Search papers, labs, and topics across Lattice.
This paper introduces "open-set supervised 3D anomaly detection," a novel problem setting where models are trained on normal samples and a few known anomalies to detect unknown anomalies in 3D point clouds. To address this, they present Open3D-AD, a method that models normal and anomalous data distributions using normal samples, simulated anomalies, and partially observed real anomalies, along with a Correspondence Distributions Subsampling technique. Experiments on a new industrial dataset (Open-Industry) and existing benchmarks demonstrate the effectiveness of Open3D-AD for detecting unseen 3D anomalies.
Training on a handful of real-world defect examples plus synthetic data lets you spot entirely new kinds of 3D manufacturing flaws.
Although self-supervised 3D anomaly detection assumes that acquiring high-precision point clouds is computationally expensive, in real manufacturing scenarios it is often feasible to collect a limited number of anomalous samples. Therefore, we study open-set supervised 3D anomaly detection, where the model is trained with only normal samples and a small number of known anomalous samples, aiming to identify unknown anomalies at test time. We present Open-Industry, a high-quality industrial dataset containing 15 categories, each with five real anomaly types collected from production lines. We first adapt general open-set anomaly detection methods to accommodate 3D point cloud inputs better. Building upon this, we propose Open3D-AD, a point-cloud-oriented approach that leverages normal samples, simulated anomalies, and partially observed real anomalies to model the probability density distributions of normal and anomalous data. Then, we introduce a simple Correspondence Distributions Subsampling to reduce the overlap between normal and non-normal distributions, enabling stronger dual distributions modeling. Based on these contributions, we establish a comprehensive benchmark and evaluate the proposed method extensively on Open-Industry as well as established datasets including Real3D-AD and Anomaly-ShapeNet. Benchmark results and ablation studies demonstrate the effectiveness of Open3D-AD and further reveal the potential of open-set supervised 3D anomaly detection.
