There are two types of data set at the link below. which one i should use.

[arsalan993]

There are two types of data set at the link below. which one i should use.

[MaybeShewill-CV]

@arsalan993 The model has been trained on Synth dataset. You are welcomed to try any dataset you want.

[arsalan993]

"Synthetic Word Dataset" is basically the Synth data set am i right..
Btw i really like your implementation but issue is it is not trained to recognize number .. so tell me should i train again with a new data set for adding number recognizer feature into it.

[MaybeShewill-CV]

@arsalan993 In my opinion you'd better train a new model on your own dataset, if you just need to recognize numbers you can change the word dictionary into a new one that only contains numbers.

[arsalan993]

yes i am planning to merge COCO text recognition dataset with and synth90 dataset and will train my model on it..
In which files i should the word dictionary.
Btw i want to recognize both alphabets and numbers..
please elaborate where exactly do i need to make changes