-of-the-art representations often suffer from over-expressivity, failing to detect many types of anomalies. Our investigation demonstrates how this over-expressivity impairs image anomaly detection in practical settings. We conclude with future directions for mitigating this issue.
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。
