Building an efficient OCR system for historical documents with little training data.
Jirí MartínekLadislav LencPavel KrálPublished in: Neural Comput. Appl. (2020)
Keyphrases
- historical documents
- training data
- document images
- handwriting recognition
- optical character recognition
- character recognition
- handwritten document images
- historical manuscripts
- word recognition
- learning algorithm
- document collections
- labeled data
- document image analysis
- printed documents
- visual features
- document processing
- handwritten text
- information extraction