正交指数约束的平滑非负矩阵分解方法及应用  

Smooth non-negative matrix factorization with orthogonal exponent constraints and its applications

作　　者：同鸣[1] 张伟[1] 吴扬成[1] 

机构地区：[1]西安电子科技大学电子工程学院,陕西西安710071

出　　处：《系统工程与电子技术》2013年第10期2221-2228,共8页Systems Engineering and Electronics

基　　金：国家自然科学基金(61072110);陕西省自然科学基金(SJ08F15);陕西省工业攻关项目(2010K06-0)资助课题

摘　　要：提出了一种正交指数约束的平滑非负矩阵分解方法,该方法将非负矩阵分解为基矩阵、列归一化平滑矩阵和系数矩阵之积,同时在目标函数中加入了正交指数约束,保证了低维特征的非负性和局部化,减小了分解误差,提高了稀疏性的调节能力。将该方法应用于数据降维、特征稀疏性比较、有遮挡人脸识别和视频运动特征提取。实验结果表明,该方法比同类方法具有更好的性能。A smooth non-negative matrix factorization with orthogonal exponent constraints is proposed. In this paper, the non-negative matrix is decomposed into the product of base matrix, column normalized smooth matrix and the coefficient matrix. At the same time, the objective function is added to the orthogonal exponent constraints, which ensures that the low dimensional characteristics are of the non-negativity and localization, re- duces the decomposition error, and improves the ability of sparseness adjustment. The proposed method is used to data dimension reduction, feature sparseness comparison, occluded face recognition and video motion feature extraction. The experimental results show that the proposed method has a better performance than other similar methods.

关 键 词：非负矩阵分解 稀疏特征 正交约束 特征提取 有遮挡人脸识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]