WGICP: Differentiable Weighted GICP-Based Lidar Odometry

TL;DR

This paper introduces WGICP, a differentiable weighted GICP method for Lidar odometry that enhances accuracy and efficiency by learning point importance, enabling better pose estimation in 3D point cloud data.

Contribution

The paper proposes a novel differentiable weighted GICP algorithm that integrates neural network-based point weighting, improving Lidar odometry accuracy and computational efficiency.

Findings

Improved accuracy on KITTI dataset

Faster pose estimation compared to traditional GICP

Enhanced robustness in SLAM applications

Abstract

We present a novel differentiable weighted generalized iterative closest point (WGICP) method applicable to general 3D point cloud data, including that from Lidar. Our method builds on differentiable generalized ICP (GICP), and we propose using the differentiable K-Nearest Neighbor (KNN) algorithm to enhance differentiability. The differentiable GICP algorithm provides the gradient of output pose estimation with respect to each input point, which allows us to train a neural network to predict its importance, or weight, in estimating the correct pose. In contrast to the other ICP-based methods that use voxel-based downsampling or matching methods to reduce the computational cost, our method directly reduces the number of points used for GICP by only selecting those with the highest weights and ignoring redundant ones with lower weights. We show that our method improves both accuracy and speed of the GICP algorithm for the KITTI dataset and can be used to develop a more robust and efficient SLAM system.