TS-Net: OCR Trained to Switch Between Text Transcription Styles.
Jan KohútMichal HradisPublished in: CoRR (2021)
Keyphrases
- text recognition
- optical character recognition
- printed documents
- ocr systems
- handwriting recognition
- text extraction
- document processing
- page layout
- document images
- character recognition
- document analysis
- text mining
- information retrieval
- manually constructed
- post processing
- scanned documents
- error correction
- text retrieval
- text documents
- recognition errors
- textual data
- text analysis
- key concepts
- training process
- hybrid algorithm
- high speed
- text processing
- database
- multilayer perceptron
- semantic information
- historical documents
- training data
- search engine
- printed text