MMD-SLAM: Structure-Enhanced Multi-Meta Gaussian Distribution-Guided Visual SLAM

2026-06-18

AI Summary

This work addresses the limitations of existing 3D Gaussian Splatting-based SLAM methods, which suffer from degraded rendering quality and map inconsistency due to the neglect of scene structural priors. To overcome these issues, the authors propose a structure-aware visual SLAM framework that innovatively integrates the Atlanta World assumption into a multi-Gaussian representation. The approach jointly optimizes camera poses through point-line fusion, constructs a Multi-Meta Gaussian representation guided by dominant scene directions, and introduces a structure-aware Gaussian evolution mechanism to enforce geometric consistency during mapping. Experimental results demonstrate significant improvements over state-of-the-art methods such as MonoGS, achieving a 48.56% reduction in ATE RMSE on ScanNet and a 5.71 dB gain in PSNR on Replica, thereby validating the effectiveness of leveraging structural priors for high-fidelity dense SLAM.

Abstract

3D Gaussian Splatting (3DGS) has significantly boosted novel view synthesis and high-fidelity scene reconstruction, expanding the potential of 3DGS-based Visual Simultaneous Localization and Mapping (SLAM) methods. However, most existing systems fail to fully exploit the underlying structural information, which limits rendering quality and often leads to inconsistent maps. To address these limitations, we propose MMD-SLAM, a structure-enhanced Visual SLAM framework that leverages the Atlanta World (AW) assumption to guide a Multi-Meta Gaussian representation for photorealistic mapping. First, we introduce a point-line fusion strategy for pose optimization, where 3D line segments are incorporated to improve tracking robustness and provide additional constraints for mapping. Second, we design a Multi-Meta Gaussian representation with dominant directions, explicitly encoding structural priors from the AW hypothesis. Finally, we propose a Gaussian evolution strategy that adapts to scene geometry and incorporates structural cues into global optimization. Extensive experiments demonstrate that these innovations enable MMD-SLAM to achieve state-of-the-art performance in both tracking accuracy and mapping quality. e.g., our method achieves a 48.56% reduction in ATE RMSE on ScanNet and a 5.71% improvement in PSNR on Replica, compared with MonoGS.

Problem

• Visual SLAM
• 3D Gaussian Splatting
• structural information
• mapping consistency
• rendering quality

Innovation

• Multi-Meta Gaussian
• Atlanta World
• point-line fusion
• Gaussian evolution
• structure-aware SLAM