SOURCE-LINKED INTELLIGENCE

PC$^2$-AD: Point Cloud Upsampling to Safeguard 3D Anomaly Detection with Resolution-constrained Edge Devices

arXiv · AI, language, vision and robotics · article · Sep 13, 2026 · UTC

Low-cost and low-resolution sensors used in edge deployments can produce test point clouds that are substantially sparser than the normal training data. This train-test sampling-resolution gap changes the local geometry available to a 3D anomaly detector. We propose PC$^2$-AD, a point cloud upsampling framework that compensates sparse test inputs before downstream detection. Target Domain Candidate Generation (TCG) adapts a pretrained upsampler to normal training geometry and generates a dense candidate pool. Geometry-Aware Candidate Filtering (GACF) selects candidates according to geometric s

Read original source ↗ Open in workspace

recordType

paper

region

Global