|
|||||
|
|
| 稀疏表示的人脸识别及其优化算法 | |
| 引用本文： | 郑轶,蔡体健. 稀疏表示的人脸识别及其优化算法[J]. 华东交通大学学报, 2012, 0(1): 10-14 |
| 作者姓名： | 郑轶 蔡体健 |
| 作者单位： | 华东交通大学轨道交通学院;华东交通大学信息工程学院;中南大学信息科学与工程学院 |
| 基金项目： | 国家自然科学基金项目(61165007);华东交通大学科研项目(09111004) |
| 摘 要： | 稀疏表示的本质就是稀疏正规化约束下的信号分解。提出一种改进的正交匹配追踪算法,使运算量较高的矩阵求逆运算转变为轻量级的向量运算或向量与矩阵的运算,可以加快逆矩阵和大矩阵乘积的求解。将此算法应用于稀疏表示的人脸识别,探讨并验证了稀疏阀值的设置和训练字典的选择对人脸识别率和识别速度的影响。 |
| 关 键 词： | 压缩感知 人脸识别 稀疏表示 快速正交匹配追踪算法 特征提取 |
Face Recognition Based on Sparse Representation and Optimization Algorithms |
|
| Zheng Yi,Cai Tijian. Face Recognition Based on Sparse Representation and Optimization Algorithms[J]. Journal of East China Jiaotong University, 2012, 0(1): 10-14 | |
| Authors: | Zheng Yi Cai Tijian |
| Affiliation: | 2,3(1.School of Railway Tracks and Transportation,East China Jiaotong University,Nanchang 330013,China;2.School of Information Engineering,East China Jiaotong University,Nanchang 330013,China;3.School of Information and Science Engineering,Central South University,Changsha 410075,China) |
| Abstract: | The essence of sparse representation is the signal decomposition under the constraint of sparse regularized.This paper proposes an improved orthogonal matching pursuit algorithm,which would convert the high complexity operations of matrix into the lightweight operation of vector or vector-matrix computing.This improved method could accelerate the solution of inverse matrix and large matrix multiplication.This paper applies the algorithm to face recognition based on sparse representation,exploring and validating the effect of the sparse valve setting and training of the dictionary on face recognition accuracy and recognition speed. |
| Keywords: | compressed sensing face recognition sparse representation orthogonal matching pursuit algorithm feature extraction |
| 本文献已被 CNKI 维普 等数据库收录！ | |