基于独立分量分析和遗传算法的人脸性别分类  

Face Gender Classification Based on Independent Component Analysis and Genetic Algorithm

作　　者：王振花[1] 穆志纯[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《计算机科学》2010年第4期197-200,264,共5页Computer Science

基　　金：国家自然科学基金项目(60573058);北京市教委重点学科共建项目(XK100080537)资助

摘　　要：性别是人脸反映的一个重要信息,通过人脸图像实现性别自动分类对大型人脸数据库的检索和识别具有重要意义。提出了一种新的结合独立分量分析(ICA)和遗传算法(GA)的人脸性别分类方法。首先采用快速独立分量分析方法(FastICA)提取人脸图像的独立基图像和投影向量,获得人脸的低维表征;然后通过遗传算法从该低维空间中选择对性别分类有利的特征子集;最后采用支持向量机进行分类。将ICA的空间局部特征提取功能、遗传算法快速寻优的特征选择功能以及SVM的强分类能力有机地结合起来。实验表明,该方法取得了很好的分类性能。The gender of a face is almost its most salient feature,and realizing automatic gender classification according to the face image will boost the performance of face retrieval and face recognition in large face database. This paper proposed a new gender classification method combining independent analysis （ICA） and genetic algorithm （GA）. The Fast- ICA algorithm was used to derive independent basic images and projection vectors out of the face images and each image was represented as a feature vector projected in the low-dimensional space spanned by the basis vectors. Then, a genetic algorithm was used to select a subset of features which seem to encode important information about gender form the low-dimensional representation. Finally, the SVM classifier was trained to perform gender classification using the selected independent-features subset. The local features extraction of ICA, the features selection ability of GA, and the strong classify ability of SVM were combined reasonably. The experiment results show that the method gets a better classifier performance.

关 键 词：人脸性别分类 独立分量分析 遗传算法 支持向量机 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]