PointLoc: Deep Pose Regressor for LiDAR Point Cloud Localization

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° 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.

arXiv commentTo appear in IEEE Sensors Journal 2021

References in corpus (7)

• DeepVO: Towards End-to-End Visual Odometry with Deep Recurrent Convolutional Neural Networks
• 3DFeat-Net: Weakly Supervised Local 3D Features for Point Cloud Registration
• Robust Attentional Aggregation of Deep Feature Sets for Multi-view 3D Reconstruction
• Milli-RIO: Ego-Motion Estimation with Low-Cost Millimetre-Wave Radar
• Learning to Localize Using a LiDAR Intensity Map
• DeepCLR: Correspondence-Less Architecture for Deep End-to-End Point Cloud Registration
• Semi-Sequential Probabilistic Model For Indoor Localization Enhancement

Cited by in corpus (3)

• NDT-Transformer: Large-Scale 3D Point Cloud Localisation using the Normal Distribution Transform Representation
• Point wise or Feature wise? Benchmark Comparison of Public Available LiDAR Odometry Algorithms in Urban Canyons
• RadarLoc: Learning to Relocalize in FMCW Radar