Learning local metrics from pairwise similarity data

作者：

Highlights：

• A new local metric learning algorithm is proposed.

• The method’s flexibility allows its use in a large variety of scenarios.

• The method outperforms or is competitive with the state-of-the-art methods on several real datasets.

摘要

•A new local metric learning algorithm is proposed.•The method’s flexibility allows its use in a large variety of scenarios.•The method outperforms or is competitive with the state-of-the-art methods on several real datasets.

论文关键词：Similarity function learning,Local metric learning,Nearest neighbors classification,Face verification

论文评审过程：Received 22 July 2016, Revised 13 March 2017, Accepted 4 April 2017, Available online 5 April 2017, Version of Record 21 November 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.04.002