Gaussian-LIC2: LiDAR-Inertial-Camera Gaussian Splatting SLAM
2025/07/05 by Xiaolei Lang, Jiajun Lv, Lang, Xiaolei +13 · 1 citation
Engineering · Computer Science · #Robotics and Sensor-Based Localization #Advanced Vision and Imaging #3D Shape Modeling and Analysis
paper · pdf · doi:10.48550/arxiv.2507.04004
Abstract
This paper presents the first photo-realistic LiDAR-Inertial-Camera Gaussian Splatting SLAM system that simultaneously addresses visual quality, geometric accuracy, and real-time performance. The proposed method performs robust and accurate pose estimation within a continuous-time trajectory optimization framework, while incrementally reconstructing a 3D Gaussian map using camera and LiDAR data, all in real time. The resulting map enables high-quality, real-time novel view rendering of both RGB images and depth maps. To effectively address under-reconstruction in regions not covered by the LiDAR, we employ a lightweight zero-shot depth model that synergistically combines RGB appearance cues with sparse LiDAR measurements to generate dense depth maps. The depth completion enables reliable Gaussian initialization in LiDAR-blind areas, significantly improving system applicability for sparse LiDAR sensors. To enhance geometric accuracy, we use sparse but precise LiDAR depths to supervise Gaussian map optimization and accelerate it with carefully designed CUDA-accelerated strategies. Furthermore, we explore how the incrementally reconstructed Gaussian map can improve the robustness of odometry. By tightly incorporating photometric constraints from the Gaussian map into the continuous-time factor graph optimization, we demonstrate improved pose estimation under LiDAR degradation scenarios. We also showcase downstream applications via extending our elaborate system, including video frame interpolation and fast 3D mesh extraction. To support rigorous evaluation, we construct a dedicated LiDAR-Inertial-Camera dataset featuring ground-truth poses, depth maps, and extrapolated trajectories for assessing out-of-sequence novel view synthesis. Both the dataset and code will be made publicly available on project page https://xingxingzuo.github.io/gaussianlic2.
Citations
- Splat-LOAM: Gaussian Splatting LiDAR Odometry and Mapping
- GS-SDF: LiDAR-Augmented Gaussian Splatting and Neural SDF for Geometrically Consistent Rendering and Reconstruction
- PINGS: Gaussian Splatting Meets Distance Fields within a Point-Based Implicit Neural Map
- GS-LIVO: Real-Time LiDAR, Inertial, and Visual Multi-sensor Fused Odometry with Gaussian Mapping
- ActiveGAMER: Active GAussian Mapping through Efficient Rendering
- DepthLab: From Partial to Complete
- Prompting Depth Anything for 4K Resolution Accurate Metric Depth Estimation
- LiV-GS: LiDAR-Vision Integration for 3D Gaussian Splatting SLAM in Outdoor Environments
- LVI-GS: Tightly-coupled LiDAR-Visual-Inertial SLAM using 3D Gaussian Splatting
- Scale Propagation Network for Generalizable Depth Completion
- MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision
- GS-LIVM: Real-Time Photo-Realistic LiDAR-Inertial-Visual Mapping with Gaussian Splatting
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second
- LI-GS: Gaussian Splatting with LiDAR Incorporated for Accurate Large-Scale Reconstruction
- Neural Surface Reconstruction and Rendering for LiDAR-Visual Systems
- FAST-LIVO2: Fast, Direct LiDAR-Inertial-Visual Odometry
- GSFusion: Online RGB-D Mapping Where Gaussian Splatting Meets TSDF Fusion
- IG-SLAM: Instant Gaussian SLAM
- Taming 3DGS: High-Quality Radiance Fields with Limited Resources
- Depth Anything V2
- RaDe-GS: Rasterizing Depth in Gaussian Splatting
- GS-Planner: A Gaussian-Splatting-based Planning Framework for Active High-Fidelity Reconstruction
- MGS-SLAM: Monocular Sparse Tracking and Gaussian Mapping with Depth Smooth Regularization
- High-quality Surface Reconstruction using Gaussian Surfels
- Gaussian Opacity Fields: Efficient Adaptive Surface Reconstruction in Unbounded Scenes
- LetsGo: Large-Scale Garage Modeling and Rendering via LiDAR-Assisted Gaussian Primitives
- Gaussian-LIC: Real-Time Photo-Realistic SLAM with Gaussian Splatting and LiDAR-Inertial-Camera Fusion
- MM-Gaussian: 3D Gaussian-based Multi-modal Fusion for Localization and Reconstruction in Unbounded Scenes
- TCLC-GS: Tightly Coupled LiDAR-Camera Gaussian Splatting for Autonomous Driving
- MM3DGS SLAM: Multi-modal 3D Gaussian Splatting for SLAM Using Vision, Depth, and Inertial Measurements
