Preserving similarity order for unsupervised clustering

作者：

Highlights：

• Our method takes the ordering of pairwise distance as the supervisory signal to learn the similarity score function.

• Our similarity score function captures both local structure and global structure of the data sample distribution.

• We propose a simple but effective strategy to identify the boundary samples from a given dataset.

摘要

•Our method takes the ordering of pairwise distance as the supervisory signal to learn the similarity score function.•Our similarity score function captures both local structure and global structure of the data sample distribution.•We propose a simple but effective strategy to identify the boundary samples from a given dataset.

论文关键词：Image clustering,Order preserving,Deep representation learning,Score function learning

论文评审过程：Received 20 May 2019, Revised 16 March 2022, Accepted 25 March 2022, Available online 27 March 2022, Version of Record 2 April 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108670