D3DLO: Deep 3D LiDAR Odometry

TL;DR

D3DLO introduces a deep learning approach for LiDAR odometry that processes raw 3D point clouds directly, achieving comparable accuracy to state-of-the-art methods with significantly fewer parameters.

Contribution

The paper presents a novel end-to-end neural network architecture for LiDAR odometry that operates directly on 3D point clouds without requiring pre-defined point correspondences.

Findings

Achieves similar performance to DeepCLR on KITTI with only 3.56% of parameters.

Plane point extraction reduces input size by 50% with marginal performance loss.

End-to-end training on KITTI demonstrates effectiveness of the proposed approach.

Abstract

LiDAR odometry (LO) describes the task of finding an alignment of subsequent LiDAR point clouds. This alignment can be used to estimate the motion of the platform where the LiDAR sensor is mounted on. Currently, on the well-known KITTI Vision Benchmark Suite state-of-the-art algorithms are non-learning approaches. We propose a network architecture that learns LO by directly processing 3D point clouds. It is trained on the KITTI dataset in an end-to-end manner without the necessity of pre-defining corresponding pairs of points. An evaluation on the KITTI Vision Benchmark Suite shows similar performance to a previously published work, DeepCLR [1], even though our model uses only around 3.56% of the number of network parameters thereof. Furthermore, a plane point extraction is applied which leads to a marginal performance decrease while simultaneously reducing the input size by up to 50%.