Delft University of Technology南亮亮教授学术报告（2025年10月30日）
报告题目：Reconstructing and Understanding Urban Environments in 3D
报告人：南亮亮教授
报告时间：2025年10月30日 14:00-16:20
报告地点：仙林校区自动化学科楼321会议室
主办单位：自动化学院、科学技术处
报告人简介：
Liangliang Nan received his Ph.D. degree in Mechatronics Engineering from the Graduate University of the Chinese Academy of Sciences in 2009. From 2009 to 2013, he was an assis
tant professor and later associate professor at Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences.
Between 2013 and 2018, he worked as a Research Scientist at the Visual Computing Center, KAUST. He is currently an Associate Professor at Delft University of Technology, where he leads the AI lab on 3D Urban Understanding (3DUU). His research interests lie at the intersection of computer vision, computer graphics, and 3D geoinformation, with a focus on understanding and modeling real-world scenes. His expertise has been recognized with numerous awards, and he served as an area chair for ICCV 2023 and ECCV 2024.Delft University of Technology. He published more than 60 papers and 12 PCTs such as IEEE TPAMI, ACM TOG, SIGGRAPH, SIGGRAPH Asia, IJCV, ICCV, CVPR, ECCV, et.al.
报告摘要：Digitalization of urban scenes is transforming how we document, analyze, and plan our built environment. Advances in laser scanning and 3D computer vision now allow us to capture cities as massive collections of 3D points (i.e., point clouds) that encode real-world geometry with remarkable detail. Yet turning these raw data into compact, accurate, and meaningful 3D models remains a key challenge. In this talk, I will outline our journey from point cloud based scene reconstruction to semantic understanding of urban models. I will introduce techniques that convert noisy, incomplete 3D scans into lightweight, watertight polygonal representations and demonstrate our effort on inferring high-level structures and semantics. Finally, I will highlight emerging trends toward large-scale, semantically rich digital twins of cities. These developments promise to support applications in urban planning, sustainability analysis, autonomous navigation, and cultural heritage preservation, bringing us closer to truly intelligent digital cities.