PointLoc: Deep Pose Regressor for LiDAR Point Cloud Localization

TL;DR

PointLoc introduces a deep learning framework that directly estimates 6-DoF poses from single LiDAR point clouds without pre-built maps, leveraging a novel PointNet-style architecture with self-attention.

Contribution

The paper presents a new end-to-end LiDAR relocalization method using a novel architecture that handles unordered point clouds for accurate pose estimation.

Findings

Achieves accurate relocalization on Oxford Radar RobotCar dataset.

Demonstrates robustness in real-world robot experiments.

Outperforms traditional methods in pose estimation accuracy.

Abstract

In this paper, we present a novel end-to-end learning-based LiDAR relocalization framework, termed PointLoc, which infers 6-DoF poses directly using only a single point cloud as input, without requiring a pre-built map. Compared to RGB image-based relocalization, LiDAR frames can provide rich and robust geometric information about a scene. However, LiDAR point clouds are unordered and unstructured making it difficult to apply traditional deep learning regression models for this task. We address this issue by proposing a novel PointNet-style architecture with self-attention to efficiently estimate 6-DoF poses from 360{\deg} LiDAR input frames.Extensive experiments on recently released challenging Oxford Radar RobotCar dataset and real-world robot experiments demonstrate that the proposedmethod can achieve accurate relocalization performance.