Keyword spotting in historical handwritten documents based on graph matching
作者：
Highlights：
• We present a novel Keyword Spotting (KWS) solution to improve the accessibility to digitalised historical handwritten documents with respect to browsing and searching.
• We introduce four novel graph representations to describe the topological characteristics of handwritten words.
• Our graph-based KWS approach can keep up or outperform several state-of-the-art template-based as well as learning-based matching approaches on four benchmark datasets.
摘要
•We present a novel Keyword Spotting (KWS) solution to improve the accessibility to digitalised historical handwritten documents with respect to browsing and searching.•We introduce four novel graph representations to describe the topological characteristics of handwritten words.•Our graph-based KWS approach can keep up or outperform several state-of-the-art template-based as well as learning-based matching approaches on four benchmark datasets.
论文关键词：Handwritten keyword spotting,Graph representation,Bipartite graph matching,Ensemble methods
论文评审过程：Received 4 September 2017, Revised 6 January 2018, Accepted 2 April 2018, Available online 6 April 2018, Version of Record 13 April 2018.
论文官网地址：https://doi.org/10.1016/j.patcog.2018.04.001
