| Id | 2929 | |
| Author | Raman T.A.; Penman S.; Kollar J. | |
| Title | SASAKI: Filling the design gap—Urban impressions with AI | |
| Reference | Raman T.A.; Penman S.; Kollar J. SASAKI: Filling the design gap—Urban impressions with AI,Artificial Intelligence in Urban Planning and Design: Technologies, Implementation, and Impacts |
|
| Keywords | ||
| Link to article | https://www.scopus.com/inward/record.uri?eid=2-s2.0-85137427984&doi=10.1016%2fB978-0-12-823941-4.00002-0&partnerID=40&md5=6a9580fc445a5aa83d522b2a553bbd9e |
|
| Abstract | Planning and design project scopes have been expanding due to the increasing complexity of work in urban environments which have added pressure on limited firm resources. As design and planning functions increasingly overlap, it is the hope that technologies like machine learning and other digital tools will evolve and proliferate to enable designers to employ data-driven analyses in informing a wider range of design decisions. This chapter proposes a potential use case for generative adversarial networks (GANs) in the planning and design process. GANs are used to develop novel impressionistic aerial imagery—or urban impressions—based on a variety of inputs common to practice such as variations of land use and/or land cover. The GANs make extensive use of readily-available, tiled, web-format maps in both their training and prediction. The generated urban impressions were evaluated based on a range of qualitative criteria, including their ability to foster increased dialogue between parties. Additionally, this research included the development of a prototype tool that creates urban impressions based on selected inputs early on in the predesign phase. Ultimately, it is the hope that this prototype “sketch tool” will allow practitioners to use GANs to test out initial “sketched” approaches in planning and design and communicate them to the client for early feedback. © 2022 Elsevier Inc. All rights reserved. |
|
| Metodology | Technique |
Find research methods used
Show Candidate Transition Variables for article (AI method)
Find category for article (AI method)
Find social impact for article (AI method)
Show WordCloud from article (AI method)
Find semantically similar articles (Semantic search)
