[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120763-en":3,"doc-seo-120763-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},120763,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","Enhancing Urban Design Through Geodata and Machine Learning","Machine learning methods are applied to urban design by extending prior work that used GIS-derived data to train a Conditional Generative Adversarial Network (CGAN) for generating alternative urban plans. The study introduces GlasGAN, a dataset built from open geospatial data from Glasgow, and validates outcomes with both quantitative and qualitative validation approaches. Results indicate the CGAN can produce geographic, contextually sympathetic urban design proposals, with output quality supported by the conducted validations.","Enhancing Urban Design Through Geodata and Machine Learning  \nAlex Mackay and Diego Pajarito Grajales  \nUrban Big Data Centre, University of Glasgow  \nJanuary 20, 2023  \nSummary  \nMachine learning (ML) methods have seen surprisingly little application in the innovation-driven field of urban design. Research by Boim, Dortheimer, and Sprecher (2022) presented a novel use of ML to generate alternative urban plans considerate of existing local practices, using data extracted from a GIS package to train a Conditional Generative Adversarial Network (CGAN) model. This paper extends that work with a novel dataset built from open geospatial data from Glasgow with results validated using additional quantitative, and qualitative validation methods. The results show the CGAN model is capable of producing geographic and contextually sympathetic urban design proposals with output quality confirmed by result validation.  \nKEYWORDS: Urban Design, GIS, Machine Learning, Conditional Generative Adversarial Networks,  \nPix2PixHD  \n1. Introduction  \nMachine learning (ML) as a branch of artificial intelligence (AI) is the algorithmic powerhouse behind many aspects of modern life. However, Urban design, known for its inextricable connection with geographic spaces, technology and innovation, remains estranged from applied artificial intelligence both academically and in practice (Khean, Fabri, Heusler, 2018; Belam, Santos, Leitao, 2019) . This is due in part to the novelty of machine learning methods, and partly the underlying skills gap surrounding them (Raman, Kollar, Penman, 2022) . Most progressive discourse on ML methods has been framed within the fields of computer science, engineering, and geography, with urban researchers only just beginning to engage with the methods.  \nThe main question preoccupying this paper is ‘How might machine learning methods in conjunction with geoinformation sciences be used to support urban design?’. Image-to-image translation using CGAN models for image infilling was identified as a very promising approach in a review of the literature. ACGAN model was also chosen for its simplicity, requiring little coding knowledge, and for its use of image data, a common data type used in design conceptualisation.  \n2. Data and methods  \nTo prepare a model capable of producing geographic and contextually sympathetic urban design proposals, the authors prepared the GlasGAN site model. It uses a paired dataset of urban map tile images, as done by Boim, Dortheimer and Sprecher (2022) but with data taken from Glasgow’s open geospatial data repositories (i.e., DigiMaps OS Maps) . Three categories of data are used in this study (see figure 1): Ground truth images ‘B’, from the geospatial visualization of open data layers built using GIS, are full unmasked images. Input label images ‘A’, are masked ground truth images. Finally, synthesized images that contains the infilled representations generated by the CGAN.  \nFigure 1-Dataset structure, showing input mask and ground truth.  \n2.1 Data  \nFigure 2-Area covered by GlasGAN: Site datasets with Anderston test area highlighted.  \nA base map of Glasgow showing building footprints, roads, tidal water, and green spaces was constructed using DigiMaps OS Maps geospatial data. The four layers were set to resemble Boim, Dortheimer, and Sprecher's simple block colors and lines using ArcGIS Pro. Once assembled, map tiles were automatically exported from the basic map using a Python script developed by Dortheimer (2022) . The script produced two sets of 900 images, one of which masked 25% of the canvas with a white square, as seen in most CGAN datasets.  \nThe test dataset, paired images not used during model training, was collected from the Anderston area of the city. The Anderston neighborhood is a key example of urban transformation having been entirely redeveloped in the mid 1950s to make way for the city’s M8 motorway. As Anderston sits within the center of the training data area, the typology ","cbCaifW2qeVxDHHa","https://ap.wps.com/l/cbCaifW2qeVxDHHa","pdf",18199529,1,11,"English","en",105,"# Summary\n# Introduction\n# Data and methods\n## Data\n## Methods\n## CGAN Training and Testing","[{\"question\":\"What problem does the paper address in urban design?\",\"answer\":\"It addresses the limited use of machine learning in innovation-driven urban design, especially the gap between ML methods and applied urban design practice.\"},{\"question\":\"What dataset and modeling approach does the paper introduce?\",\"answer\":\"It builds GlasGAN from Glasgow open geospatial data and uses a CGAN framework, specifically Pix2PixHD, to perform image-to-image translation for map infilling.\"},{\"question\":\"How is the model trained and tested?\",\"answer\":\"Training is run in a Google Colab notebook with datasets uploaded to Google Drive, using Pix2PixHD’s cloned source code and a dedicated execution setup for running the notebook code.\"}]","Enhancing Urban Design Through Geodata and Machine Learning | PDF",1785731907,28,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"enhancing-urban-design-through-geodata-and-machine-learning","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/enhancing-urban-design-through-geodata-and-machine-learning/120763/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the paper address in urban design?","Question",{"text":75,"@type":76},"It addresses the limited use of machine learning in innovation-driven urban design, especially the gap between ML methods and applied urban design practice.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What dataset and modeling approach does the paper introduce?",{"text":80,"@type":76},"It builds GlasGAN from Glasgow open geospatial data and uses a CGAN framework, specifically Pix2PixHD, to perform image-to-image translation for map infilling.",{"name":82,"@type":73,"acceptedAnswer":83},"How is the model trained and tested?",{"text":84,"@type":76},"Training is run in a Google Colab notebook with datasets uploaded to Google Drive, using Pix2PixHD’s cloned source code and a dedicated execution setup for running the notebook code.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]