基于稀疏表示的人脸姿态估计研究  

Face Poses Estimation Based on Sparse Representation

作　　者：廖海斌[1] 邱益鸣 陈庆虎[2] 

机构地区：[1]湖北科技学院计算机科学与技术学院,湖北咸宁437100 [2]武汉大学电子信息学院,湖北武汉430072

出　　处：《电视技术》2015年第13期40-44,共5页Video Engineering

基　　金：国家自然科学基金项目(61271256);河南省重大科技攻关项目(072SGZS38042);湖北科技学院博士启动基金项目(BK1418)

摘　　要：针对人脸光照、遮挡、身份、表情等因素变化的人脸姿态估计难题,结合稀疏表示分类(SRC)方法的优秀识别性能,对SRC理论进行了深入分析,并将其应用于人脸姿态分类。为了解决姿态估计中人脸光照、噪声和遮挡变化问题,将人脸姿态离散化为不同的子空间,每个子空间对应一个类别,据此,提出基于字典学习与稀疏约束的人脸姿态识别方法。通过在公开的XJTU和PIE人脸库上实验表明:所研究的方法对人脸光照、噪声和遮挡变化具有鲁棒性。According to the challenges in face pose estimation under different illuminations, occlusions, identity, expressions, and so on, combining with the excellent classification performance of sparse representation classification （ SRC ）, a deep analysis on the theory of SRC and its application in face pose classification are made. In order to handle challenges such as variation of face illumination, noises and occlusion, a robust face pose estimation method based on dictionary learning and sparse representation is presented. In which face poses are discrete into different subspaces, each subspace corresponding to a class. Several experiments are performed on XJTU and PIE databases. Recognition results show that the proposed method is suitable for efficient face pose recognition under illumination, noises and occlusion variations.

关 键 词：人脸姿态估计 稀疏表示 子空间学习 人脸识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]