- Navigate to license_plate_generator folder
- Open cmd
- type "python Generator_augmentation -n [number of image to generate] -s save -i [directory to save image]
- Navigate to root folder
- Open cmd
- python training.py ** automatically save weight data in every each epoch
I use Evaluate_Predict_Showcase.ipynb to evluate the model.
OS : Windows 10
Python : 3.7
Tensorflow : 2.3.0
Keras : 2.4.3
File | Description |
---|---|
Model.py | Contain CRNN model |
parameter.py | Contain parameters |
training.py | Model training file |
Image_Generator.py | Image Preprocessing and formating before .fit() |
save_model.py | Saving trained model as .h5 |
LSTM+BN5--thai-v3.hdf5 | Contain Trained CRNN weight |
Model_LSTM+BN5--thai-v3.h5 | Contain Trained CRNN model |
This model still need be fine-tuned with real license plate images.