Building3D is an urban-scale dataset for building roof modelling from aerial LiDAR point clouds. It consists of more than 160 thousand buildings, covering 16 cities in Estonia about 998 Km^2. Besides mesh models and real-world LiDAR point clouds, it is the first time to release wireframe models. We believe that our Building3D will facilitate future research on urban modelling, mesh simplification, point cloud completion and semantic segmentation etc.

Building3D Preview


If you find our dataset useful, please cite the following paper:

  title={Building3D: An Urban-Scale Dataset and Benchmarks for Learning Roof Structures from Point Clouds},
  author={Ruisheng Wang, Shangfeng Huang, Hongxin Yang},
  booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision},
  year = {2023}


  • 2023-7-21 Entry-level dataset
  • 2023-4-15 3D LabelMe
  • 2023-1-10 Publish Building3D website

Paper Preview


Building3D dataset and Benchmark © 2022 is licensed under CC BY-NC-SA 4.0. This means that you must attribute the work in the manner specified by the authors, you may not use this work for commercial purposes and if you alter, transform, or build upon this work, you may distribute the resulting work only under the same license.