Locality-constrained collaborative representation based discriminative projection

作者：Wei Lai^*; Xu Feifei

来源：Journal of Computational Information Systems, 2014, 10(8).

DOI：10.12733/jcis9801

• Better performance
• Collaborative representations
• Data set
• Face image database
• Local structure
• Related algorithms
• Subspace learning

Collaborative representation based classifier (CRC) has been successfully applied to pattern classification. In this paper, we proposed a locality-constrained collaborative representation based discriminant projection (LCRCDP) for feature extraction. LCRCDP seeks a subspace in which the with-class reconstruction residual of a given data set is minimized and the between-class reconstruction residual is maximized. Then CRC can achieve better performance in the obtained subspace. The experiments on face image databases including ORL, YALE and CMU PIE show that LCRCDP is superior to the related algorithms.

• 单位
  上海海事大学