Uncertainty-Driven 3-D Gaussian Splatting for Robust Real-Time RGB-D SLAM
Abstract
μSLAM is a real-time RGB-D simultaneous localization and mapping system that combines a sparse feature-based backbone with an uncertainty-aware 3D Gaussian field. It uses uncertainty to guide keyframe selection and map refinement, and a coarse-to-fine loop-closure procedure to improve global consistency. The system is evaluated on public benchmarks and real-world robotic sequences.
Type
Publication
IEEE Transactions on Automation Science and Engineering
