SGAD-SLAM: Splatting Gaussians at Adjusted Depth for Better Radiance Fields in RGBD SLAM

2026-03-22

AI Summary

This work addresses the slow convergence and limited rendering fidelity commonly encountered in existing RGB-D SLAM systems that employ 3D Gaussian representations. To overcome these limitations, we propose a pixel-aligned, simplified Gaussian representation that enables each Gaussian to adaptively adjust its depth along the viewing ray. Furthermore, we model the depth distribution within local pixel neighborhoods as a Gaussian, facilitating faster inter-frame alignment. Our approach significantly enhances both rendering fidelity and tracking efficiency while preserving system scalability. Extensive experiments on standard benchmarks demonstrate that our method outperforms state-of-the-art approaches, achieving notable improvements in view synthesis quality, camera pose accuracy, runtime speed, and memory footprint.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: State Estimation
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

3D Gaussian Splatting (3DGS) has made remarkable progress in RGBD SLAM. Current methods usually use 3D Gaussians or view-tied 3D Gaussians to represent radiance fields in tracking and mapping. However, these Gaussians are either too flexible or too limited in movements, resulting in slow convergence or limited rendering quality. To resolve this issue, we adopt pixel-aligned Gaussians but allow each Gaussian to adjust its position along its ray to maximize the rendering quality, even if Gaussians are simplified to improve system scalability. To speed up the tracking, we model the depth distribution around each pixel as a Gaussian distribution, and then use these distributions to align each frame to the 3D scene quickly. We report our evaluations on widely used benchmarks, justify our designs, and show advantages over the latest methods in view rendering, camera tracking, runtime, and storage complexity. Please see our project page for code and videos at https://machineperceptionlab.github.io/SGAD-SLAM-Project .

Problem

• 3D Gaussian Splatting
• RGBD SLAM
• radiance fields
• rendering quality
• convergence speed

Innovation

• 3D Gaussian Splatting
• RGBD SLAM
• depth-adjusted Gaussians
• radiance fields
• real-time tracking