The code has been mainly referred to the STREET model. See inference output for ocr example. The model trains both Thai and English in one charset. The input is an textline image, and the output is a ...
There is a lack of publicly available datasets at word/line level for Devanagari character recognition. We created this dataset containing Marathi vocabulary of ~12k word images and thier ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results