...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Journal of Theoretical and Applied Information Technology >COMPARING THE SIMILARITIES MEASUREMENT OF FACE EXPRESSION-RECOGNITION BASED ON 2DLDA MODIFICATION METHOD
【24h】

COMPARING THE SIMILARITIES MEASUREMENT OF FACE EXPRESSION-RECOGNITION BASED ON 2DLDA MODIFICATION METHOD

机译：基于2DLDA修正方法的人脸表情识别相似度测量

获取原文
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Facial expression recognition is the development of face recognition in an environment of pattern recognition (feature recognition). Research on facial expression recognition is very useful in many fields, for example in the field of human computer interaction, in this case the computer recognizes facial expressions of the user, then the computer programmatically perform the appropriate instructions to the facial expression of the user. Facial expressions can also be used as a measure of customer satisfaction with public services. In this study, the facial expression recognition applications were built to measure customer satisfaction with the process of feature extraction using the Modified Two Dimensional Linear Discriminant Analysis (Modified 2DLDA) to obtain input characteristics on each face. 2DLDA modification method is the development of methods 2DLDA; which may have the similarity measurement using Euclidean Distance, Manhattan Distance, and Two Dimensional Correlation Coefficient. The combination of these test methods uses Jaffe database which is a database that contains Japanese female facial expression. The highest test results using the Euclidean Distance is 88.57%, the Manhattan Distance method is 89.92%, and the method Two Dimensional Correlation Coefficient of 90.48%.
机译：面部表情识别是在模式识别（特征识别）环境中面部识别的发展。关于面部表情识别的研究在许多领域中非常有用，例如在人机交互领域，在这种情况下，计算机识别用户的面部表情，然后计算机以编程方式对用户的面部表情执行适当的指令。面部表情也可以用来衡量客户对公共服务的满意度。在这项研究中，构建了面部表情识别应用程序，以使用改进的二维线性判别分析（改进的2DLDA）在特征提取过程中测量客户满意度，以获取每张脸部的输入特征。 2DLDA修改方法是2DLDA方法的发展;使用欧几里得距离，曼哈顿距离和二维相关系数进行相似度测量。这些测试方法的组合使用Jaffe数据库，该数据库包含日本女性面部表情。使用欧几里德距离的最高测试结果为88.57％，曼哈顿距离法为89.92％，二维相关系数法为90.48％。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号