Asad-Raza

I am able to convert to onnx model using tf2onnx.convert. import tensorflow as tf import keras_ocr d=keras_ocr.detection.Detector(weights='clovaai_general') tf.saved_model.save(obj=d.model, export_dir='Data') !pip install -U tf2onnx !python -m tf2onnx.convert --saved-model 'Data' --output 'Data/craft_mlt_25k.onnx'...

The generated ONNX model is working good with onnxruntime

Hi , Did you get answer to the question? Im also interested in knowing the answer for this question. Does Keras OCR use pretrained model of pyTorch CRAFT implementation?

In my experimentation Im able to train some of the layers while freezing others in the following way : detector = keras_ocr.detection.Detector(weights= 'clovaai_general') convLayer = "conv_cls" for Layer in detector.model.layers:...

I had the similar issue. Downgrading the opencv-python to 4.4.0.44 worked me as well

> Maybe this would be helpful to you? #130 Thanks eduOS.

@tahakoudri No , I am trying [Keras implementation of CRAFT](https://github.com/faustomorales/keras-ocr) for training. They have provided [documentation ](https://keras-ocr.readthedocs.io/en/latest/examples/end_to_end_training.html)for training. You may find [this](https://github.com/faustomorales/keras-ocr/issues/40#issuecomment-631582529) comment also useful.

Im also facing this issue. Setting link_threshold to 1.5 or above does not solve it. Is there any solution for this issue?

> > Im also facing this issue. Setting link_threshold to 1.5 or above does not solve it. Is there any solution for this issue? > > can you show some...

> No , I have not tried rotating the image. I will try this. Thank you for the suggestion