Generative Pre-Trained Transformer for Design Concept Generation: An Exploration
DS 116: Proceedings of the DESIGN2022 17th International Design Conference
Year: 2022
Editor: Mario Štorga, Stanko Škec, Tomislav Martinec, Dorian Marjanović
Author: Qihao Zhu, Jianxi Luo
Series: DESIGN
Institution: Singapore University of Technology and Design, Singapore
Section: Artificial Intelligence and Data-Driven Design
Page(s): 1825-1834
DOI number: https://doi.org/10.1017/pds.2022.185
ISSN: 2732-527X (Online)
Abstract
Novel concepts are essential for design innovation and can be generated with the aid of data stimuli and computers. However, current generative design algorithms focus on diagrammatic or spatial concepts that are either too abstract to understand or too detailed for early phase design exploration. This paper explores the uses of generative pre-trained transformers (GPT) for natural language design concept generation. Our experiments involve the use of GPT-2 and GPT-3 for different creative reasonings in design tasks. Both show reasonably good performance for verbal design concept generation.
Keywords: early design phase, idea generation, generative design, natural language generation, generative pre-trained transformer