AI 2D-3D Generation in Architectural Design: A Divine Hand or a Pandora's Box

• 出版年: 2024 IEEE 22nd Student Conference on Research and Development, SCOReD 2024
• 第一著者: 2-s2.0-85219511098
• フォーマット: Conference paper
• 言語: English
• 出版事項: Institute of Electrical and Electronics Engineers Inc. 2024
• オンライン･アクセス: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85219511098&doi=10.1109%2fSCOReD64708.2024.10872721&partnerID=40&md5=a75f81a2e5efa2bc304edf5f58d666c8

This study explores the integration of AI-powered 2D-to-3D generation tools in architectural design workflows, focusing on open-source ComfyUI and its 3D pack extension. A mixed-methods approach combining workflow demonstrations with architecture students and quantitative survey analysis reveals both potential applications and pedagogical implications. While not yet practical for professional use, AIgenerated 3D models serve as valuable educational tools for conceptual design, facilitating spatial visualization and design exploration. Survey results (N=126) revealed strong student support for AI integration in architectural education, driven more by its future potential than current capabilities. This forward-looking perspective and ability to separate educational benefits from job security concerns indicate students' preparedness for an AI-influenced profession. This study demonstrates how AI-driven 3D generation tools can be effectively integrated into architectural design workflows and education, balancing technological innovation with pedagogical needs. © 2024 IEEE.