文献类型：期刊文献

英文题名：A product form design method integrating Kansei engineering and diffusion model

作者：Yang, Chaoxiang[1];Liu, Fei[1];Ye, Junnan[1]

机构：[1]East China Univ Sci & Technol, Sch Art Design & Media, 130 Meilong Rd, Shanghai 200237, Peoples R China

年份：2023

卷号：57

外文期刊名：ADVANCED ENGINEERING INFORMATICS

收录：;EI(收录号：20232714359986);WOS:【SCI-EXPANDED(收录号:WOS:001029165400001)】；

基金：The authors would like to thank all of the anonymous referees for the comments and suggestions, which have helped to improve the paper. In addition, the authors wish to thank the Shanghai Pujiang Program and the Chinese Universities Scientific Fund, for financially supporting under Contract No. 2020PJC025 and No. JKZ01212202.

语种：英文

外文关键词：Product form design; User emotional demands; Kansei engineering; Diffusion model

摘要：The experience economy has shifted consumer demands from the functional to the emotional, and the emotional demands of the user have become a key design consideration. At the same time, product form design often relies solely on the knowledge and experience of designers, resulting in uneven design results and difficult quality assurance. To this end, this paper proposes a product form design method that applies an image generation algorithm oriented toward satisfying the emotional demands of users. Firstly, product images from the network are collected and processed to build a dataset of product images, which is used to train the Diffusion Model (DM) to generate product images that differ from the dataset. Secondly, the Kansei factors are obtained by clustering Kansei words from online reviews using Factor Analysis (FA) and then calculating the weights of the Kansei factors by the Analytic Hierarchy Process (AHP). Thirdly, a questionnaire is distributed to obtain user scores on the Kansei factors of the product images, and the Kansei evaluation value is calculated by weighting, then a prediction model is constructed using Support Vector Regression (SVR) to score and filter the generated images. Finally, the designer selects the highest-scoring images for detailing and tests the effectiveness of the design through user satisfaction. Using the ear thermometer as an example, we have created a product form that meets the emotional demands of users and verifies the scientific validity and effectiveness of the proposed method.