C3D-AD：通过可学习指导的核注意力实现持续三维异常检测

C3D-AD: Toward Continual 3D Anomaly Detection via Kernel Attention with Learnable Advisor

August 2, 2025

作者: Haoquan Lu, Hanzhe Liang, Jie Zhang, Chenxi Hu, Jinbao Wang, Can Gao

cs.AI

摘要

三维异常检测（3D Anomaly Detection, AD）在识别高精度工业产品中的异常或缺陷方面展现出巨大潜力。然而，现有方法通常以类别特定的方式进行训练，且缺乏从新兴类别中学习的能力。本研究提出了一种名为持续三维异常检测（Continual 3D Anomaly Detection, C3D-AD）的持续学习框架，该框架不仅能学习多类点云的通用表示，还能处理随时间出现的新类别。具体而言，在特征提取模块中，为了高效地从不同任务的多类产品中提取通用局部特征，引入了带随机特征层的核注意力机制（Kernel Attention with random feature Layer, KAL），该机制对特征空间进行归一化。随后，为了正确且持续地重建数据，提出了一种高效的带可学习指导的核注意力机制（Kernel Attention with learnable Advisor, KAA），该机制在编码器和解码器中学习新类别的信息，同时摒弃冗余的旧信息。最后，为了保持跨任务表示的一致性，设计了表示排练损失函数，提出了带参数扰动的重建模块（Reconstruction with Parameter Perturbation, RPP），确保模型记住先前类别的信息并返回适应类别的表示。在三个公开数据集上的大量实验验证了所提方法的有效性，分别在Real3D-AD、Anomaly-ShapeNet和MulSen-AD上实现了66.4%、83.1%和63.4%的平均AUROC性能。

English

3D Anomaly Detection (AD) has shown great potential in detecting anomalies or defects of high-precision industrial products. However, existing methods are typically trained in a class-specific manner and also lack the capability of learning from emerging classes. In this study, we proposed a continual learning framework named Continual 3D Anomaly Detection (C3D-AD), which can not only learn generalized representations for multi-class point clouds but also handle new classes emerging over time.Specifically, in the feature extraction module, to extract generalized local features from diverse product types of different tasks efficiently, Kernel Attention with random feature Layer (KAL) is introduced, which normalizes the feature space. Then, to reconstruct data correctly and continually, an efficient Kernel Attention with learnable Advisor (KAA) mechanism is proposed, which learns the information from new categories while discarding redundant old information within both the encoder and decoder. Finally, to keep the representation consistency over tasks, a Reconstruction with Parameter Perturbation (RPP) module is proposed by designing a representation rehearsal loss function, which ensures that the model remembers previous category information and returns category-adaptive representation.Extensive experiments on three public datasets demonstrate the effectiveness of the proposed method, achieving an average performance of 66.4%, 83.1%, and 63.4% AUROC on Real3D-AD, Anomaly-ShapeNet, and MulSen-AD, respectively.