作者: Jinghua Wang , Qi Zhu , Zizhu Fan , Yong Xu
DOI:
关键词:
摘要: In this paper, we extend the idea of sparse representation into high dimensional feature space induced by kernel function, and propose a based test sample classification algorithm (KTSRC) for first time. The KTSRC is on assumption that can be linearly represented part training samples in space. Although explicit form unknown, implement trick. experimental results show achieves promising performance face recognition, outperforms state-of-the-art methods.