期刊文献+

线性回归分析在人脸识别中的应用研究

Research and Application in Face Recognition with Linear Regression Classification
下载PDF
导出
摘要 线性回归分类(LRC)算法中,借助一个依赖线性子空间的单一对象类模型,开发了一个线性模型,作为特定类库的线性组合来描述探针图像,并且借助于最小二乘法及其为了支持具有最小重构误差的类而制定的决策,解决了逆问题,但是并不能解决连续闭塞问题。基于此,提出了一种新颖的基于近邻子空间分类的识别方法,模块化线性回归分类(MLRC)算法。将LRC算法进行模块化,并且引入了一种基于距离的本征融合(DEF)算法用于决策。在FERET及ORL上的实验表明,与其它几种常用的方法相比较,MLRC算法在处理人脸识别问题上取得了更好的结果。 A novel approach of face identification by formulating the pattern recognition problem in terms of linear regression is presented.Using a fundamental concept that patterns from a single-object class lie on a linear subspace,a linear model representing a probe image as a linear combination of class-specific galleries is developed.The inverse problem is solved using the least-squares method and the decision is ruled in favor of the class with the minimum reconstruction error.The proposed Linear Regression Classification(LRC) algorithm falls in the category of nearest subspace classification.The algorithm is extensively evaluated on several standard databases under a number of exemplary evaluation protocols reported in the face recognition literature.A comparative study with state-of-the-art algorithms clearly reflects the efficacy of the proposed approach.For the problem of contiguous occlusion,a Modular LRC approach is proposed,introducing a novel Distance-based Evidence Fusion(DEF) algorithm.Experiments on the FERET and a challenging passport face database shows that the proposed method can achieve better results compared with other common solutions to the human face recognition problem.
作者 易月娥 宋娟
出处 《科学技术与工程》 北大核心 2013年第17期4994-4998,共5页 Science Technology and Engineering
基金 湖南省科技厅科学研究项目(2012GK3063) 湖南省科技资助计划博士后专项(2012RS4027)资助
关键词 人脸识别 线性回归分类 近邻子空间分类 子空间学习 face recognition linear regression classification nearest subspace classification subspace learning
  • 相关文献

参考文献9

  • 1Zhang Z,Wang J, Zha H. Adaptive manifold learning. IEEE Trans-actions on Pattern Analysis and Machine Intelligence, 2012; 34( 1):131—137.
  • 2Chen S,Liu J, Zhou Z. Making FLDA applicable to face recognitionwith one sample per person. Pattern Recognition, 2007 ; 37 ( 7 ):1553—1555.
  • 3sparse representation. IEEE Trans Pattern Analysis and Machine In-telligence, 2009; 31(2) :210—227.
  • 4Yan S, Liu J, Tang X,et al. A parameter-free framework for generalsupervised subspace learning. IEEE Transactions on Information Fo-rensics and Security, 2007 ;2( 1 ) :69-76.
  • 5Connolly J F,Granger E, Sabourin R. An adaptive classification sys-tem for video-based face recognition. Information Sciences, 2012 ;192(1) :50—70.
  • 6Haiiz F, Shafie A A, Mustafah Y M. Face recognition from singlesample per person by learning of generic discriminant vectors. Proce-.
  • 7Xie Z,Liu G,Fang Z, Face recognition based on combination of hu-man perception and local binary pattern. Lecture Notes in ComputerScience, 2012; 7202:365—373.
  • 8Jiang X, Mandal B, Kot A. Eigenfeature regularization and extractionin face recognition. IEEE Trans. Pattern Analysis and Machine Intel-ligence, 2008; 30(3) :383—394.
  • 9Wang R, Shan S, Chen X,et al. Manifold-manifold distance and itsapplication to face recognition with image sets. IEEE Transactions onPattern Analysis and Machine Intelligence, 2012; 21(10):4466-4479.

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部