Large-scale vocabularies with local graph diffusion and mode seeking

作者：

Highlights：

•

• This paper is the first to introduce graph diffusion to very large-scale clustering.

• Our mode-seeking like algorithm finds modes with the defined diffusion process.

• It builds large vocabularies fast as it is not iterative and can be done in parallel.

• It compares favorably to previously clustering methods commonly used in CBIR.

摘要

•This paper is the first to introduce graph diffusion to very large-scale clustering.•Our mode-seeking like algorithm finds modes with the defined diffusion process.•It builds large vocabularies fast as it is not iterative and can be done in parallel.•It compares favorably to previously clustering methods commonly used in CBIR.

论文关键词：Image retrieval,Local graph diffusion,Mode-seeking,Large-scale clustering

论文评审过程：Received 27 May 2017, Revised 3 November 2017, Accepted 16 January 2018, Available online 6 February 2018, Version of Record 6 February 2018.

论文官网地址：https://doi.org/10.1016/j.image.2018.01.004