arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2607.01698cs.CV
收藏

面向大规模场景重建的结构感知高斯泼溅

Signal Structure-Aware Gaussian Splatting for Large-Scale Scene Reconstruction

Tongji University(同济大学);Shanghai Innovation Institute(上海创新研究院);Shanghai Westwell Technology Co., Ltd(上海西井科技股份有限公司);Beijing University of Technology(北京工业大学);Johns Hopkins University(约翰霍普金斯大学)

Weiyi Xue, Fan Lu, Chi Zhang, Tianhang Wang, Sanqing Qu, Zehan Zheng, Boyuan Zheng, Junqiao Zhao, Guang Chen

AI总结:

针对大规模场景中稀疏观测区域导致的高斯过度密集和冗余问题,提出SIG调度器,通过信号结构恢复视角同步图像监督与高斯频率,并引入球约束高斯控制优化,显著提升效率与渲染质量。

摘要:

3D高斯泼溅在新视角合成中展现出显著潜力。与小规模场景不同,大规模场景不可避免地包含稀疏观测区域,其初始点过于稀疏。在这种情况下,用高频图像监督从低频稀疏点初始化的高斯往往会导致不受控制的密集化和冗余基元,降低效率和质量。直观上,这个问题可以通过调度策略缓解,这些策略可分为两种范式:通过密集化调节目标信号频率,以及通过图像分辨率调节采样频率。然而,先前的调度策略主要是硬编码的,未能感知场景频率的收敛行为。为此,我们从信号结构恢复的角度重新构建场景重建问题,并提出了SIG,一种新颖的调度器,它同步图像监督与高斯频率。具体来说,我们推导了3D表示的平均采样频率和带宽,然后基于场景频率收敛调节训练图像分辨率和高斯密集化过程。此外,我们引入了球约束高斯,利用初始化点云的空间先验来控制高斯优化。我们的框架实现了频率一致、几何感知和无漂浮物的训练,在大规模场景的效率和渲染质量上均以显著优势达到了最先进性能。代码可在以下网址获取:this https URL

英文摘要:

3D Gaussian Splatting has demonstrated remarkable potential in novel view synthesis. In contrast to small-scale scenes, large-scale scenes inevitably contain sparsely observed regions with excessively sparse initial points. In this case, supervising Gaussians initialized from low-frequency sparse points with high-frequency images often induces uncontrolled densification and redundant primitives, degrading both efficiency and quality. Intuitively, this issue can be mitigated with scheduling strategies, which can be categorized into two paradigms: modulating target signal frequency via densification and modulating sampling frequency via image resolution. However, previous scheduling strategies are primarily hardcoded, failing to perceive the convergence behavior of scene frequency. To address this, we reframe the scene reconstruction problem from the perspective of signal structure recovery and propose SIG, a novel scheduler that synchronizes image supervision with Gaussian frequencies. Specifically, we derive the average sampling frequency and bandwidth of 3D representations, and then regulate the training image resolution and the Gaussian densification process based on scene frequency convergence. Furthermore, we introduce Sphere-Constrained Gaussians, which leverage the spatial prior of initialized point clouds to control Gaussian optimization. Our framework enables frequency-consistent, geometry-aware, and floater-free training, achieving state-of-the-art performance by a substantial margin in both efficiency and rendering quality in large-scale scenes. The code is available at: https://github.com/weiyixue999/Signal_Structure_Aware_Gaussian

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