Abstract
Abstract. In recent research, fully supervised Deep Learning (DL) techniques and large amounts of pointwise labels are employed to train a segmentation network to be applied to buildings’ point clouds. However, fine-labelled buildings’ point clouds are hard to find and manually annotating pointwise labels is time-consuming and expensive. Consequently, the application of fully supervised DL for semantic segmentation of buildings’ point clouds at LoD3 level is severely limited. To address this issue, we propose a novel label-efficient DL network that obtains per-point semantic labels of LoD3 buildings’ point clouds with limited supervision. In general, it consists of two steps. The first step (Autoencoder – AE) is composed of a Dynamic Graph Convolutional Neural Network-based encoder and a folding-based decoder, designed to extract discriminative global and local features from input point clouds by reconstructing them without any label. The second step is semantic segmentation. By supplying a small amount of task-specific supervision, a segmentation network is proposed for semantically segmenting the encoded features acquired from the pre-trained AE. Experimentally, we evaluate our approach based on the ArCH dataset. Compared to the fully supervised DL methods, we find that our model achieved state-of-the-art results on the unseen scenes, with only 10% of labelled training data from fully supervised methods as input.
Highlights
In recent years, 3D buildings’ point cloud representation enables and promotes new applications in many fields such as Cultural Heritage preservation (Pierdicca et al, 2020), Construction Engineering (Ham, Golparvar-Fard, 2015), Emergency Decision-making (Fazeli et al, 2016), and Smart Cities (Hu et al, 2018)
Inspired by the success of deep neural networks (DNNs) used in Computer Vision to accomplish subset tasks, Deep Learning (DL) approaches have appeared in the last few years for understanding 3D point clouds (Cao et al, 2020)
We have presented an effective label-efficient unsupervised network for LoD3 buildings' point clouds semantic segmentation
Summary
3D buildings’ point cloud representation enables and promotes new applications in many fields such as Cultural Heritage preservation (Pierdicca et al, 2020), Construction Engineering (Ham, Golparvar-Fard, 2015), Emergency Decision-making (Fazeli et al, 2016), and Smart Cities (Hu et al, 2018). Point clouds of buildings generally provide the representation of the entire building including only a few types of architectural elements with no semantic information, limiting the efficient exploitation in the abovementioned application domains (Czerniawski, Leite, 2020). It’s essential to investigate the methods of extracting semantic information from 3D buildings’ point clouds to acquire high Level-of-Details (LoDs) modelling, see Wang and Kim (2019). In the buildings’ point cloud domain, DL techniques played an essential role in numerous applications, such as indoor (Wang et al, 2018), urban (Kumar et al, 2019) and buildings’ scenes (Huang et al, 2019) analysis. Even though important results were achieved, the existing DL approaches for 3D building point clouds are strongly supervised, and these methods have substantial demands for finely labelled data, see Meng et al (2020)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
