此文的目的有三个：第一，当地连续均值量化变换特征是提出照明和传感器敏感操作在目标识别上。其次，注册稀疏Winnows网络分割，提出了加快原分类。最后，特点和分类相结合对于正面人脸检测任务。检测结果列 为MIT + CMU系统和BioID数据库。关于这人脸检测器，接收器操作特征曲线BioID数据库产生最好的结果公布。对于结果麻省理工学院的中央结算系统+数据库相当于国家的最先进的脸探测器。一个人脸检测算法的MATLAB版本可以从http://www.mathworks.com/matlabce

1_On the Dimensionality Reduction for Sparse Representation based Face Recognition 2_Fingerprint Pore Matching based on Sparse Representation 3_Microaneurysm (MA) Detection via Sparse Representation Classifier with MA and Non-MA Dictionary

This paper is a report CMU CVPR2013 article, proposed a sparse encoding based on contour features, referred to as HSC (Histogram of Sparse Code), and in target detection beyond the HOG (Histogram of Gradient) this paper introduces ideas and calculati