基于生成对抗网络的健身榨汁杯设计研究  

Research on the Design of Fitness Juicing Cup based on Generative Adversarial Network

作　　者：周宏伟[1] 张人水 Zhou Hongwei;Zhang Renshui(College of Art and Design,Beijing University of Technology,Bejing 100124,China)

机构地区：[1]北京工业大学艺术设计学院,北京100124

出　　处：《艺术与设计（理论版）》2023年第4期123-126,共4页Art and Design

摘　　要：文章为生成对抗网络与工业设计结合探索创新解决思路,通过研究AI生成对抗网络衍生模型,以健身榨汁杯为设计实践,提出融合运用不同衍生模型的设计模式。运用文献调研法梳理生成对抗网络衍生模型特点,通过健身榨汁杯设计流程进行设计验证。设计验证部分,结合市场调研与用户调研的数据应用Stack GAN与Cycle GAN确定产品功能与外观设计,基于问卷调研与用户反馈结果应用Deep Exampler-based与Style GAN完成系列产品配色设计及理想模特生成。文章总结了工业设计中生成对抗网络衍生模型的多种应用模式,最终通过实例验证该算法的有效性、可行性,为工业设计实践提供参考。The article explores innovative solutions for the combination of generative adversarial networks and industrial design.By studying the derivative models of Al generative adversarial networks,the article proposes a design model that integrates the use of different derivative models by using a fitness juice cup as a design practice.The paper uses literature research to sort out the characteristics of the derivative models of generative adversarial networks and the design verification through the design process of fitness juice cup.In the design validation part,StackGAN and CycleGAN are applied to determine the product function and appearance design based on the data of market research and user research,and Deep Exampler-based and StyleGAN are applied to complete the color scheme and ideal model generation based on the results of questionnaire research and user feedback.The article summarizes various application modes of generative adversarial network-derived models in industrial design,and finally,the efficiency and feasibility of the proposed algorithm will be verified by example to provide reference for industrial design practice.

关 键 词：生成对抗网络 工业设计 人工智能 榨汁杯 计算机辅助设计 

分 类 号：TB472[一般工业技术—工业设计]