Keyword spotting in unconstrained handwritten Chinese documents using contextual word model

作者：

Highlights：

• We propose a contextual word model for keyword spotting from handwritten Chinese documents.

• The contextual word model combines character classifier, geometric and linguistic contexts.

• Promising results were obtained on a large handwriting database CASIA-HWDB.

• The geometric and linguistic contexts improve the spotting performance significantly.

摘要

•We propose a contextual word model for keyword spotting from handwritten Chinese documents.•The contextual word model combines character classifier, geometric and linguistic contexts.•Promising results were obtained on a large handwriting database CASIA-HWDB.•The geometric and linguistic contexts improve the spotting performance significantly.

论文关键词：Keyword spotting,Chinese handwritten documents,Word similarity,Contextual word model

论文评审过程：Received 13 June 2012, Revised 22 August 2013, Accepted 10 October 2013, Available online 21 October 2013.

论文官网地址：https://doi.org/10.1016/j.imavis.2013.10.003