C3D-AD: Toward Continual 3D Anomaly Detection via Kernel Attention with Learnable Advisor
By: Haoquan Lu , Hanzhe Liang , Jie Zhang and more
Potential Business Impact:
Finds hidden flaws in 3D objects automatically.
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.
Similar Papers
MC3D-AD: A Unified Geometry-aware Reconstruction Model for Multi-category 3D Anomaly Detection
CV and Pattern Recognition
Finds flaws in many different 3D objects.
A Lightweight 3D Anomaly Detection Method with Rotationally Invariant Features
CV and Pattern Recognition
Find weird spots in 3D shapes, no matter how they're turned.
PointAD+: Learning Hierarchical Representations for Zero-shot 3D Anomaly Detection
CV and Pattern Recognition
Finds hidden flaws in 3D objects using pictures.