GI-SLAM: Gaussian-Inertial SLAM
Xulang Liu, Ning Tan
TL;DR
GI-SLAM tackles the challenge of robust dense SLAM by integrating inertial data into a 3D Gaussian Splatting framework. It introduces an IMU loss and a covisibility-based, motion-constrained keyframe strategy to jointly optimize pose and a differentiable Gaussian-based map across monocular, stereo, and RGBD sensors. The method demonstrates competitive tracking RMSE and enhanced photometric rendering metrics on EuRoC and TUM-RGBD, with ablation results confirming the value of IMU data and motion-aware keyframe selection. This work advances real-time, photorealistic dense SLAM by unifying inertial sensing with a flexible Gaussian-based scene representation, improving robustness in challenging, texture-poor, and dynamic environments.
Abstract
3D Gaussian Splatting (3DGS) has recently emerged as a powerful representation of geometry and appearance for dense Simultaneous Localization and Mapping (SLAM). Through rapid, differentiable rasterization of 3D Gaussians, many 3DGS SLAM methods achieve near real-time rendering and accelerated training. However, these methods largely overlook inertial data, witch is a critical piece of information collected from the inertial measurement unit (IMU). In this paper, we present GI-SLAM, a novel gaussian-inertial SLAM system which consists of an IMU-enhanced camera tracking module and a realistic 3D Gaussian-based scene representation for mapping. Our method introduces an IMU loss that seamlessly integrates into the deep learning framework underpinning 3D Gaussian Splatting SLAM, effectively enhancing the accuracy, robustness and efficiency of camera tracking. Moreover, our SLAM system supports a wide range of sensor configurations, including monocular, stereo, and RGBD cameras, both with and without IMU integration. Our method achieves competitive performance compared with existing state-of-the-art real-time methods on the EuRoC and TUM-RGBD datasets.