conscienceli / iternet Goto Github PK

when I run the predict.py,I found the N_patches_img(22801) in crop_predict.py is too big to run.

Hello,

Congratulations for your work, it is inspiring and very impressive.
I was wondering if you could provide a link pointing on where you found the DRIVE test set predictions, because I have looked online and cannot find the ground truths for the test set to report my own results.
Many thanks in advance.

Can you provide the image names as a list that were used for training and testing for ChaseDB and STARE ?

Thanks!

hi, i m getting above mentioned error when running train.py file even when my hdf5 has been generated for training data. please help. Thanks

I didn't found in your code,would you please relase the code ?

I didnt change the train code,use the model "checkpoint = ModelCheckpoint(save_path, monitor='final_out_loss', verbose=1, save_best_only=True, mode='min')",iteration=3, DATASET='STARE',batch_size=32, epochs=200.
predict(batch_size=32, epochs=200, iteration=3, stride_size=3, DATASET='STARE'),and
and result is lower than your paper.
Area under the ROC curve: 0.962542154639597
Area under Precision-Recall curve: 0.8794006909319013
Jaccard similarity score: 0.9677991055713587
F1 score (F-measure): 0.7959445466207131

Confusion matrix:[[1113200 12113]
[ 27496 77250]]
ACCURACY: 0.9677991055713587
SENSITIVITY: 0.7374983292918107
SPECIFICITY: 0.9892358837052446
PRECISION: 0.8644517305820082
Can you help me??where i did in wrong way?

I have some issues to run your code: 'Model' object has no attribute '_get_distribution_strategy'
Which version of Keras and TF did you use?

Which version of packages (tensorflow, keras etc.) have you used? Can you produce requirements file of this project by the following script?
pip freeze > requirements.txt

With this definition, we drew a curve of θ versus C(θ) (refer to the supplementary material for some examples).
We adopt the area under this curve as connectivity metric (abbreviated to Conn.).

So could you tell me what is the code?

What are the areas of the code I have to change to train it on my custom dataset?

Download weights from the link and uploaded to my drive. Can't load the weights on local/Google Colab. It is stuck running on the same cell for some time.

Hello,
I'm trying to use only tensorflow in the code, so I'm changing the Keras imports to tensorflow.keras, like this:

from tensorflow.keras import losses
from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, ReduceLROnPlateau
from tensorflow.keras.layers import Input, MaxPooling2D
from tensorflow.keras.layers import concatenate, Conv2D, Conv2DTranspose, Dropout, ReLU, BatchNormalization, Activation
from tensorflow.keras.layers import add, multiply
from tensorflow.keras.models import Model
from tensorflow.keras.optimizers import Adam

However, the graph gets disconnected at dropout_19_1 in the third iteration:
I get the following error:

Graph disconnected: cannot obtain value for tensor Tensor("dropout_19_1/Identity:0", shape=(None, None, None, 32), dtype=float32) at layer "concatenate_4".

The error is coming from this line:
model = Model(inputs=[inputs], outputs=[outs[-1]])

Any thoughts?

Thank you