Nothing Stands Still (NSS)
收藏资源简介:
NSS数据集由苏黎世联邦理工学院创建,专注于捕捉和分析建筑内部在建设和翻新过程中的大规模空间和时间变化。数据集包含6个大型建筑内部区域,每个区域在不同时间阶段被重复捕捉,总计27个时间阶段。数据集旨在为计算机视觉和机器人应用提供一个挑战性的基准,特别是在处理动态环境中的3D点云注册问题。NSS数据集不仅包含标准的双点云配准任务,还评估多点云配准,以创建一个连贯的时空地图。
The NSS dataset was created by ETH Zurich, focusing on capturing and analyzing large-scale spatio-temporal changes within building interiors during construction and renovation processes. The dataset includes 6 large-scale building interior regions, each repeatedly captured across different time phases, totaling 27 time phases overall. It aims to provide a challenging benchmark for computer vision and robotic applications, particularly for solving 3D point cloud registration issues in dynamic environments. The NSS dataset not only incorporates the standard two-view point cloud registration task, but also evaluates multi-view point cloud registration to build a coherent spatio-temporal map.

- 1Nothing Stands Still: A Spatiotemporal Benchmark on 3D Point Cloud Registration Under Large Geometric and Temporal Change苏黎世联邦理工学院 · 2023年



