Efficient Unsupervised Point Cloud Anomaly Detector Using Local-Global Features
The author proposes the PointCore framework for unsupervised point cloud anomaly detection, utilizing joint local-global features to reduce computational complexity and improve inference accuracy.
