Graph-GSReg: Leveraging 3D Scene Graphs for Gaussian Splatting Registration

2026-06-29

AI Summary

This work addresses key challenges in multi-view 3D Gaussian splatting fusion—namely, inefficient registration, reliance on extensive training data, and susceptibility to occlusion-induced artifacts. It introduces, for the first time, a 3D scene graph into this task and proposes a training-free, graph-driven registration framework. By constructing a 3D scene graph that integrates semantic and structural information, the method achieves a globally consistent high-order representation of the scene. Coupled with a self-supervised test-time optimization strategy, it effectively mitigates post-fusion voids and floating artifacts. Evaluated on both real-world and synthetic datasets, the approach demonstrates superior registration accuracy and rendering quality, significantly outperforming existing methods.

Abstract

Merging multiple 3D Gaussian Splatting (3DGS) scenes into a single unified Gaussian representation is essential for large-scale 3D mapping and long-term map management. Despite its importance, this area remains underexplored, and existing solutions exhibit several limitations. Learning-based methods attempt direct correspondence between Gaussian primitives and require training on large 3DGS datasets. Image-based optimization methods depend heavily on coarse initialization from generic foundation models and often incur expensive refinement. We present \ourmodel. Our method constructs a 3D scene graph from a 3DGS and its rendered images, \textit{reformulating 3DGS registration as a graph registration problem}. The proposed 3D scene graph represents each 3DGS at a higher-level representation, enabling a globally consistent understanding of semantic information and structural context for accurate registration. To further construct a seamless unified scene, we introduce a Self-Supervised Test-Time Optimization. Naively merging two 3D Gaussian scenes often suffers from occlusion artifacts such as hollows and floaters. To alleviate this issue, we refine the merged Gaussians to preserve visual consistency between the original scenes and the merged scene. We evaluate our method on real and synthetic benchmarks, demonstrating competitive registration accuracy and merged scene rendering quality.

Problem

• 3D Gaussian Splatting
• scene registration
• 3D mapping
• scene merging
• occlusion artifacts

Innovation

• 3D Gaussian Splatting
• Scene Graph
• Graph Registration
• Self-Supervised Optimization
• 3D Scene Fusion