What type of dataset is being used whether the boxes are for each single characters or it is for a word and what is the annotation type for the same ?

faustomorales / keras-ocr

A packaged and flexible version of the CRAFT text detector and Keras CRNN recognition model.

https://keras-ocr.readthedocs.io/

MIT License

1.37k stars 349 forks source link

Open kctiintern21 opened 2 years ago

kctiintern21 commented 2 years ago

As I need to annotate the dataset from scratch I am not understanding the format what ICDAR and other dataset have done for the annotation part?