Linear reconstruction measure steered nearest neighbor classification framework

作者：

Highlights：

• Introduce and provide a theoretical justification for linear reconstruction measure (LRM).

• Analyze the role of regularization items in regularized LRM.

• Analyze the advantages of LRM over the conventional point-to-point measure (C-PtP).

• Present the LRM steered nearest neighbor classification framework (LRM_NNCF).

摘要

•Introduce and provide a theoretical justification for linear reconstruction measure (LRM).•Analyze the role of regularization items in regularized LRM.•Analyze the advantages of LRM over the conventional point-to-point measure (C-PtP).•Present the LRM steered nearest neighbor classification framework (LRM_NNCF).

论文关键词：Linear reconstruction measure (LRM),Pattern classification,Classifier,Match learning,Face recognition

论文评审过程：Received 6 September 2012, Revised 22 August 2013, Accepted 16 October 2013, Available online 4 November 2013.

论文官网地址：https://doi.org/10.1016/j.patcog.2013.10.018