EgoNN: Egocentric Neural Network for Point Cloud Based 6DoF Relocalization at the City Scale

TL;DR

EgoNN introduces a neural network that efficiently extracts global and local descriptors from large point clouds for accurate 6DoF relocalization at city scale, combining global retrieval with local matching.

Contribution

The paper proposes a fully convolutional neural network architecture for extracting descriptors from large point clouds, enabling efficient two-stage 6DoF relocalization.

Findings

Effective global descriptor retrieval for city-scale point clouds

Accurate local feature matching with robust pose estimation

Open-source code and pretrained models available

Abstract

The paper presents a deep neural network-based method for global and local descriptors extraction from a point cloud acquired by a rotating 3D LiDAR. The descriptors can be used for two-stage 6DoF relocalization. First, a course position is retrieved by finding candidates with the closest global descriptor in the database of geo-tagged point clouds. Then, the 6DoF pose between a query point cloud and a database point cloud is estimated by matching local descriptors and using a robust estimator such as RANSAC. Our method has a simple, fully convolutional architecture based on a sparse voxelized representation. It can efficiently extract a global descriptor and a set of keypoints with local descriptors from large point clouds with tens of thousand points. Our code and pretrained models are publicly available on the project website.