Skip to content

Opening book details…

About this document

LPD-Net for 3D Point Cloud Recognition by alex.muravev is a document available to read on EtoBox.

The document presents LPD-Net, a novel deep neural network designed for large-scale place recognition and environment analysis using 3D point clouds. It introduces two key modules for adaptive local feature extraction and graph-based neighborhood aggregation, enabling the extraction of discriminative global descriptors. LPD-Net outperforms existing methods like PointNetVLAD and demonstrates robustness against varying weather and lighting conditions.

Author
alex.muravev
Language
EN