Training Strategies for OCR Systems for Historical Documents
Artificial Intelligence Applications and Innovations (2019)
Research topics:
Dialogue Act RecognitionAbstract
This paper presents an overview of training strategies for optical character recognition of historical documents. The main issue is the lack of the annotated data and its quality. We summarize several ways of synthetic data preparation. The main goal of this paper is to show and compare possibilities how to train a convolutional recurrent neural network classifier using the synthetic data and its combination with a real annotated dataset.
Authors
BibTex
@InProceedings{10.1007/978-3-030-19823-7_30, author = "Mart{\'i}nek, Ji{\v{r}}{\'i} and Lenc, Ladislav and Kr{\'a}l, Pavel", editor = "MacIntyre, John and Maglogiannis, Ilias and Iliadis, Lazaros and Pimenidis, Elias", title = "Training Strategies for OCR Systems for Historical Documents", booktitle = "Artificial Intelligence Applications and Innovations", month = "24-26 May", year = "2019", publisher = "Springer International Publishing", address = "Cham", pages = "362--373", doi = "10.1007/978-3-030-19823-7_30", abstract = "This paper presents an overview of training strategies for optical character recognition of historical documents. The main issue is the lack of the annotated data and its quality. We summarize several ways of synthetic data preparation. The main goal of this paper is to show and compare possibilities how to train a convolutional recurrent neural network classifier using the synthetic data and its combination with a real annotated dataset.", isbn = "978-3-030-19823-7" }
