ShortcutBreaker: 多类无监督异常检测中的低秩噪声瓶颈与频率滤波块
ShortcutBreaker: Low-Rank Noisy Bottleneck and Frequency Filtering Block for Multi-Class Unsupervised Anomaly Detection
AI总结:
本文提出ShortcutBreaker框架,通过低秩噪声瓶颈和全局扰动注意力解决多类无监督异常检测中的身份捷径问题,实验表明其在四个数据集上均取得优异的AUROC性能。
摘要:
多类无监督异常检测(MUAD)近年来受到越来越多的关注,因为它旨在开发一个统一的模型来处理多个类别的异常检测,即消除为不同对象训练单独模型的需要,从而节省大量计算资源。在MUAD设置下,尽管先进的基于Transformer的架构带来了显著的性能提升,但身份捷径仍然存在:它们直接将输入复制到输出,缩小了正常和异常案例之间的重建误差差距,从而使得两者更难区分。因此,我们提出了ShortcutBreaker,一种新的统一特征重建框架用于MUAD任务,具有两个关键创新来解决捷径问题。首先,利用矩阵秩不等式,我们设计了一个低秩噪声瓶颈(LRNB)将高维特征投影到低秩潜在空间,并理论上证明其能够防止平凡的身份重现。其次,利用ViTs的全局建模能力而不是仅仅关注局部特征,我们引入了一个全局扰动注意力来防止解码器中的信息捷径。在四个广泛使用的异常检测基准上进行了大量实验,包括三个工业数据集(MVTec-AD、ViSA和Real-IAD)和一个医疗数据集(Universal Medical)。所提出的方法在这些四个数据集上分别取得了99.8%、98.9%、90.6%和87.8%的图像级AUROC,一致优于之前的MUAD方法在不同场景下的表现。我们的代码将被发布。
英文摘要:
Multi-class unsupervised anomaly detection (MUAD) has garnered growing research interest, as it seeks to develop a unified model for anomaly detection across multiple classes, i.e., eliminating the need to train separate models for distinct objects and thereby saving substantial computational resources. Under the MUAD setting, while advanced Transformer-based architectures have brought significant performance improvements, identity shortcuts persist: they directly copy inputs to outputs, narrowing the gap in reconstruction errors between normal and abnormal cases, and thereby making the two harder to distinguish. Therefore, we propose ShortcutBreaker, a novel unified feature-reconstruction framework for MUAD tasks, featuring two key innovations to address the issue of shortcuts. First, drawing on matrix rank inequality, we design a low-rank noisy bottleneck (LRNB) to project highdimensional features into a low-rank latent space, and theoretically demonstrate its capacity to prevent trivial identity reproduction. Second, leveraging ViTs global modeling capability instead of merely focusing on local features, we incorporate a global perturbation attention to prevent information shortcuts in the decoders. Extensive experiments are performed on four widely used anomaly detection benchmarks, including three industrial datasets (MVTec-AD, ViSA, and Real-IAD) and one medical dataset (Universal Medical). The proposed method achieves a remarkable image-level AUROC of 99.8%, 98.9%, 90.6%, and 87.8% on these four datasets, respectively, consistently outperforming previous MUAD methods across different scenarios Our code will be released..
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。