GSORB-SLAM: Gaussian Splatting SLAM benefits from ORB features and Transmittance information

TL;DR

GSORB-SLAM introduces a robust dense SLAM framework combining 3D Gaussian Splatting with ORB features, employing novel optimization and viewpoint selection techniques to improve accuracy and scene modeling.

Contribution

It presents a new integrated SLAM system that enhances 3D Gaussian Splatting with geometric optimization, adaptive scene modeling, and hybrid viewpoint selection.

Findings

Achieves 16.2% lower RMSE than ORB-SLAM2.

Improves PSNR by 3.93 dB over 3DGS-SLAM.

Demonstrates state-of-the-art performance across multiple datasets.

Abstract

The emergence of 3D Gaussian Splatting (3DGS) has recently ignited a renewed wave of research in dense visual SLAM. However, existing approaches encounter challenges, including sensitivity to artifacts and noise, suboptimal selection of training viewpoints, and the absence of global optimization. In this paper, we propose GSORB-SLAM, a dense SLAM framework that integrates 3DGS with ORB features through a tightly coupled optimization pipeline. To mitigate the effects of noise and artifacts, we propose a novel geometric representation and optimization method for tracking, which significantly enhances localization accuracy and robustness. For high-fidelity mapping, we develop an adaptive Gaussian expansion and regularization method that facilitates compact yet expressive scene modeling while suppressing redundant primitives. Furthermore, we design a hybrid graph-based viewpoint selection mechanism that effectively reduces overfitting and accelerates convergence. Extensive evaluations across various datasets demonstrate that our system achieves state-of-the-art performance in both tracking precision-improving RMSE by 16.2% compared to ORB-SLAM2 baselines-and reconstruction quality-improving PSNR by 3.93 dB compared to 3DGS-SLAM baselines. The project: https://aczheng-cai.github.io/gsorb-slam.github.io/