Watch Live Streaming on:
Beijing (GMT+8): 10:00 p.m., April 14
New York(EST): 10:00 a.m., April 14
Melbourne(GMT+11): 12:00 a.m., April 15
London(GMT): 03:00 p.m., April 14
A Text-to-3D Generative Furniture Design Tool Based on Augmented Reality

Wei Wu
Multi-disciplinary designer
Master of Design Studies, Havard GSD
Wei Wu is a multi-disciplinary designer delving into the intersection of speculative design and emerging technologies. Her interactive works utilize techniques such as robotic fabrication, animation, and extended reality to encourage reflection and examination of the implications of technology on society and the political environment. She has a background in architecture and holds a master’s degree from Harvard Graduate School of Design.
She has previously worked at RoboticPlus, utilizing robotic printing and cutting to design parametric architectural elements and installations, resulting in the world's most extensive carbon fiber roof. Her work was showcased at the ACADIA conference using robotic arms and body tracking to expand the idea of hand drawing and sculpture. She has taught at Harvard University and MIT, offering classes on digital fabrication, photogrammetry, and wearable technology. She has also researched at Hongkong University and MIT on Mixed Reality game design and Brain-Computer Interfaces. Her work has been featured in Archdaily, Dezeen, Archinect, Bustler, Architect Magazine, Archposition, Wallpaper, Fortune China, Animation World Network and other design magazine and platform.
Abstract
The recent breakthrough in generative artificial intelligence (AI) systems, particularly those that generate corresponding content with natural text prompts, have brought transformative changes in various industries, such as furniture design. Recent research has demonstrated that generating high-quality 3D models based on textual descriptions (text-to-3D) has the potential to be used in actual design applications to improve work efficiency. However, the existing design tools are limited for inexperienced users, as they require a stronger spatial awareness skill to concretize 3D designs. In this study, we propose a digital design tool utilizing augmented reality (AR) to visualize and edit text-to-3D generative furniture designs. The proposed design tool allows users to directly generate 3D models by inputting textual descriptions. The generated models can be clearly visualized and interactively modified through an AR device according to design requirements, such as aesthetics and ergonomics. Several examples demonstrate that the proposed design tool can increase design efficiency by providing an interactive and immersive design experience.
Keywords: Generative artificial intelligence; Text-to-3D; Augmented reality (AR); Furniture design
Host
Yufan Xie
Master of Architecture, University of Southern California
Yufan Xie is a multi-disciplinary designer practicing at the intersection of of architecture, generative design and sound synthesis, focusing on multi-sensory narratives driven by generative algorithms, 3d printing and spatial computing.
Comments