arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.30342cs.CVcs.RO
收藏

基于各向异性可见性场的不确定性驱动的3D高斯溅射主动建图

Uncertainty-driven 3D Gaussian Splatting Active Mapping via Anisotropic Visibility Field

Georgia Institute of Technology(佐治亚理工学院)

Shangjie Xue, Jesse Dill, Dhruv Ahuja, Frank Dellaert, Panagiotis Tsiotras, Danfei Xu

AI总结:

提出GAVIS框架,通过各向异性可见性场量化3DGS的不确定性,并基于最大信息增益实现主动建图,在精度和效率上显著优于现有方法。

摘要:

我们提出了高斯溅射各向异性可见性场(GAVIS),这是一个用于3DGS中不确定性量化和主动建图的新框架。我们的关键洞察是,训练视图中未见的区域会导致3DGS产生不可靠的预测。为了解决这个问题,我们引入了一种原则性且高效的方法来量化3DGS中的可见性场,定义为每个粒子相对于训练视图的各向异性可见性,并使用球谐函数表示。得到的可见性场被集成到基于贝叶斯网络的不确定性感知3DGS光栅化器中,实现了对合成视图的实时(200 FPS)不确定性量化。在此基础上,进一步在最大信息增益框架内执行主动建图。跨多种环境的广泛实验表明,GAVIS在精度和效率上始终且显著优于先前的方法。此外,除了独立使用外,我们的方法还可以事后应用于改进现有方法的性能。

英文摘要:

We present Gaussian Splatting Anisotropic Visibility Field (GAVIS), a novel framework for uncertainty quantification and active mapping in 3DGS. Our key insight is that regions unseen from the training views yield unreliable predictions from the 3DGS. To address this, we introduce a principled and efficient method for quantifying the visibility field in 3DGS, defined as the anisotropic visibility of each particle with respect to the training views, and represented using spherical harmonics. The resulting visibility field is integrated into a Bayesian Network-based uncertainty-aware 3DGS rasterizer, enabling real-time (200 FPS) uncertainty quantification for synthesized views. Active mapping is further performed within a maximum information gain framework building on this formulation. Extensive experiments across diverse environments demonstrate that GAVIS consistently and significantly outperforms prior approaches in both accuracy and efficiency. Moreover, beyond standalone use, our method can be applied post-hoc to improve the performance of existing approaches.

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。

↑