SAGE-ICP: Semantic Information-Assisted ICP

TL;DR

SAGE-ICP is a novel semantic information-assisted ICP method that enhances LiDAR-based pose estimation accuracy in large-scale environments by integrating deep semantic features into the registration process, even with some semantic errors.

Contribution

The paper introduces SAGE-ICP, a new approach that effectively incorporates semantic information into ICP for improved localization accuracy and robustness in large-scale scenes.

Findings

Outperforms baseline methods on KITTI datasets

Improves localization accuracy with semantic integration

Maintains real-time performance faster than sensor frame rate

Abstract

Robust and accurate pose estimation in unknown environments is an essential part of robotic applications. We focus on LiDAR-based point-to-point ICP combined with effective semantic information. This paper proposes a novel semantic information-assisted ICP method named SAGE-ICP, which leverages semantics in odometry. The semantic information for the whole scan is timely and efficiently extracted by a 3D convolution network, and these point-wise labels are deeply involved in every part of the registration, including semantic voxel downsampling, data association, adaptive local map, and dynamic vehicle removal. Unlike previous semantic-aided approaches, the proposed method can improve localization accuracy in large-scale scenes even if the semantic information has certain errors. Experimental evaluations on KITTI and KITTI-360 show that our method outperforms the baseline methods, and improves accuracy while maintaining real-time performance, i.e., runs faster than the sensor frame rate.