VBGS-SLAM: Variational Bayesian Gaussian Splatting Simultaneous Localization and Mapping

2026-04-02

AI Summary

This work addresses the limitations of existing 3D Gaussian Splatting-based SLAM methods, which rely on deterministic pose optimization, suffer from sensitivity to initialization, and are prone to catastrophic forgetting. We propose the first approach that integrates variational Bayesian inference into 3DGS-SLAM, formulating a generative probabilistic framework that jointly optimizes camera poses and scene Gaussian parameters while explicitly modeling their posterior uncertainties. Leveraging the conjugate structure of multivariate Gaussians, our method enables efficient closed-form updates, preserving the high rendering efficiency of 3D Gaussian splatting while significantly enhancing system robustness and mitigating pose drift. Experiments demonstrate superior long-term tracking accuracy and high-quality novel view synthesis across diverse synthetic and real-world scenes.

Abstract

3D Gaussian Splatting (3DGS) has shown promising results for 3D scene modeling using mixtures of Gaussians, yet its existing simultaneous localization and mapping (SLAM) variants typically rely on direct, deterministic pose optimization against the splat map, making them sensitive to initialization and susceptible to catastrophic forgetting as map evolves. We propose Variational Bayesian Gaussian Splatting SLAM (VBGS-SLAM), a novel framework that couples the splat map refinement and camera pose tracking in a generative probabilistic form. By leveraging conjugate properties of multivariate Gaussians and variational inference, our method admits efficient closed-form updates and explicitly maintains posterior uncertainty over both poses and scene parameters. This uncertainty-aware method mitigates drift and enhances robustness in challenging conditions, while preserving the efficiency and rendering quality of existing 3DGS. Our experiments demonstrate superior tracking performance and robustness in long sequence prediction, alongside efficient, high-quality novel view synthesis across diverse synthetic and real-world scenes.

Problem

• 3D Gaussian Splatting
• SLAM
• pose optimization
• catastrophic forgetting
• initialization sensitivity

Innovation

• Variational Inference
• 3D Gaussian Splatting
• Uncertainty Quantification
• SLAM
• Probabilistic Modeling