Stereo 3D Gaussian Splatting SLAM for Outdoor Urban Scenes

TL;DR

This paper introduces BGS-SLAM, a novel outdoor stereo SLAM system using 3D Gaussian Splatting that operates solely on RGB stereo images, achieving high accuracy without active sensors or LiDAR.

Contribution

BGS-SLAM is the first outdoor 3D Gaussian Splatting SLAM system that uses only stereo RGB images and deep stereo networks for depth estimation.

Findings

Outperforms existing 3DGS SLAM methods in outdoor environments

Achieves high tracking accuracy and detailed mapping

Operates effectively without active depth sensors or LiDAR

Abstract

3D Gaussian Splatting (3DGS) has recently gained popularity in SLAM applications due to its fast rendering and high-fidelity representation. However, existing 3DGS-SLAM systems have predominantly focused on indoor environments and relied on active depth sensors, leaving a gap for large-scale outdoor applications. We present BGS-SLAM, the first binocular 3D Gaussian Splatting SLAM system designed for outdoor scenarios. Our approach uses only RGB stereo pairs without requiring LiDAR or active sensors. BGS-SLAM leverages depth estimates from pre-trained deep stereo networks to guide 3D Gaussian optimization with a multi-loss strategy enhancing both geometric consistency and visual quality. Experiments on multiple datasets demonstrate that BGS-SLAM achieves superior tracking accuracy and mapping performance compared to other 3DGS-based solutions in complex outdoor environments.