摘要
提出一种快速的完备鉴别保局投影算法(FCDLPP).FCDLPP算法只需使用一次瘦QR分解就可求得保局类内散布的零空间的鉴别矢量,然后再进行一次广义特征值分解求得保局类内散布的主元空间的鉴别矢量.另外,FCDLPP对零空间的不规则鉴别特征和主元空间的规则鉴别特征进行融合.理论分析和实验结果表明,FCDLPP算法不论在计算复杂度还是识别率上都比完备的鉴别保局投影算法有更好的性能和效果.
Fast complete discriminant locality preserving projections (FCDLPP) is proposed. There is only one step of economic QR faetorization for FCDLPP algorithm to obtain the optimal diseriminant vectors in the null space of locality preserving within-class scatter. Then, one step of eigen-decomposition is used to obtain the optimal diseriminant vectors in the principal space of the locality preserving within-class scatter. Besides, FCDLPP fuses the regular diseriminant features in the principal space and irregular discriminant features in the null space. Theoretical analyses and experimental results show that the proposed FCDLPP outperforms complete diseriminant locality preserving projections (CDLPP) on computational speed and recognition rates.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2011年第6期804-809,共6页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金项目(No.60873151
60632050)
江苏省普通高校研究生科研创新计划项目(No.178)资助
关键词
人脸识别
完备鉴别保局投影(CDLPP)
QR分解
Face Recognition, Complete Discriminant Locality Preserving Projections (CDLPP), QR factorization