Hello all,
I’m having problem figuring out the parameters of ctc loss in keras with tensorflow backend. Can anyone please help me with it?. I’m using 1D convolutions for the task. I get error related to dimension. I tried going through the OCR example which is on Keras website but did not get much about the format. Can anyone provide some reference on what actual needs to be passed to keras ctc loss?
Thanks in advance.
Regards,
Atul