saadwazir / histoseg Goto Github PK

I want to ask a question.

When you were calculating F1, did you use the algorithm provided by the GlaS competition official website to calculate each object?
That is, is your calculation the same as instance segmentation?

Or,
is your calculation in f1 score is calculated by pixel?

Thank you.

Sorry I don't know why my training .npy file didn't work, could you please tell me why? Or provide .npy file for train?

The following are the errors that result from training with .npy file：

Traceback (most recent call last):
File "e:\HistoSeg\HistoSeg-Tensorflow\HistoSeg_Train.py", line 629, in
results=model.fit(X_train, y_train, batch_size=batch_arg, epochs=epochs_arg, callbacks=callbacks, validation_data=(X_test, y_test) , verbose = 1)
File "E:\Users\15199\anaconda3\envs\Histoseg\lib\site-packages\tensorflow\python\keras\engine\training_v1.py", line 793, in fit
return func.fit(
File "E:\Users\15199\anaconda3\envs\Histoseg\lib\site-packages\tensorflow\python\keras\engine\training_arrays_v1.py", line 644, in fit
return fit_loop(
File "E:\Users\15199\anaconda3\envs\Histoseg\lib\site-packages\tensorflow\python\keras\engine\training_arrays_v1.py", line 380, in model_iteration
batch_outs = f(ins_batch)
File "E:\Users\15199\anaconda3\envs\Histoseg\lib\site-packages\tensorflow\python\keras\backend.py", line 4067, in call
fetched = self._callable_fn(*array_vals,
File "E:\Users\15199\anaconda3\envs\Histoseg\lib\site-packages\tensorflow\python\client\session.py", line 1483, in call
ret = tf_session.TF_SessionRunCallable(self._session._session,
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
(0) INVALID_ARGUMENT: assertion failed: [labels out of bound] [Condition x < y did not hold element-wise:] [x (metrics/mean_io_u/confusion_matrix/control_dependency:0) = ] [0 0 0...] [y (metrics/mean_io_u/confusion_matrix/Cast_2:0) = ] [2]
[[{{function_node metrics_mean_io_u_confusion_matrix_assert_less_Assert_AssertGuard_false_7653}}{{node Assert}}]]
[[expanded_conv_11_project_BN/cond/then/_980/FusedBatchNormV3/_6151]]
(1) INVALID_ARGUMENT: assertion failed: [labels out of bound] [Condition x < y did not hold element-wise:] [x (metrics/mean_io_u/confusion_matrix/control_dependency:0) = ] [0 0 0...] [y (metrics/mean_io_u/confusion_matrix/Cast_2:0) = ] [2]
[[{{function_node metrics_mean_io_u_confusion_matrix_assert_less_Assert_AssertGuard_false_7653}}{{node Assert}}]]

Thanks.

Hi,

Could you please provide the .npy file for test?

Or
Could you please provide the sample for how to build the .npy file?

Thank you!

Hi, I have installed all the requirements that you listed and have trained the model using your original .npy files.

When I attempt to test the model, it informs me:
"You are trying to load a weight file containing 112 layers into a model with 109 layers."

Without being able to test the model, I cannot confirm that your experimental results are real.

In the GlaS dataset, your IoU is only 76.73. According to the mathematical definition of IoU and Dice coefficient, your Dice is absolutely impossible to reach 99.09, so your Dice coefficient calculation is obviously wrong. You should check your code implementation of Dice coefficient calculation.

在GlaS数据集里，你的IoU只有76.73，根据IoU和Dice系数的数学定义，你的Dice是绝对不可能达到99.09的，所以你的Dice系数计算显然是错误的。你应该检查一下你关于Dice系数计算这部分的代码实现。