摘要
微表情自动分析是计算机视觉研究方向之一。在刑侦、临床医学、商业谈判、公共安全等场景下的微表情分析技术具有重要研究和应用价值。为了梳理微表情自动分析领域研究现状及发展方向,对常用微表情数据集和数据预处理方法进行整理。基于微表情特征,全面整理和对比微表情检测和识别任务各类算法以及实验方法和验证指标。可以帮助研究人员更加快捷、全面了解该领域研究现状,存在的问题和未来发展方向。
Micro-expression automatic analysis method is one of the research directions of computer vision.Micro-expression analysis technology has important research and application value in criminal investigation,clinical medicine,business negotiation,public security and other scenarios.In order to sort out the research status and development direction in the field of automatic micro-expression analysis,this paper sorted out the commonly used micro-expression datasets and data preprocessing methods.Based on micro-expression features,various algorithms,this paper comprehensively sorted and compared experimental methods and verification indicators for micro-expression spotting and recognition tasks.It can help researchers more quickly and comprehensively understand the current research status,existing problems and future development directions in this field.
作者
周伟航
肖正清
钱育蓉
马玉民
公维军
帕力旦·吐尔逊
Zhou Weihang;Xiao Zhengqing;Qian Yurong;Ma Yumin;Gong Weijun;Palladium Turson(Software College,Xinjiang University,Urumqi 830046,China;Key Laboratory of Software Engineering,Xinjiang University,Urumqi 830046,China;Key Laboratory of Signal Detection&Processing in Xinjiang Uygur Auto-nomous Region,Xinjiang University,Urumqi 830046,China;Mathematics&Systems Science College,Xinjiang University,Urumqi 830046,China;Xinjiang Normal University,Urumqi 830000,China)
出处
《计算机应用研究》
CSCD
北大核心
2022年第7期1921-1932,共12页
Application Research of Computers
基金
国家自然科学基金资助项目(61966035)
国家自然科学基金联合基金重点项目(U1803261)
新疆科技厅国际合作项目(2020E01023)
新疆第三次科考项目(2021xjkk1404)。
关键词
微表情识别
微表情检测
微表情数据集
micro-expression recognition
micro-expression spotting
micro-expression datasets
