def train_(base_path):
data, anno_viable, anno_dead = read_data(base_path)
print("loaded!!")
anno_viable = np.expand_dims(anno_viable, axis=-1)
anno_dead = np.expand_dims(anno_dead, axis=-1)
print("expanded!!")
mean = np.mean(data)
print("mean finished!!")
std = np.std(data)
print("std finished!!")
data_ = (data - mean) / std
train_data = data_[:50]
train_anno_viable = anno_viable[:50]
train_anno_dead = anno_dead[:50]
val_data = data_[50:]
val_anno_viable = anno_viable[50:]
val_anno_dead = anno_dead[50:]
print('-' * 30)
print(
'Creating and compiling the fully convolutional regression networks.')
print('-' * 30)
if sys.argv[1] == 'unet':
model = buildModel_U_net(input_dim=(504, 376, 3))
filename = 'cell_counting_viable_unet.hdf5'
elif sys.argv[1] == 'fcrna':
model = buildModel_FCRN_A_v2(input_dim=(504, 376, 3))
filename = 'cell_counting_viable_fcrna.hdf5'
else:
raise ValueError(
'The first command line argument should be "unet" or "fcrna"')
learn(filename, train_data, train_anno_viable, val_data, val_anno_viable,
model)
if sys.argv[1] == 'unet':
model = buildModel_U_net(input_dim=(504, 376, 3))
filename = 'cell_counting_dead_unet.hdf5'
elif sys.argv[1] == 'fcrna':
model = buildModel_FCRN_A_v2(input_dim=(504, 376, 3))
filename = 'cell_counting_dead_fcrna.hdf5'
else:
raise ValueError(
'The first command line argument should be "unet" or "fcrna"')
learn(filename, train_data, train_anno_dead, val_data, val_anno_dead,
model)
def train_(base_path):
data, anno = read_data(base_path)
anno = np.expand_dims(anno, axis = -1)
mean = np.mean(data)
std = np.std(data)
data_ = (data - mean) / std
train_data = data_[:150]
train_anno = anno[:150]
val_data = data_[150:]
val_anno = anno[150:]
print('-'*30)
print('Creating and compiling the fully convolutional regression networks.')
print('-'*30)
model = buildModel_U_net(input_dim = (256,256,3))
model_checkpoint = ModelCheckpoint('cell_counting.hdf5', monitor='loss', save_best_only=True)
model.summary()
print('...Fitting model...')
print('-'*30)
change_lr = LearningRateScheduler(step_decay)
datagen = ImageDataGenerator(
featurewise_center = False, # set input mean to 0 over the dataset
samplewise_center = False, # set each sample mean to 0
featurewise_std_normalization = False, # divide inputs by std of the dataset
samplewise_std_normalization = False, # divide each input by its std
zca_whitening = False, # apply ZCA whitening
rotation_range = 30, # randomly rotate images in the range (degrees, 0 to 180)
width_shift_range = 0.3, # randomly shift images horizontally (fraction of total width)
height_shift_range = 0.3, # randomly shift images vertically (fraction of total height)
zoom_range = 0.3,
shear_range = 0.,
horizontal_flip = True, # randomly flip images
vertical_flip = True, # randomly flip images
fill_mode = 'constant',
dim_ordering = 'tf')
# Fit the model on the batches generated by datagen.flow().
model.fit_generator(datagen.flow(train_data,
train_anno,
batch_size = 16
),
samples_per_epoch = train_data.shape[0],
nb_epoch = 192,
callbacks = [model_checkpoint, change_lr],
)
model.load_weights('cell_counting.hdf5')
A = model.predict(val_data)
mean_diff = np.average(np.abs(np.sum(np.sum(A,1),1)-np.sum(np.sum(val_anno,1),1))) / (100.0)
print('After training, the difference is : {} cells per image.'.format(np.abs(mean_diff)))
# train_data = data_[:12]
train_data = (X_train - np.mean(X_train)) / np.std(X_train)
train_anno = anno[:len(train_set)]
# val_data = data_[12:]
val_data = (X_test - np.mean(X_test)) / np.std(X_test)
val_anno = anno[len(train_set):]
return train_data, train_anno, val_data, val_anno
train_data, train_anno, val_data, val_anno = load_data()
#%% Creat the model
model = buildModel_U_net(input_dim=train_data.shape[1:])
model.load_weights('trained_model.hdf5')
#%% Detection
def detect(data=val_data, threshold=0.6):
start = time.time()
A = model.predict(val_data)
print("\nConsumed time: \t%.2f\t s\n" % (time.time() - start))
#mean_diff = np.average(np.abs(np.sum(np.sum(A,1),1)-np.sum(np.sum(val_anno,1),1))) / (100.0)
#print('After training, the difference is : {} cells per image.'.format(np.abs(mean_diff)))
preds_test = np.where(A > 0, A / 100, A)
preds_test = (preds_test + 1) / 2
from tqdm import tqdm
#%%
# Set some parameters
IMG_WIDTH = 1024
IMG_HEIGHT = 512
IMG_CHANNEL = 3
seed = 42
random.seed = seed
np.random.seed = seed
#%% Creat the model
print('#' * 30)
print('Creating and compiling the fully convolutional regression networks.')
print('#' * 30)
model = buildModel_U_net(input_dim=(IMG_HEIGHT, IMG_WIDTH, IMG_CHANNEL))
model.load_weights('trained_model.hdf5')
print('#' * 30)
print('Model loaded.')
print('#' * 30)
#%% Read image
def read_image(img_path):
img_name = os.path.basename(img_path)[:-4]
img = cv2.imread(img_path)
img = cv2.resize(img, (IMG_WIDTH, IMG_HEIGHT),
interpolation=cv2.INTER_AREA)
img = np.asarray(img)