Boosting offline handwritten text recognition in historical documents with few labeled lines

Abstract

In this paper, we face the problem of offline handwritten text recognition (HTR) in historical documents when few labeled samples are available and some of them contain errors in the train set. Three main contributions are developed. First we analyze how to perform transfer learning (TL) from a massive database to a smaller historical database, analyzing which layers of the model need a finetuning process. Second, we analyze methods to efficiently combine TL and data augmentation (DA). Finally, an algorithm to mitigate the effects of incorrect labelings in the training set is proposed. The methods are analyzed over the ICFHR 2018 competition database, Washington and Parzival. Combining all these techniques, we demonstrate a remarkable reduction of CER (up to 6% in some cases) in the test set with little complexity overhead.

Recognition and Information Extraction in Historical Handwritten Tables: Toward Understanding Early $$20^{th}$$ Century Paris Census

“…With only 4 annotated pages (3 for training and 1 for validation), we reach an average CER of 5.11%. Then, the more we annotate data of a target dataset, the less improvement we get by new annotated lines which confirms the results obtained in [2]. The next challenge to address will be to minimize the number of new lines needed to benefit from writer specialization.…”