基于流形学习和稀疏约束的快速特征提取算法  

Fast Feature Extraction Algorithm Based on Manifold Learning and Sparsity Constraints

作　　者：任迎春[1,2] 王志成[1] 陈宇飞[1] 赵卫东[1] 彭磊[3] REN Ying-chun WANG Zhi-cheng CHEN Yu-fei ZHAO Wei-dong PENG Lei(CAD Research Center, Tongii University, Shanghai 201804, China College of Mathematics,Physics and Information Engineering,Jiaxing University,Jiaxing 314001,China College of Information Engineering,Taishan Medical University,Taian 271016,China)

机构地区：[1]同济大学CAD研究中心,上海201804 [2]嘉兴学院数理与信息工程学院,嘉兴314001 [3]泰山医学院信息工程学院,泰安271016

出　　处：《计算机科学》2016年第8期277-281,296,共6页Computer Science

基　　金：国家自然科学基金(61103070;11301226);浙江省自然科学基金(LQ13A010017);山东省自然科学基金(ZR2015FL005)资助

摘　　要：针对稀疏保持投影算法在特征提取过程中无监督和L1范数优化的计算量较大的问题,提出一种基于流形学习和稀疏约束的快速特征提取算法。首先通过逐类PCA构造级联字典,并基于该字典通过最小二乘法快速学习稀疏保持结构;其次构造用于描述不同子流形距离的局部类间散度函数;然后整合所学习到的稀疏表示信息和局部类间散度信息以达到既考虑判别效率又保持稀疏表示结构的目的;所提算法最终转化为一个求解广义特征值问题。在公共人脸数据库(Yale,ORL和Extended Yale B)中的测试结果验证了该方法的可行性和有效性。Aiming at the problems of being unsupervised and time-consuming of L1 norm optimization in the existing sparsity preserving projection, by integrating the sparse representation information with the manifold structure of the data, a novel algorithm for fast feature extraction, named sparsity preserving discriminative learning （SPDL）, was pro- posed. SPDL first creates a concatenated dictionary by class-wise PCA decompositions and learns the sparse representa- tion structure of each sample under the constructed dictionary using the least square method. Secondly, a local between- class separability function is defined to characterize the scatter of the samples in different sub-manifolds. Then SPDL in- tegrates the learned sparse representation information with the local between-class relationship to construct a discrimi- nant function. Finally, the proposed method is transformed into a problem of solving the generalized eigenvalue. Exten- sive experimental results on several public face databases demonstrate the effectiveness of the proposed approach.

关 键 词：特征提取 稀疏表示 主元分析 流形学习 人脸识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]