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

2024-10-15

arXiv.org

AI Summary

Existing dense visual SLAM methods based on 3D Gaussian Splatting (3DGS) suffer from sensitivity to noise and artifacts, suboptimal view selection, and lack of global optimization—limiting both pose estimation accuracy and reconstruction fidelity. This paper proposes GSORB-SLAM, the first tightly coupled dense SLAM framework integrating 3DGS with ORB features. It introduces geometric representation enhancement to improve tracking robustness; designs opacity-aware modeling with adaptive Gaussian duplication and regularization to balance compactness and high-fidelity reconstruction; and establishes a hybrid graph-optimized view selection strategy to suppress overfitting and accelerate convergence. Experiments demonstrate that GSORB-SLAM achieves a 16.2% reduction in pose estimation RMSE compared to ORB-SLAM2 and a 3.93 dB PSNR improvement over 3DGS-SLAM, attaining state-of-the-art performance in both localization accuracy and reconstruction quality.

Technology Category

• Intelligent Robots: State Estimation
• Computer Vision: 3D Computer Vision
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

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/

Problem

• Enhance SLAM accuracy with 3DGS and ORB integration.
• Reduce noise and artifacts in dense SLAM systems.
• Improve scene modeling with adaptive Gaussian expansion.

Innovation

• Integrates 3DGS with ORB features
• Novel geometric representation for tracking
• Adaptive Gaussian expansion for mapping