File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)

Article: Building layout generation using site-embedded GAN model

TitleBuilding layout generation using site-embedded GAN model
Authors
Keywords3D visualization
Building layout
Design scenario
Generative adversarial network (GAN)
Generative design
Issue Date25-Apr-2023
PublisherElsevier
Citation
Automation in Construction, 2023, v. 151 How to Cite?
Abstract

Building layout generation has entered a new era in recent years, leveraging state-of-the-art deep generative methods to learn morphological properties of exiting urban structures and synthesize building alternatives responsive to local context. However, most existing research generally follows an image-to-image translation idea, while overlooking the impact of site/design attributes on building configuration, making their results less performative. Besides, most synthesized layouts are commonly displayed in 2D pixelized images, limiting further performance evaluation and informed decision-making. This study, therefore, proposes a novel GAN-based model, namely site-embedded generative adversarial networks (ESGAN) for automated building layout generation. Both qualitative and quantitative results in New York City indicate ESGAN is capable of synthesizing visually realistic and semantically reasonable layouts. This end-to-end generative system can not only encode a conditional vector to improve performance in different design scenarios but also display synthesized layouts at different levels of detail for human-system interaction.


Persistent Identifierhttp://hdl.handle.net/10722/335691
ISSN
2023 Impact Factor: 9.6
2023 SCImago Journal Rankings: 2.626
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorJiang, F-
dc.contributor.authorMa, J-
dc.contributor.authorWebster, CJ-
dc.contributor.authorLi, X-
dc.contributor.authorGan, VJ-
dc.date.accessioned2023-12-21T08:35:45Z-
dc.date.available2023-12-21T08:35:45Z-
dc.date.issued2023-04-25-
dc.identifier.citationAutomation in Construction, 2023, v. 151-
dc.identifier.issn0926-5805-
dc.identifier.urihttp://hdl.handle.net/10722/335691-
dc.description.abstract<p><a href="https://www.sciencedirect.com/topics/engineering/building-layout" title="Learn more about Building layout from ScienceDirect's AI-generated Topic Pages">Building layout</a> generation has entered a new era in recent years, leveraging state-of-the-art deep <a href="https://www.sciencedirect.com/topics/engineering/generative-method" title="Learn more about generative methods from ScienceDirect's AI-generated Topic Pages">generative methods</a> to learn morphological properties of exiting urban structures and synthesize building alternatives responsive to local context. However, most existing research generally follows an image-to-image translation idea, while overlooking the impact of site/design attributes on building configuration, making their results less performative. Besides, most synthesized layouts are commonly displayed in 2D pixelized images, limiting further performance evaluation and informed decision-making. This study, therefore, proposes a novel GAN-based model, namely site-embedded generative adversarial networks (ESGAN) for automated building layout generation. Both qualitative and quantitative results in New York City indicate ESGAN is capable of synthesizing visually realistic and semantically reasonable layouts. This end-to-end generative system can not only encode a conditional vector to improve performance in different design scenarios but also display synthesized layouts at different levels of detail for human-system interaction.</p>-
dc.languageeng-
dc.publisherElsevier-
dc.relation.ispartofAutomation in Construction-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.subject3D visualization-
dc.subjectBuilding layout-
dc.subjectDesign scenario-
dc.subjectGenerative adversarial network (GAN)-
dc.subjectGenerative design-
dc.titleBuilding layout generation using site-embedded GAN model-
dc.typeArticle-
dc.identifier.doi10.1016/j.autcon.2023.104888-
dc.identifier.scopuseid_2-s2.0-85153580942-
dc.identifier.volume151-
dc.identifier.isiWOS:001032016300001-
dc.identifier.issnl0926-5805-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats