def train():
print('Loading and preprocessing train data...')
print('-' * 30)
imgs_train, imgs_mask_train = load_train_data()
print('Loaded train images: ', imgs_train.shape, imgs_mask_train.shape)
print('-' * 30)
# Normalization of the train set (Exp 4)
imgs_train = rescale_intensity(imgs_train, in_range=(0, 1))
imgs_train = imgs_train.astype(np.float32)
imgs_mask_train = imgs_mask_train.astype(np.float32)
print('Creating and compiling model...')
print('-' * 30)
model = unet()
#Saving the weights and the loss of the best predictions we obtained
model_checkpoint = ModelCheckpoint(
'/data/flavio/anatiel/models/dissertacao/unet_exp4_200epc_best_int16_clahe.h5',
monitor='val_loss',
save_best_only=True,
mode="min")
print('Fitting model...')
print('-' * 30)
history = model.fit(imgs_train,
imgs_mask_train,
batch_size=BATCH_SIZE,
epochs=EPOCHS,
verbose=1,
shuffle=True,
validation_split=0.1,
callbacks=[model_checkpoint])
model.save(
'/data/flavio/anatiel/models/dissertacao/unet_exp4_200epc_last_int16_clahe.h5'
)
# convert the history.history dict to a pandas DataFrame:
hist_df = pd.DataFrame(history.history)
# save to json:
hist_json_file = '/data/flavio/anatiel/models/dissertacao/unet_exp4_history_200epc_int16_clahe.json'
with open(hist_json_file, mode='w') as f:
hist_df.to_json(f)
print("history saved")
plt.plot(history.history['dice_coef'])
plt.plot(history.history['val_dice_coef'])
plt.title('Model dice coeff')
plt.ylabel('Dice coeff')
plt.xlabel('Epoch')
plt.legend(['Train', 'Val'], loc='upper left')
# save plot to file
plt.savefig(
'/data/flavio/anatiel/models/dissertacao/unet_exp4_plot_200epc_int16_clahe.png'
)
def train():
print('Loading and preprocessing train data...')
print('-' * 30)
imgs_train, imgs_mask_train = load_train_data()
print('Loaded train images: ', imgs_train.shape, imgs_train.dtype,
imgs_mask_train.shape, imgs_mask_train.dtype)
print('-' * 30)
# mean = np.mean(imgs_train) # mean for data centering
# std = np.std(imgs_train) # std for data normalization
# Normalization of the train set (Exp 1)
# imgs_train = imgs_train.astype(np.float32)
# imgs_train -= mean
# imgs_train /= std
# imgs_train = imgs_train.astype(np.float32)
# imgs_mask_train = imgs_mask_train.astype(np.float32)
print('Creating and compiling model...')
print('-' * 30)
model = unet()
#Saving the weights and the loss of the best predictions we obtained
model_checkpoint = ModelCheckpoint(
f'/data/flavio/anatiel/models/models_kfold/unet_ds{DS}_150epc_best_k{K}.h5',
monitor='val_loss',
save_best_only=True,
mode="min")
print('Fitting model...')
print('-' * 30)
history = model.fit(imgs_train,
imgs_mask_train,
batch_size=BATCH_SIZE,
epochs=EPOCHS,
verbose=1,
shuffle=True,
validation_split=0.1,
callbacks=[model_checkpoint])
model.save(
f'/data/flavio/anatiel/models/models_kfold/unet_ds{DS}_150epc_last_k{K}.h5'
)
# convert the history.history dict to a pandas DataFrame:
hist_df = pd.DataFrame(history.history)
# save to json:
hist_json_file = f'/data/flavio/anatiel/models/models_kfold/unet_ds{DS}_150epc_k{K}.json'
with open(hist_json_file, mode='w') as f:
hist_df.to_json(f)
print("history saved")
plt.plot(history.history['dice_coef'])
plt.plot(history.history['val_dice_coef'])
plt.title('Model dice coeff')
plt.ylabel('Dice coeff')
plt.xlabel('Epoch')
plt.legend(['Train', 'Val'], loc='upper left')
# save plot to file
plt.savefig(
f'/data/flavio/anatiel/models/models_kfold/unet_ds{DS}_150epc_k{K}.png'
)
def train():
print('Loading and preprocessing train data...')
print('-' * 30)
imgs_train, imgs_mask_train = load_train_data()
print('Loaded train images: ', imgs_train.shape, imgs_mask_train.shape)
print('-' * 30)
imgs_val, imgs_mask_val = load_val_data()
print('Loaded val images: ', imgs_val.shape, imgs_mask_val.shape)
print('-' * 30)
imgs_train = imgs_train.astype('float32')
mean = np.mean(imgs_train) # mean for data centering
std = np.std(imgs_train) # std for data normalization
#Normalization of the train set
imgs_train -= mean
imgs_train /= std
imgs_mask_train = imgs_mask_train.astype('float32')
print('Creating and compiling model...')
print('-' * 30)
model = unet()
#Saving the weights and the loss of the best predictions we obtained
model_checkpoint = ModelCheckpoint(
'/data/flavio/anatiel/models/unet2d/weights_unet_masked_lung_clahe_500epc.h5',
monitor='val_loss',
save_best_only=True)
# checkpoint = ModelCheckpoint('/data/flavio/anatiel/models/unet2d/weights_train_unet_masked_lung_blur_3epc.h5', monitor='dice', verbose=1, save_best_only=True,save_weights_only=True, mode='max')
# checkpoint2 = ModelCheckpoint('/data/flavio/anatiel/models/unet2d/weights_val_unet_masked_lung_blur_3epc.h5', monitor='val_dice', verbose=1, save_best_only=True,save_weights_only=True, mode='max')
print('Fitting model...')
print('-' * 30)
history = model.fit(imgs_train,
imgs_mask_train,
batch_size=BATCH_SIZE,
epochs=EPOCHS,
verbose=1,
shuffle=True,
validation_split=0.2,
callbacks=[model_checkpoint])
# history = model.fit(imgs_train, imgs_mask_train, batch_size=BATCH_SIZE, epochs=EPOCHS, verbose=1, shuffle=True, callbacks=[checkpoint,checkpoint2], validation_data=(imgs_val, imgs_mask_val))
model.save(
'/data/flavio/anatiel/models/unet2d/weights_unet_masked_lung_clahe_500epc_last.h5'
)
# convert the history.history dict to a pandas DataFrame:
hist_df = pd.DataFrame(history.history)
# save to json:
print("saving history...")
hist_json_file = 'anatiel/unet/history_masked_lung_clahe_500epc.json'
with open(hist_json_file, mode='w') as f:
hist_df.to_json(f)
print("history saved")
plt.plot(history.history['dice_coef'])
plt.plot(history.history['val_dice_coef'])
plt.title('Model dice coeff')
plt.ylabel('Dice coeff')
plt.xlabel('Epoch')
plt.legend(['Train', 'Test'], loc='upper left')
# save plot to file
plt.savefig('anatiel/unet/plot_dice_masked_lung_clahe_500epc.png')
plt.show()