揭示3D高斯点云渲染优化中的稳定结构与密度依赖规律
Analysis of Converged 3D Gaussian Splatting Solutions: Density Effects and Prediction Limit
- 通过可学习探针分析渲染最优解的参数分布规律
- 稀疏区域参数受视角异质性影响，出现耦合失效现象
- 提出密度感知策略，提升训练鲁棒性，适配自适应架构
我们研究了标准多视角优化下3D高斯点云（3DGS）解中浮现的结构。将其称为渲染最优参考（ROR），并分析其统计特性，发现混合尺度结构和跨场景的双峰辐射率分布具有稳定性。为理解这些参数的决定因素，我们采用可学习探针，训练预测器从点云重建ROR，无需渲染监督。分析揭示出基本的密度分层机制：密集区域参数与几何相关，可实现无渲染预测；稀疏区域则在各架构下系统性失败。通过方差分解证明，可见性异质性导致稀疏区域几何与外观参数间以协方差主导的耦合关系。这揭示了ROR的双重属性：点云足以表征的几何原语，以及需多视角约束的视图合成原语。我们提出密度感知策略以增强训练鲁棒性，并讨论了自适应平衡前馈预测与渲染修正的架构启示。
原文摘要 · Abstract (English)
We investigate what structure emerges in 3D Gaussian Splatting (3DGS) solutions from standard multi-view optimization. We term these Rendering-Optimal References (RORs) and analyze their statistical properties, revealing stable patterns: mixture-structured scales and bimodal radiance across diverse scenes. To understand what determines these parameters, we apply learnability probes by training predictors to reconstruct RORs from point clouds without rendering supervision. Our analysis uncovers fundamental density-stratification. Dense regions exhibit geometry-correlated parameters amenable to render-free prediction, while sparse regions show systematic failure across architectures. We formalize this through variance decomposition, demonstrating that visibility heterogeneity creates covariance-dominated coupling between geometric and appearance parameters in sparse regions. This reveals the dual character of RORs: geometric primitives where point clouds suffice, and view synthesis primitives where multi-view constraints are essential. We provide density-aware strategies that improve training robustness and discuss architectural implications for systems that adaptively balance feed-forward prediction and rendering-based refinement.
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。