首页> 外文会议>International conference on electronic engineering, communication and management;EECM 2011 >Facial Expression Recognition Using Kernel Locality Preserving Projections
【24h】

Facial Expression Recognition Using Kernel Locality Preserving Projections

机译:使用内核局部性保留投影的面部表情识别

获取原文

摘要

Considering the nonlinear manifold structure of facial expression images, a new facial expression recognition method based on kernel locality preserving projections (KLPP) is proposed in this paper. KLPP is used to extract the low-dimensional embedded data representations from the original extracted high-dimensional local binary patterns (LBP) facial features. The experimental results on the popular Cohn-Kanade facial expression database demonstrate that KLPP obtains the best accuracy of 86.94%, outperforming the other used methods including locality preserving projections (LPP) and kernel Isomap (KIsomap).
机译:针对面部表情图像的非线性流形结构,提出了一种基于核局部保留投影(KLPP)的面部表情识别新方法。 KLPP用于从原始提取的高维局部二进制模式(LBP)面部特征中提取低维嵌入式数据表示。在流行的Cohn-Kanade面部表情数据库上的实验结果表明,KLPP的最佳准确性为86.94%,优于其他使用的方法,包括局部性保留投影(LPP)和内核Isomap(KIsomap)。

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