includeamin / wbc-classification

Classification of WBC ( White Blood Cells ) with CNN . (Convolutional Neural Network)

WBC-Classification

Classification of WBC ( White Blood Cells ) with CNN . (Convectional Neural Network)

DATASET

• Dataset

Requirements

Keras==2.1.5
numpy==1.18.1
matplotlib==3.1.1
imutils==0.5.3
scikit_learn==0.23.1

Train and save trained model

Use this command to train the model and save model

git clone 
cd WBC-Classification
python3 learning.py -d ./CNN/datasets/TRAIN/ -m ./TrainedModel/model_epoch_100.hdf5

after train you will see result plot:

Test the model

python3 load_test_model.py -d ./CNN/datasets/TEST -m ./TrainedModel/model_epoch_100.hdf5

NOTE

default epoch count is 100. for change it, just edit cp variable in learning.py

Todo

• more customizable

wbc-classification's Issues

Accuracy values are not the same as described by the file

Using your model as a reference, I trained the said dataset based on a given model but my accuracy values are much lower. Can you please look into it?

where is datasets and model file?

datasets is missing otherwise put a link for datasets.....model file also not include please update you Readme file for instruction help to others

vjvgjv

bhkhkb

value error low >= high

File "test_model.py", line 75, in
idxs = np.random.randint(0, len(imagePaths), size=(10,))
File "mtrand.pyx", line 630, in numpy.random.mtrand.RandomState.randint
File "bounded_integers.pyx", line 1228, in numpy.random.bounded_integers._rand_int64
ValueError: low >= high

value error for test_model.py file

File "test_model.py", line 75, in
idxs = np.random.randint(0, len(imagePaths), size=(10,))
File "mtrand.pyx", line 630, in numpy.random.mtrand.RandomState.randint
File "bounded_integers.pyx", line 1228, in numpy.random.bounded_integers._rand_int64
ValueError: low >= high

Unsupported image shape: ()
preprocessor not enough values to unpack (expected 3, got 2)
Unsupported image shape: ()