생성형 AI를 활용한 실내디자인 협업 방안 연구 = Collaborative Methods for Interior Design Using Generative AI

다국어 초록 (Multilingual Abstract)

The purpose of the study is to develop AI collaboration tools and prompt design techniques that can be utilized in the interior design process, and to propose a collaborative approach to collect, explore, analyse and refine interior architectural information to generate design alternatives that reflect the interior designer’s intentions. The research methodology first explores the concept of AI-designer collaboration through a theoretical review. Next, summarise AI collaboration tools and prompting techniques that can be applied to the interior design process, and propose an AI-based interior design collaboration model, which is then applied to a real interior design project to evaluate its practical use. The research results are as follows. 1) Information Collection Stage: Tools like Clova Note, Archisketch, and Pinterest AI are used to effectively understand client requirements, organize information, and improve efficiency in the early design stages. 2) Information Exploration Stage: Clova Note organizes consultation data, while ChatGPT and Archisketch categorize and store client preferences. The CoT technique helps break down client needs and provide clear design directions. 3) Information Analysis Stage: ChatGPT and Archisketch create design prompts and perspective images, combining functional and emotional prompts to reflect client needs and emotions. Techniques like ICL & GKP are used to generate design themes and new ideas. 4) Information Refinement Stage: Midjourney’s Edit and Retexture features refine images, and GKP techniques incorporate improvements, resulting in finalized designs that align with client expectations.