%0 Journal Article %T A Study of Bag-of-Visual-Words Representations for Handwritten Keyword Spotting %A David Aldavert %A Marçal Rusiñol %A Ricardo Toledo %A Josep Llados %J International Journal on Document Analysis and Recognition %D 2015 %V 18 %N 3 %I Springer Berlin Heidelberg %@ 1433-2833 %F David Aldavert2015 %O DAG; ADAS; 600.055; 600.061; 601.223; 600.077; 600.097 %O exported from refbase (http://refbase.cvc.uab.es/show.php?record=2679), last updated on Thu, 28 Mar 2019 09:49:58 +0100 %X The Bag-of-Visual-Words (BoVW) framework has gained popularity among the document image analysis community, specifically as a representation of handwritten words for recognition or spotting purposes. Although in the computer vision field the BoVW method has been greatly improved, most of the approaches in the document image analysis domain still rely on the basic implementation of the BoVW method disregarding such latest refinements. In this paper, we present a review of those improvements and its application to the keyword spotting task. We thoroughly evaluate their impact against a baseline system in the well-known George Washington dataset and compare the obtained results against nine state-of-the-art keyword spotting methods. In addition, we also compare both the baseline and improved systems with the methods presented at the Handwritten Keyword Spotting Competition 2014. %K Bag-of-Visual-Words %K Keyword spotting %K Handwritten documents %K Performance evaluation %U http://dx.doi.org/10.1007/s10032-015-0245-z %P 223-234