基于判别公共向量的单训练样本人脸识别  

Making discriminative common vectors applicable to face recognition with one training image per person

作　　者：李瑞东[1] 祝磊[2] 余党军[1] 陈偕雄[3] 

机构地区：[1]金华职业技术学院信息工程学院,浙江金华321017 [2]浙江大学电气工程学院,浙江杭州310027 [3]浙江大学信息学院,浙江杭州310028

出　　处：《浙江大学学报（理学版）》2008年第2期181-184,共4页Journal of Zhejiang University（Science Edition）

基　　金：浙江省自然科学基金资助项目(Y106164)

摘　　要：提出一种新的基于判别公共向量(Discriminative Common Vector)的单样本人脸识别算法.该方法基于人脸类内方差相似的假定,通过引入单人多样本的辅助人脸集来估计类内方差,解决了单训练样本情况下样本类内方差无法估计的问题.在FERET人脸库的测试结果表明,在面部细节、光照、表情变化的情况下,该方法都具有较好的识别效果.At present, there are many methods for frontal view face recognition. However, few of them can work well when only one sample image per class is available. A new method based on Discriminative Common Vector in case of one training sample scenario was presented. Our approach is based on the assumption that human faces share similar intrapersonal variation. The intrapersonal variation of the training set can be estimated from the collected generic face set. The proposed method was tested and evaluated by using the FERET face database. Nearest neighborhood （NN） algorithm was used to construct classifiers. The experimental results showed that the proposed method has good performance when facial detail, illumination condition, and face expression change.

关 键 词：判别公共向量 单训练样本 人脸识别 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]