[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-648297-105":59,"doc-detail-648297-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","lod-sketch-extraction-from-architectural-models-using-generative-ai-dataset-construction-for-multi-level-architectural-design-generation","LoD Sketch Extraction from Architectural Models Using Generative AI - Dataset Construction for Multi-Level Architectural Design Generation","","The paper focuses on multi-level representations in architectural design, emphasizing that smooth transitions from conceptual massing to detailed modeling require consistent LoD (Level of Details). Traditional LoD modeling is manual, slow, labor-intensive, and can introduce geometric inconsistencies. Generative AI enables multi-level models from sketch inputs, yet progress is limited by the lack of high-quality paired LoD training data. An automatic LoD sketch extraction framework uses progressive simplification for geometrically consistent, hierarchically coherent outputs, validated by SSIM and Hausdorff metrics.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/lod-sketch-extraction-from-architectural-models-using-generative-ai-dataset-construction-for-multi-level-architectural-design-generation/648297/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/lod-sketch-extraction-from-architectural-models-using-generative-ai-dataset-construction-for-multi-level-architectural-design-generation/648297.png","ImageObject",300,407,{"name":92,"@type":93},"eBook King","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-25","2026-10-23",true,{"@type":102,"interactionType":103,"userInteractionCount":8},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"Why is multi-level LoD representation important in architectural design?","Question",{"text":112,"@type":113},"Multi-level LoD supports a progressive transition from conceptual massing to detailed modeling. Consistent representation across LoD levels helps maintain smooth evolution of geometry and semantics.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"What key problem limits generative AI for multi-level architectural generation?",{"text":117,"@type":113},"Effective training needs high-quality paired data across multiple LoD levels, but existing datasets often contain only a single LoD and lack samples showing geometric detail evolution.",{"name":119,"@type":110,"acceptedAnswer":120},"How does the proposed framework evaluate geometric consistency across LoD levels?",{"text":121,"@type":113},"It reports SSIM values for LoD3→LoD2 and LoD2→LoD1 transitions and uses normalized Hausdorff distances to measure controlled geometric deviation during abstraction.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},648297,1792919768,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":8,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":52,"language":139,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":140,"faqs":141,"seo_title":142,"seo_description":67,"update_tm":143,"read_time":144},962088006270,"https://ap-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45","LoD Sketch Extraction from Architectural Models Using Generative AI  \nDataset Construction for Multi-Level Architectural Design Generation  \nXusheng Du 1, Athiwat Kongkaeo2, Ye Zhang3, and Haoran Xie4,5  \n1,4 Japan Advanced Institute of Science and Technology.  \n2Chulalongkorn University.  \n3Tianjin University.  \n5 Waseda University.  \n[1](1s2320034@jaist.ac.jp)[s2320034@jaist.ac.jp](1s2320034@jaist.ac.jp), 0009-0008-1086-2081  \nAbstract. For architectural design, the representation across multiple Levels of Details (LoD) is essential for achieving a smooth transition from conceptual massing to detailed modeling. However, traditional LoD modeling processes rely on manual operations, which are timeconsuming, labor-intensive, and prone to geometric inconsistencies. The rapid advancement of generative artificial intelligence (AI) has opened new possibilities for generating multi-level architectural models from sketch inputs. Nevertheless, its application remains limited by the lack of high-quality paired LoD training data. To address this issue, we propose an automatic LoD sketch extraction framework using generative AI models, which progressively simplifies high-detail architectural models to automatically generate geometrically consistent and hierarchically coherent multi-LoD representation. The proposed framework integrates computer vision techniques with generative AI methods to establish a progressive extraction pipeline that transitions from detailed representations to volumetric abstractions. Experimental results demonstrate that the proposed method maintains strong geometric consistency across LoD levels, achieving SSIM values of 0.7319 and 0.7532 for the transitions from LoD3 to LoD2 and from LoD2 to LoD1, respectively. The corresponding normalized Hausdorff distances are 25.1% and 61.0% of the image diagonal, reflecting controlled geometric deviation during abstraction. These results verified that the proposed framework effectively preserves global structure while achieving progressive semantic simplification with different LoD, providing reliable data and technical support for AIdriven multi-level architectural generation and hierarchical modeling.  \nKeywords. Level of Detail (LoD), Sketch Dataset, Generative AI, Architectural Design Generation  \n1. Introduction  \nIn architectural design practice, architectural models are typically represented at  \n–LEAVE THIS WHITE BOX ON PAGE 01!!–If it has moved, you can cut and paste it back to page 1, right click on the boundary and choose 'More Layout Options...' and then under 'Vertical', choose 'Absolute position' -24 cm (below Page).  \ndifferent Levels of Detail (LoD) to accommodate the progressive transition from conceptual expression to detailed design. According to the CityGML standard (OGC, 2012), LoD1 denotes coarse volumetric blocks, LoD2 adds simplified roof geometry and semantic components, and LoD3 includes detailed façade elements such as windows and doors. However, constructing and maintaining these multi-LoD models in digital twin and Building Information Modeling (BIM) systems still rely heavily on manual reconstruction and refinement, which are time-consuming, labor-intensive, and prone to geometric inconsistencies across LoD levels.  \nThe rapid advancement of Generative Artificial Intelligence (AI) has opened up new possibilities for architectural design. Generative models such as Stable Diffusion (Rombach et al., 2022) and conditional generation frameworks like ControlNet (Zhang et al., 2023) can synthesize architectural images from simple inputs such as text or sketches. This capability enables multi-level architectural representations across LoD stages based on sketch inputs, offering a novel paradigm for automated design exploration. Despite its enormous potential, the application of generative AI in architectural design still faces several critical challenges. First, the effective training of such models requires high-quality paired data across multiple LoD","cbCaiektiXfLFYp7","https://ap.wps.com/l/cbCaiektiXfLFYp7","pdf",786981,"English","# Abstract\n# 1. Introduction\n# 2. Related Work\n## 2.1. LOD MODELING","[{\"question\":\"Why is multi-level LoD representation important in architectural design?\",\"answer\":\"Multi-level LoD supports a progressive transition from conceptual massing to detailed modeling. Consistent representation across LoD levels helps maintain smooth evolution of geometry and semantics.\"},{\"question\":\"What key problem limits generative AI for multi-level architectural generation?\",\"answer\":\"Effective training needs high-quality paired data across multiple LoD levels, but existing datasets often contain only a single LoD and lack samples showing geometric detail evolution.\"},{\"question\":\"How does the proposed framework evaluate geometric consistency across LoD levels?\",\"answer\":\"It reports SSIM values for LoD3→LoD2 and LoD2→LoD1 transitions and uses normalized Hausdorff distances to measure controlled geometric deviation during abstraction.\"}]","LoD Sketch Extraction from Architectural Models Using Generative AI - Dataset Construction for Multi-Level Architectural Design Generation | PDF",1792793004,25]