Handwritten Chess Scoresheet Dataset (HCS)

Research Tasks

Latin Handwriting Recognition in Chess Scoresheets

2021-07-04 (v. 1)

Contact author

Owen Eicher

Colorado School of Mines



This work is licensed under a Creative Commons Attribution 3.0 Unported License.


We live in an era where deep learning can be applied to a vast number of complex problems, and yet surprisingly little work has been done on handwritten chess move recognition. A handwriting recognition framework is particularly useful with chess since players are bound to record moves by hand in order to prevent computer aided cheating. Chess is one of the most popular board game today and its popularity is growing faster than ever. This dataset seeks to catalyze future research and development in handwriting recognition with a focus on chess scoresheets.


Character recognition accuracy and word recognition accuray are both decent metrics for successful handwriting recognition models. These are found by calculating the percentage of correctly identified characters and words by a given model respectively. This dataset includes ground truth labels for all moves to assist in calculating these values.


No comments on this dataset yet.

Add your comment

In order to comment on a dataset you need to be logged on
Register Now!


In order to rate this dataset you need to be logged on
Register Now!