【摘要】 在基于流形正则化的框架下提出了一种半监督学习算法(MLapRLS)并将其用于人脸识别.首先构建所有样本的最近邻图来估计数据空间的几何结构,并对多变量线性回归的目标函数增加该流形正则化项,得到针对多类问题的MLapRLS.该方法能充分利用少量有标签样本和大量易于获取的无标签样本来帮助学习以提取有效特征.在Extended YaleB和CMU PIE人脸数据库上的实验结果证明了该方法的有效性.更多还原

【Abstract】 Considering the semi-supervised learning framework based on manifold regularization,a method called MLapRLS was proposed and applied to face recognition.In MLapRLS,a nearest neighbor graph was constructed first to model the intrinsic geometrical structure of the sample space,and then the graph structure was incorporated into the objective function of the multivariate linear regression as a regularization term.Aiming to extract effective features for the semi-supervised multi-class problem,MLapRLS can make use of both labeled samples and large numbers of easily collected unlabeled samples.Experimental results on extended YaleB and CMU PIE face databases demonstrate the effectiveness of the presented method.更多还原