摘要
在生成式AI爆发的今天,探索生成式人工智能参与设计领域刻不容缓。文章旨在研究如何提升共情设计理念下的AIGC生成内容的准确性。首先介绍共情设计的概念和国内外研究情况,然后从AIGC图像生成的概念、技术发展以及挑战等方面展开讨论,探究AIGC在共情设计领域的潜在应用。文章通过提出共情设计预训练模型方法,并详述了优化提示词输入的逻辑探讨以及图像扩散模型的训练方法,构建提高生成内容准确性的方法。
With the explosion of Generative AI,it is urgent to explore Artificial Intelligence Generated Content participation in the design field.The article aims to investigate how to improve the accuracy of AIGC generated content under the concept of Empathic design.It first introduces the concept of Empathic design and domestic and international research,and then discusses the concept,technological development,and challenges of AIGC image generation to explore the potential applications of AIGC in the field of Empathic design.The article constructs a method to improve the accuracy of generated content by proposing a pre-training model method for Empathic design and detailing a logical exploration of optimising cue word inputs as well as a training method for image diffusion models.
作者
罗洁
丁干干
Luo Jie;Ding Qianqian(College of Fine Arts and Design,Guangzhou University,Guangzhou 510006,China)
关键词
共情设计
生成式人工智能(AIGC)
预训练模型
提示词
扩散模型
empathic design
Artificial Intelligence Generated Content(AIGC)
pre-trained models
cue words
diffusion models