BEV-SLD: Self-Supervised Scene Landmark Detection for Global Localization with LiDAR Bird's-Eye View Images

TL;DR

BEV-SLD introduces a self-supervised LiDAR-based localization method using bird's-eye-view images to detect scene-specific landmarks, enabling robust and consistent global localization across diverse environments.

Contribution

It proposes a novel self-supervised approach leveraging BEV images for scene landmark detection, improving robustness and consistency in LiDAR-based global localization.

Findings

Outperforms state-of-the-art localization methods in various environments.

Provides consistent landmark detection across different scenes.

Demonstrates robustness in campus, industrial, and forest settings.

Abstract

We present BEV-SLD, a LiDAR global localization method building on the Scene Landmark Detection (SLD) concept. Unlike scene-agnostic pipelines, our self-supervised approach leverages bird's-eye-view (BEV) images to discover scene-specific patterns at a prescribed spatial density and treat them as landmarks. A consistency loss aligns learnable global landmark coordinates with per-frame heatmaps, yielding consistent landmark detections across the scene. Across campus, industrial, and forest environments, BEV-SLD delivers robust localization and achieves strong performance compared to state-of-the-art methods.