Label noise correction and application in crowdsourcing

作者：

Highlights：

• This paper focuses on label noise correction, which keeps silence in literature.

• A self-training based noise correction method is proposed.

• A clustering-based correction method is proposed.

• A previous method is adapted for label noise correction only.

• Investigate the performance of noise correction algorithms in crowdsourced image labeling.

摘要

•This paper focuses on label noise correction, which keeps silence in literature.•A self-training based noise correction method is proposed.•A clustering-based correction method is proposed.•A previous method is adapted for label noise correction only.•Investigate the performance of noise correction algorithms in crowdsourced image labeling.

论文关键词：Noise correction,Crowdsourcing,Image processing,Classification

论文评审过程：Received 11 January 2016, Revised 2 September 2016, Accepted 3 September 2016, Available online 9 September 2016, Version of Record 14 September 2016.

论文官网地址：https://doi.org/10.1016/j.eswa.2016.09.003