摘要
随着智能家居的快速发展,国内外学者已经对交互界面进行了多方面的研究.立足虚拟人技术,重点探索智能家居中虚拟人的潜在需求.基于设计师的人物设计经验,将各异的人脸五官特征归类为有限个解剖维度,以此作为设计规则,设计人脸五官的卡通风格化模型模板,比对人脸图像特征构建模板库.在此基础上,结合人脸特征识别及基于美学计算的特征——卡通形象模版库映射关系,实现从二维真实人脸图像到三维卡通人脸的自动生成.该生成方法轻量且能够满足用户的个性化需求,并已投入实际应用.
With the rapid development of smart home,researchers around the world have carried out numerous studies on its interactive interface.Based on virtual human technology,this paper explores the potential de-mand of virtual humans in smart home scenarios.Through the character design experience of designers,dif-ferent facial features are grouped into limited anatomical dimensions as design rules.Under the guidance of these rules,cartoon stylized facial model templates can be made via comparing face image features to build a template library.Combined with face feature recognition as well as the mapping relationship between the face features and the cartoon model template-based aesthetic computing,automatic generation from two-dimnsional real face images to three-dimensional cartoon faces is realized.This generation method is lightweight,meets users’personalized needs,and has been put into practical application.
作者
姚智皓
乔玥涵
徐千尧
Yao Zhihao;Qiao Yuehan;Xu Qianyao(Academy of Arts&Design,Tsinghua University,Beijing 100084;The Future Laboratory,Tsinghua University,Beijing 100084)
出处
《计算机辅助设计与图形学学报》
EI
CSCD
北大核心
2023年第2期221-229,共9页
Journal of Computer-Aided Design & Computer Graphics
基金
国家自然科学基金(62172252)
中国博士后科学基金(2021M701914).
关键词
人机自然交互
虚拟人
智能家居
natural human-computer interaction
virtual human
smart home