基于GAN的手绘草图图像翻译研究综述  

Research on freehand sketch to image translation based on generative adversarial networks

作　　者：王建欣 史英杰[2] 刘昊 黄海峤[4] 杜方 Wang Jianxin;Shi Yingjie;Liu Hao;Huang Haiqiao;Du Fang(School of Arts&Sciences,Beijing Institute of Fashion Technology,Beijing 100029,China;School of Business,Beijing Institute of Fashion Technology,Beijing 100029,China;Fashion Accessory Art&Engineering College,Beijing Institute of Fashion Technology,Beijing 100029,China;School of Fashion,Beijing Institute of Fashion Technology,Beijing 100029,China;School of Information Engineering,Ningxia University,Yinchuan 750021,China)

机构地区：[1]北京服装学院文理学院,北京100029 [2]北京服装学院商学院,北京100029 [3]北京服装学院服饰艺术与工程学院,北京100029 [4]北京服装学院服装艺术与工程学院,北京100029 [5]宁夏大学信息工程学院,银川750021

出　　处：《计算机应用研究》2022年第8期2249-2256,共8页Application Research of Computers

基　　金：国家自然科学基金资助项目(61502279,62062058);北京服装学院重点科研项目(2021A-02);北京服装学院青年拔尖人才培养计划资助项目(YS22-1005096);北京市服装产业数字化工程技术研究中心科研项目(KJCX20801-30299/016)。

摘　　要：目前基于GAN的手绘草图图像翻译工作处于起步阶段,分析了草图图像翻译面临的挑战性问题,从无控制和精细控制的草图图像翻译两个方面对基于GAN的草图图像翻译研究工作进行分析,并对生成图像的评估方法进行总结。基于已有研究工作的总结归纳,对该领域未来可能的发展趋势进行了展望,为该领域研究人员拓展研究思路提供了线索。At present,sketch to image translation based on GAN is in its infancy.This paper analyzed the challenging problems on sketch to image translation,and summarized the work based on GAN from two aspects of uncontrolled sketch to image translation and finely controlled sketch to image translation.This paper also summarized the method of evaluating generative image.Based on the summary of the existing research,this paper proposed the possible future development trends in this field,which provides clues for researchers to expand their research ideas.

关 键 词：手绘草图 图像翻译 生成式对抗网络 图像合成 解耦 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]