VPGS-SLAM: Voxel-based Progressive 3D Gaussian SLAM in Large-Scale Scenes
This addresses the scalability challenge for visual SLAM systems in robotics and AR/VR applications, though it is an incremental advancement over existing 3DGS methods.
The paper tackles the problem of 3D Gaussian Splatting-based SLAM being limited to small scenes and prone to memory issues in large-scale environments, proposing VPGS-SLAM which scales to arbitrary indoor and outdoor scenes with improved robustness and accuracy.
3D Gaussian Splatting has recently shown promising results in dense visual SLAM. However, existing 3DGS-based SLAM methods are all constrained to small-room scenarios and struggle with memory explosion in large-scale scenes and long sequences. To this end, we propose VPGS-SLAM, the first 3DGS-based large-scale RGBD SLAM framework for both indoor and outdoor scenarios. We design a novel voxel-based progressive 3D Gaussian mapping method with multiple submaps for compact and accurate scene representation in large-scale and long-sequence scenes. This allows us to scale up to arbitrary scenes and improves robustness (even under pose drifts). In addition, we propose a 2D-3D fusion camera tracking method to achieve robust and accurate camera tracking in both indoor and outdoor large-scale scenes. Furthermore, we design a 2D-3D Gaussian loop closure method to eliminate pose drift. We further propose a submap fusion method with online distillation to achieve global consistency in large-scale scenes when detecting a loop. Experiments on various indoor and outdoor datasets demonstrate the superiority and generalizability of the proposed framework. The code will be open source on https://github.com/dtc111111/vpgs-slam.