- HGS-Mapping: Online Dense Mapping Using Hybrid Gaussian Representation in Urban Scenes
- GSDF: 3DGS Meets SDF for Improved Rendering and Reconstruction
- DBA-Fusion: Tightly Integrating Deep Dense Visual Bundle Adjustment with Multiple Sensors for Large-Scale Localization and Mapping
- GaussNav: Gaussian Splatting for Visual Navigation
- MCD: Diverse Large-Scale Multi-Campus Dataset for Robot Perception
- Splat-Nav: Safe Real-Time Robot Navigation in Gaussian Splatting Maps
- How NeRFs and 3D Gaussian Splatting are Reshaping SLAM: a Survey
- LIV-GaussMap: LiDAR-Inertial-Visual Fusion for Real-time 3D Radiance Field Map Rendering
- N3-Mapping: Normal Guided Neural Non-Projective Signed Distance Fields for Large-scale 3D Mapping
- Street Gaussians: Modeling Dynamic Urban Scenes with Gaussian Splatting
- pixelSplat: 3D Gaussian Splats from Image Pairs for Scalable Generalizable 3D Reconstruction
- DrivingGaussian: Composite Gaussian Splatting for Surrounding Dynamic Autonomous Driving Scenes
- Gaussian Splatting SLAM
- Gaussian-SLAM: Photo-realistic Dense SLAM with Gaussian Splatting
- SplaTAM: Splat, Track & Map 3D Gaussians for Dense RGB-D SLAM
- NeuSG: Neural Implicit Surface Reconstruction with 3D Gaussian Splatting Guidance
- Periodic Vibration Gaussian: Dynamic Urban Scene Reconstruction and Real-time Rendering
- Scaffold-GS: Structured 3D Gaussians for View-Adaptive Rendering
- Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular, Stereo, and RGB-D Cameras
- Multi-Scale 3D Gaussian Splatting for Anti-Aliased Rendering
- SuGaR: Surface-Aligned Gaussian Splatting for Efficient 3D Mesh Reconstruction and High-Quality Mesh Rendering
- GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting
- EmerNeRF: Emergent Spatial-Temporal Scene Decomposition via Self-Supervision
- LONER: LiDAR Only Neural Representations for Real-Time SLAM
- 3D Gaussian Splatting for Real-Time Radiance Field Rendering
- 3D Gaussian Splatting for Real-Time Radiance Field Rendering
- Co-SLAM: Joint Coordinate and Sparse Parametric Encodings for Neural Real-Time SLAM
- NF-Atlas: Multi-Volume Neural Feature Fields for Large Scale LiDAR Mapping
- Point-SLAM: Dense Neural Point Cloud-based SLAM
- NeRF-LOAM: Neural Implicit Representation for Large-Scale Incremental LiDAR Odometry and Mapping
- Continuous-Time Fixed-Lag Smoothing for LiDAR-Inertial-Camera SLAM
- NICER-SLAM: Neural Implicit Scene Encoding for RGB SLAM
- NeuS2: Fast Learning of Neural Implicit Surfaces for Multi-view Reconstruction
- ESLAM: Efficient Dense SLAM System Based on Hybrid Representation of Signed Distance Fields
- Unifying Flow, Stereo and Depth Estimation
- NeRF-SLAM: Real-Time Dense Monocular SLAM with Neural Radiance Fields
- SHINE-Mapping: Large-Scale 3D Mapping Using Sparse Hierarchical Implicit Neural Representations
- Orbeez-SLAM: A Real-time Monocular Visual SLAM with ORB Features and NeRF-realized Mapping
- Deep Patch Visual Odometry
- MVSFormer: Multi-View Stereo by Learning Robust Image Features and Temperature-based Depth
- FAST-LIVO: Fast and Tightly-coupled Sparse-Direct LiDAR-Inertial-Visual Odometry
- Instant neural graphics primitives with a multiresolution hash encoding
- NICE-SLAM: Neural Implicit Scalable Encoding for SLAM
- M2DGR: A Multi-sensor and Multi-scenario SLAM Dataset for Ground Robots
- Plenoxels: Radiance Fields without Neural Networks
- Urban Radiance Fields
- R3LIVE: A Robust, Real-time, RGB-colored, LiDAR-Inertial-Visual tightly-coupled state Estimation and mapping package
- CLINS: Continuous-Time Trajectory Estimation for LiDAR-Inertial System
- DROID-SLAM: Deep Visual SLAM for Monocular, Stereo, and RGB-D Cameras
- FAST-LIO2: Fast Direct LiDAR-inertial Odometry
- FAST-LIO2: Fast Direct LiDAR-Inertial Odometry
- Volume Rendering of Neural Implicit Surfaces
- NeuS: Learning Neural Implicit Surfaces by Volume Rendering for Multi-view Reconstruction
- LVI-SAM: Tightly-coupled Lidar-Visual-Inertial Odometry via Smoothing and Mapping
- MVSNeRF: Fast Generalizable Radiance Field Reconstruction from Multi-View Stereo
- iMAP: Implicit Mapping and Positioning in Real-Time
- pixelNeRF: Neural Radiance Fields from One or Few Images
- LIC-Fusion 2.0: LiDAR-Inertial-Camera Odometry with Sliding-Window Plane-Feature Tracking
- ORB-SLAM3: An Accurate Open-Source Library for Visual, Visual–Inertial, and Multimap SLAM
- LIC-Fusion: LiDAR-Inertial-Camera Odometry
- Direct Sparse Odometry
- ImageNet classification with deep convolutional neural networks
Cited by
Related