Gaussian-LIC2 Released on arXiv

Our latest work Gaussian-LIC2 is now available on arXiv! 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. By employing a lightweight zero-shot depth model to generate dense depth maps from sparse LiDAR measurements and RGB cues, our system enables reliable Gaussian initialization in LiDAR-blind areas and significantly improves applicability for sparse LiDAR sensors. We also demonstrate improved pose estimation under LiDAR degradation scenarios via tightly incorporating photometric constraints from the Gaussian map into the continuous-time factor graph optimization. Code and dataset will be publicly available.