摘要
研究文本智能生成图像所蕴含的艺术性及其背后的重构与复制,进而探讨“艺术创作与艺术复制”所引发的问题与解决方案。从现有的文本智能生成图像技术切入,梳理当前的主要应用方法,提炼出技术特点。在此基础上探索文本智能生成图像技术的创作过程、生成图像特点和艺术特性。从纵向角度与传统图像、数字图像进行比较分析,探讨人与人工智能在图像创作中因主体差异性而产生的创作过程的差异。最终探讨文本智能生成图像在艺术重构与艺术复制过程中的创造性与复制性问题。文本智能生成图像是对人类的艺术性创作的深度学习,是对人类已有艺术作品的重构,但是也无法避免对人类已有艺术作品的复制,创作行为与复制行为之间的合理性成为该技术亟待解决的问题。
This paper aims to study the artistry contained in textual intelligent image generation works and the reconstruction and reproduction behind it,and then to explore the problems and solutions caused by“artistic creation and artistic reproduction”.Starting from the existing text intelligent image generation technology,combing the current main application methods and extracting the technical characteristics.On this basis,this paper explores the creation process,the characteristics of the generated image and the artistic characteristics of the text intelligent image generation technology.From a longitudinal perspective,it compares and analyzes with traditional images and digital images,and explores the differences in the creation process caused by the main differences between humans and artificial intelligence in image creation.Finally,it discusses the creativity and reproducibility of text intelligent image generation technology in the process of artistic reconstruction and artistic reproduction.The intelligent image generation technology of text is a deep learning of human artistic creation,and a reconstruction of human existing artistic works,but it cannot avoid the duplication of human existing artistic works.The rationality between creation behavior and replication behavior has become an urgent problem to be solved by this technology.
作者
刘佳
萧惠丹
LIU Jia;XIAO Huidan(School of Arts&Communication,Beijing Normal University,Beijing 100875,China)
出处
《工业工程设计》
2021年第5期60-64,100,共6页
INDUSTRIAL&ENGINEERING DESIGN
关键词
人工智能
文本生成图像
可复制性
DALL·E
数字化图像
artificial intelligence
text intelligent image generation technology
replicability
DALL·E
digitized image
