arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2305.19590cs.CV
收藏

神经核曲面重建

Neural Kernel Surface Reconstruction

NVIDIA(英伟达);University of Toronto(多伦多大学);Vector Institute(向量研究所)

Jiahui Huang, Zan Gojcic, Matan Atzmon, Or Litany, Sanja Fidler, Francis Williams

AI总结:

本研究提出基于Neural Kernel Fields的神经核曲面重建方法,通过紧支撑核、梯度拟合等解决其局限,可高效处理大规模含噪点云,在多类重建基准上达SOTA。

摘要:

我们提出了一种从大规模、稀疏且含噪的点云中重建三维隐式曲面的新方法。该方法基于最新提出的Neural Kernel Fields(神经核场,简称NKF)表示构建,具备与NKF相近的泛化能力,同时解决了其主要局限:(a)通过紧支撑核函数可扩展至大型场景,支持使用内存高效的稀疏线性求解器;(b)通过梯度拟合求解实现对噪声的鲁棒性;(c)降低了训练要求,可从任何密集带方向点的数据集学习,甚至能混合由不同尺度的物体和场景组成的训练数据。我们的方法能在数秒内重建数百万个点,并以外核方式处理超大型场景。在包含单个物体、室内场景和室外场景的重建基准测试中,我们取得了当前最优的结果。

英文摘要:

We present a novel method for reconstructing a 3D implicit surface from a large-scale, sparse, and noisy point cloud. Our approach builds upon the recently introduced Neural Kernel Fields (NKF) representation. It enjoys similar generalization capabilities to NKF, while simultaneously addressing its main limitations: (a) We can scale to large scenes through compactly supported kernel functions, which enable the use of memory-efficient sparse linear solvers. (b) We are robust to noise, through a gradient fitting solve. (c) We minimize training requirements, enabling us to learn from any dataset of dense oriented points, and even mix training data consisting of objects and scenes at different scales. Our method is capable of reconstructing millions of points in a few seconds, and handling very large scenes in an out-of-core fashion. We achieve state-of-the-art results on reconstruction benchmarks consisting of single objects, indoor scenes, and outdoor scenes.

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。

↑