An improved distance-based relevance feedback strategy for image retrieval

作者：

Highlights：

• The performance of existing nearest neighbor approaches is studied in a CBIR context.

• Several issues on the application of NN methods to the CBIR problem are analyzed.

• An improved relevance feedback algorithm also based on the NN paradigm is proposed.

• The accuracy of the estimates is improved by considering locality of labeled samples.

• Experimental results evidence significant improvements in most cases.

摘要

•The performance of existing nearest neighbor approaches is studied in a CBIR context.•Several issues on the application of NN methods to the CBIR problem are analyzed.•An improved relevance feedback algorithm also based on the NN paradigm is proposed.•The accuracy of the estimates is improved by considering locality of labeled samples.•Experimental results evidence significant improvements in most cases.

论文关键词：CBIR,Image retrieval,Relevance feedback,Nearest neighbor

论文评审过程：Received 18 December 2012, Revised 13 May 2013, Accepted 18 July 2013, Available online 27 July 2013.

论文官网地址：https://doi.org/10.1016/j.imavis.2013.07.004