【作者】 李晗； 刘宇波； 邓巧明； 胡凯；

【Author】 Li Han;Liu Yubo;Deng Qiaoming;Hu Kai;School of architecture,South China University of Technology,State Key Laboratory of Subtropical Building Science;

【机构】 华南理工大学建筑学院,亚热带建筑科学国家重点实验室；

【摘要】 将不同的元素进行创新性的融合以营造"陌生感",是创意产生的重要途径。建筑设计中常通过融合不同风格来产生全新的形态或流派。然而,由于人类设计师的经验和信息检索能力有限,这种方法更适用于能够快速检索大量数据库并且没有偏好的人工智能。本文以具有复杂性和文化底蕴的太湖石元素在现代建筑形态中的转译为例,探索以扩散模型为例的深度学习模型潜空间辅助三维找形的潜力,并探讨三维到三维的生成式设计在建筑找形中的应用方法。更多还原

【Abstract】 Innovatively blending different elements to evoke a sense of unfamiliarity is a crucial approach to generating creativity.In architectural design,the fusion of various styles is often employed to create new forms or genres.However,due to the limited experiences of human designers,this method is better suited for an artificial intelligence with the ability to rapidly access vast databases without preferences.This article takes the example of the translation of the complex and culturally rich elements of Taihu-stone into modern architectural forms to explore the potential of using deep learning models,specifically diffusion models,to assist in the three-dimensional form-finding process.Furthermore,it investigates the application of 3 D-to-3 D generative design in architectural form-finding.This research aims to be concise and academically oriented.更多还原