ICFHR2014 Competition on Word Recognition from Historical Documents: ANncestry Word REcognition from Segmented Historical Documents (ANWRESH)

This paper describes the ANWRESH (Anncestry Word Recognition from Segmented Historical Documents)competition held at the 14th International Conference on Frontiers in Handwriting Recognition (ICFHR-2014). This competition uses the ANWRESH dataset selected from the 1930 US Census collection including word bounding box and field lexicon data. Five teams submitted systems for recognizing six fields including Surname, Given Name(s), Age, Birth Place, Marital Status, and Relation. Results are reported, along with brief descriptions of each system.