DSC: Deep Scan Context Descriptor for Large-Scale Place Recognition

TL;DR

This paper introduces Deep Scan Context, a novel global descriptor for LiDAR-based place recognition that uses raw point clouds and graph neural networks to achieve robust and competitive results in large-scale environments.

Contribution

The paper presents a new descriptor, Deep Scan Context, which leverages raw point clouds and graph neural networks for improved place recognition performance.

Findings

DSC outperforms existing methods on the KITTI dataset.

DSC is robust to scene variations.

Uses only raw point clouds without semantic information.

Abstract

LiDAR-based place recognition is an essential and challenging task both in loop closure detection and global relocalization. We propose Deep Scan Context (DSC), a general and discriminative global descriptor that captures the relationship among segments of a point cloud. Unlike previous methods that utilize either semantics or a sequence of adjacent point clouds for better place recognition, we only use raw point clouds to get competitive results. Concretely, we first segment the point cloud egocentrically to acquire centroids and eigenvalues of the segments. Then, we introduce a graph neural network to aggregate these features into an embedding representation. Extensive experiments conducted on the KITTI dataset show that DSC is robust to scene variants and outperforms existing methods.