Improving Book OCR by Adaptive Language and Image Models

   Abstract