Exploring graph neural networks for semantic enrichment: Room type classification

Wang, Zijian; Sacks, Rafael; Yeung, Timson

doi:10.1016/j.autcon.2021.104039

Cited by 50 publications

(25 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, machine-learning algorithms have been applied to the semantic enrichment of generating models. For example, Wang et al [162] applied graph neural networks to classify rooms, enrich apartment models with their room types, and further compliance-checking tasks.…”

Section: Fast Modelling With Semantic Enrichmentmentioning

confidence: 99%

A Review of Digital Twin Applications in Civil and Infrastructure Emergency Management

Cheng

Hou

2023

Buildings

View full text Add to dashboard Cite

Natural disasters can cause severe damages to civil infrastructure and lead to extensive economic losses and casualties. To improve the emergency response capability of civil infrastructure under extreme circumstances such as natural disasters and human-caused hazards, intelligent technology for infrastructure emergency management has been extensively studied. As an emerging paradigm of interdisciplinary convergence, digital twins (DTs) can integrate intelligent technology into different stages of emergency management and provide a new solution for the emergency management of civil infrastructure (EMCI). However, applications of DT in EMCI have several limitations and are mostly case by case. However, the sector needs more generalisable lessons to address the greater value of DT in the context of EMCI. To address this gap, we first carry out a systematic literature review and analyse the latest progress and previous research deficiencies of DT by taking the scientometrical approach. Next, a framework is proposed to explain how DT can be applied to the mitigation, preparation, response, and recovery stages of EMCI. Lastly, the trends and prospects of DT applications in EMCI are discussed. Overall, the knowledge gained from this study will promote the research and development of more-viable DTs to address the sector’s demand for emergency management.

show abstract

Section: Fast Modelling With Semantic Enrichmentmentioning

confidence: 99%

A Review of Digital Twin Applications in Civil and Infrastructure Emergency Management

Cheng

Hou

2023

Buildings

View full text Add to dashboard Cite

show abstract

“…Existing work indicates a significant potential for ML and DL methods to address classification needs in BIM in general and space function classification in particular. Existing space function classifiers use space feature vectors or space connectivity graphs as input (Bloch and Sacks, 2018;Wang, Sacks and Yeung, 2022). However, the applicability of DL image segmentation methods to space function classification has not been studied.…”

Section: Related Workmentioning

confidence: 99%

SFS-A68: a Dataset for the Segmentation of Space Functions in Apartment Buildings

Ziaee¹,

Suter²

2022

Proceedings of the 29th EG-ICE International Workshop on Intelligent Computing in Engineering

View full text Add to dashboard Cite

Analyzing building models for usable area, building safety, or energy analysis requiresfunction classification data of spaces and related objects. Automated space function classificationis desirable to reduce input model preparation effort and errors. Existing space function classifiersuse space feature vectors or space connectivity graphs as input. The application of deep learning(DL) image segmentation methods to space function classification has not been studied. As aninitial step towards addressing this gap, we present a dataset, SFS-A68, that consists of input andground truth images generated from 68 digital 3D models of space layouts of apartment buildings.The dataset is suitable for developing DL models for space function segmentation. We use thedataset to train and evaluate an experimental space function segmentation network based ontransfer learning and training from scratch. Test results confirm the applicability of DL imagesegmentation for space function classification.

show abstract

“…Based on Rec-GNNs or Conv-GNNs, Skip-GNNs add "skip connection" across layers to avoid over smoothing problems and increased noise in deeper networks (Zhou et al, 2020). Researchers in the discipline of computer science have conducted several surveys or reviews on GNNs (Bronstein et al, 2017;Zhang et al, 2019;Zhou et al, 2020;Wu et al, 2021b;Wang et al, 2022;Zhang et al, 2022). Comprehensive explanations of graphs and GNNs can be found in these studies.…”

Section: Graph Neural Network (Gnns)mentioning

confidence: 99%

“…GNNs have been widely applied to chemistry, biology, recommendation systems, traffic networks, etc (Zhou et al, 2020). Some studies in the AEC sector also investigated the applications of GNNs such as BIM segmentation enrichment (Wang et al, 2022), point cloud semantic segmentation (Feng et al, 2021), and prediction of building energy consumption (Hu et al, 2022). However, the adoption of GNNs in AEC is still in its infancy.…”

Section: Introductionmentioning

confidence: 99%

Graph neural networks in building life cycle: a review

Yilong¹,

Wang²,

Reza³

et al. 2022

Proceedings of the 2022 European Conference on Computing in Construction

View full text Add to dashboard Cite

Graph neural networks (GNNs) have attracted much attention in the field of machine learning because of their excellent performance on graph data. Graph data in the architecture, engineering and construction (AEC) sector is very common, such as bubble diagrams for space planning and point clouds for scan-to-BIM models. Some studies in AEC have adopted GNNs to solve practical problems. However, there has been a limited focus on the outcomes of these studies. Therefore, this paper aims to review the applications of GNNs in the building lifecycle. A wide range of existing literature was retrieved. The result shows that the adoption of GNNs is still in its infancy but has been increasing dramatically in recent years. Ten application domains were identified from the planning stage to the operation stage. In addition, the challenges and opportunities of GNNs adoption in AEC were discussed providing directions for future research.

show abstract

Exploring graph neural networks for semantic enrichment: Room type classification

Cited by 50 publications

References 28 publications

A Review of Digital Twin Applications in Civil and Infrastructure Emergency Management

A Review of Digital Twin Applications in Civil and Infrastructure Emergency Management

SFS-A68: a Dataset for the Segmentation of Space Functions in Apartment Buildings

Graph neural networks in building life cycle: a review

Contact Info

Product

Resources

About