Abstract:One of the most challenges for anomaly detection (AD) is how to learn one unified and generalizable model to adapt to multi-class especially cross-class settings: the model is trained with normal samples from seen classes with the objective to detect anomalies from both seen and unseen classes. In this work, we propose a novel Proposal Masked Anomaly Detection (PMAD) approach for such challenging multi- and cross-class anomaly detection. The proposed PMAD can be adapted to seen and unseen classes by two key designs: MAE-based patch-level reconstruction and prototype-guided proposal masking. First, motivated by MAE (Masked AutoEncoder), we develop a patch-level reconstruction model rather than the image-level reconstruction adopted in most AD methods for this reason: the masked patches in unseen classes can be reconstructed well by using the visible patches and the adaptive reconstruction capability of MAE. Moreover, we improve MAE by ViT encoder-decoder architecture, combinational masking, and visual tokens as reconstruction objectives to make it more suitable for anomaly detection. Second, we develop a two-stage anomaly detection manner during inference. In the proposal masking stage, the prototype-guided proposal masking module is utilized to generate proposals for suspicious anomalies as much as possible, then masked patches can be generated from the proposal regions. By masking most likely anomalous patches, the “shortcut reconstruction” issue (i.e., anomalous regions can be well reconstructed) can be mostly avoided. In the reconstruction stage, these masked patches are then reconstructed by the trained patch-level reconstruction model to determine if they are anomalies. Extensive experiments show that the proposed PMAD can outperform current state-of-the-art models significantly under the multi- and especially cross-class settings. Code will be publicly available at https://github.com/xcyao00/PMAD.

Revitalizing Reconstruction Models for Multi-class Anomaly Detection via Class-Aware Contrastive Learning

Learning Discrimination from Contaminated Data: Multi-Instance Learning for Unsupervised Anomaly Detection

Hierarchical Semi-supervised Contrastive Learning for Contamination-Resistant Anomaly Detection

A Unified Model for Multi-class Anomaly Detection

A Diffusion-Based Framework for Multi-Class Anomaly Detection

Context-aware Feature Reconstruction for Class-Incremental Anomaly Detection and Localization

Enhancing Multi-Class Anomaly Detection via Diffusion Refinement with Dual Conditioning

Toward Multi-class Anomaly Detection: Exploring Class-aware Unified Model against Inter-class Interference

Learning Unified Reference Representation for Unsupervised Multi-class Anomaly Detection

Reconstruction-based Multi-Normal Prototypes Learning for Weakly Supervised Anomaly Detection

Unilaterally Aggregated Contrastive Learning with Hierarchical Augmentation for Anomaly Detection

Structural Teacher-Student Normality Learning for Multi-Class Anomaly Detection and Localization

AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios

Learning Feature Inversion for Multi-class Anomaly Detection under General-purpose COCO-AD Benchmark

Context Enhancement with Reconstruction as Sequence for Unified Unsupervised Anomaly Detection

One-for-All: Proposal Masked Cross-Class Anomaly Detection

ResAD: A Simple Framework for Class Generalizable Anomaly Detection

Anomaly Detection by Adapting a pre-trained Vision Language Model

Long-Tailed Anomaly Detection with Learnable Class Names

UAC-AD: Unsupervised Adversarial Contrastive Learning for Anomaly Detection on Multi-modal Data in Microservice Systems

Robust Anomaly Detection from Partially Observed Anomalies with Augmented Classes