Search papers, labs, and topics across Lattice.
This paper introduces ArcAD, a novel calibration framework designed to enhance supervised anomaly detection in cold-start scenarios where normal samples are limited. By employing a push-pull learning paradigm, ArcAD effectively constructs compact normal boundaries and synthesizes pseudo-anomalies to improve anomaly discrimination. Extensive experiments reveal that ArcAD significantly outperforms existing state-of-the-art methods across various datasets, demonstrating its robustness in both single-class and multi-class settings under data scarcity.
ArcAD reshapes cold-start anomaly detection by synthesizing pseudo-anomalies and clustering limited normal samples, leading to unprecedented performance gains.
The deployment of Industrial Anomaly Detection (IAD) in real-world manufacturing frequently encounters a challenging cold-start bottleneck, in which limited normal samples fail to represent the full normal distribution and only a few anomalies are available. Under such a regime, existing methods struggle to form compact normal boundaries and fail to effectively exploit supervised signals from rare defects. To address this challenge, we propose Anomaly-Rectified Cold-start AD (ArcAD), a plug-and-play calibration framework for reconstruction-based IAD baselines. ArcAD follows a push-pull learning paradigm to construct a compact and discriminative normal boundary under data scarcity. On the one hand, ArcAD projects limited normal samples onto a hypersphere and pulls them into multiple compact clusters to maximize coverage of the normal manifold. On the other hand, it synthesizes pseudo-anomalies on the hypersphere and leverages real anomalies to push the boundary inward and sharpen anomaly discrimination. Extensive experiments on MVTec-AD, VisA, Real-IAD, and MANTA demonstrate that ArcAD significantly outperforms state-of-the-art supervised and unsupervised methods in both single-class and multi-class settings under cold-start conditions. Code is available at: https://github.com/LGC-AD/ArcAD.
