Improving state-of-the-art OCR through high-precision document-specific modeling.
Andrew KaeGary B. HuangCarl DoerschErik G. Learned-MillerPublished in: CVPR (2010)
Keyphrases
- high precision
- high recall
- document images
- document processing
- high reliability
- printed documents
- post processing
- document analysis
- scanned documents
- keywords
- high accuracy
- text documents
- achieve high precision
- optical character recognition
- character recognition
- document clustering
- semantic information
- end to end
- document classification
- database
- retrieval systems
- web documents
- higher level
- information retrieval systems
- preprocessing
- information retrieval