LVI-GS：使用3D高斯泼溅的紧耦合激光雷达-视觉-惯性SLAM
LVI-GS: Tightly-coupled LiDAR-Visual-Inertial SLAM using 3D Gaussian Splatting
- The University of Hong Kong（香港大学）
- HKU-TCL Joint Research Center for Artificial Intelligence（香港大学-TCL人工智能联合研究中心）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
提出LVI-GS，一种紧耦合激光雷达-视觉-惯性SLAM框架，利用3D高斯泼溅实现实时高保真建图，通过金字塔训练和深度损失增强几何感知，性能优于现有3D重建系统。
AI中文摘要：
3D高斯泼溅（3DGS）已展现出其在快速渲染和高保真建图方面的能力。本文介绍了LVI-GS，一种基于3DGS的紧耦合激光雷达-视觉-惯性建图框架，该框架利用激光雷达与图像传感器的互补特性，捕获3D场景的几何结构和视觉细节。为此，3D高斯从着色后的激光雷达点初始化，并通过可微渲染进行优化。为实现高保真建图，我们引入基于金字塔的训练方法，以有效学习多级特征，并结合源自激光雷达测量的深度损失，增强几何特征感知。通过精心设计的高斯地图扩展、关键帧选择、线程管理及自定义CUDA加速策略，我们的框架实现了实时的照片级逼真建图。数值实验评估了我们的方法相比最先进的3D重建系统的优越性能。
英文摘要：
3D Gaussian Splatting (3DGS) has shown its ability in rapid rendering and high-fidelity mapping. In this paper, we introduce LVI-GS, a tightly-coupled LiDAR-Visual-Inertial mapping framework with 3DGS, which leverages the complementary characteristics of LiDAR and image sensors to capture both geometric structures and visual details of 3D scenes. To this end, the 3D Gaussians are initialized from colourized LiDAR points and optimized using differentiable rendering. In order to achieve high-fidelity mapping, we introduce a pyramid-based training approach to effectively learn multi-level features and incorporate depth loss derived from LiDAR measurements to improve geometric feature perception. Through well-designed strategies for Gaussian-Map expansion, keyframe selection, thread management, and custom CUDA acceleration, our framework achieves real-time photo-realistic mapping. Numerical experiments are performed to evaluate the superior performance of our method compared to state-of-the-art 3D reconstruction systems.
