CHING-YI TSAI

Intro

Interior design focuses on creating spaces that are both functional and visually appealing. In traditional settings, even a simple room design can require multiple meetings and extensive work, as interior designers strive to align with homeowners’ preferences in layout, furniture, style, colors, and materials. To streamline this process, we introduce SpaceDreaming, a generative AI-based tool for initial interior design exploration. It enables both homeowners and designers to quickly iterate through a wide array of AI-generated, photorealistic design options, each customized to specific room layouts and personal tastes. We carried out both formative and summative studies involving 18 homeowners and 20 interior designers to refine and assess RoomDreaming.

Homeowners found that RoomDreaming expanded the scope and depth of their design exploration, enhancing both efficiency and satisfaction. Designers noted that an hour of collaborative work using RoomDreaming could produce results that might otherwise take several days of conventional meetings and additional days or weeks to develop and refine designs.

My role in the projects involves the use of prompt-engineering, the design of system architecture, and user study.

Notice

This work is currently in the conditionally acceptance phase for the ACM CHI’24 paper track. The PDF will be provided on request.

Contributor: Shun-Yu Wang, Wei-Chung Su, Serena Chen, Ching-Yi Tsai, Marta Misztal, Katherine M. Cheng, Alwena Lin, Yu Chen, Mike Y. Chen