提升3D高斯溅射的位姿优化稳定性，应对初始位姿和几何误差挑战
Rethinking Pose Refinement in 3D Gaussian Splatting under Pose Prior and Geometric Uncertainty
- 用蒙特卡洛采样结合费舍尔信息优化，显式建模位姿与几何不确定性
- 在室内外多个基准上，定位精度提升且抗噪声能力显著增强
- 无需重训练或额外监督，适合真实场景中鲁棒位姿估计应用
3D高斯溅射（3DGS）作为一种强大的场景表示方法，正被广泛用于视觉定位与位姿精修。然而，尽管其具有高质量的可微分渲染能力，基于3DGS的位姿精修仍对初始相机位姿和重建几何高度敏感。本文深入分析这一问题，识别出两大不确定性来源：(i) 位姿先验不确定性，常源于回归或检索模型输出单一确定性估计；(ii) 几何不确定性，由3DGS重建中的不完美导致，误差会传播至PnP求解器。这些不确定性可能扭曲重投影几何并破坏优化稳定性，即使渲染外观仍合理。为此，我们提出一种重定位框架，结合蒙特卡洛位姿采样与基于费舍尔信息的PnP优化。该方法显式建模位姿与几何不确定性，且无需重训练或额外监督。在多样化的室内外基准测试中，该方法持续提升定位精度，并显著增强在位姿与深度噪声下的稳定性。
原文摘要 · Abstract (English)
3D Gaussian Splatting (3DGS) has recently emerged as a powerful scene representation and is increasingly used for visual localization and pose refinement. However, despite its high-quality differentiable rendering, the robustness of 3DGS-based pose refinement remains highly sensitive to both the initial camera pose and the reconstructed geometry. In this work, we take a closer look at these limitations and identify two major sources of uncertainty: (i) pose prior uncertainty, which often arises from regression or retrieval models that output a single deterministic estimate, and (ii) geometric uncertainty, caused by imperfections in the 3DGS reconstruction that propagate errors into PnP solvers. Such uncertainties can distort reprojection geometry and destabilize optimization, even when the rendered appearance still looks plausible. To address these uncertainties, we introduce a relocalization framework that combines Monte Carlo pose sampling with Fisher Information-based PnP optimization. Our method explicitly accounts for both pose and geometric uncertainty and requires no retraining or additional supervision. Across diverse indoor and outdoor benchmarks, our approach consistently improves localization accuracy and significantly increases stability under pose and depth noise.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。