diff --git a/demo.py b/demo.py
index 6fa32ea..9396a74 100644
--- a/demo.py
+++ b/demo.py
@@ -6,7 +6,7 @@
 import numpy as np
 
 from data_generator import generate_trimap, random_choice, get_alpha_test
-from model import build_encoder_decoder_net
+from model import build_encoder_decoder
 from utils import get_final_output, safe_crop
 
 if __name__ == '__main__':
@@ -14,7 +14,7 @@
     channel = 4
 
     model_weights_path = 'models/model.62-0.0524.hdf5'
-    model = build_encoder_decoder_net()
+    model = build_encoder_decoder()
     model.load_weights(model_weights_path)
     print(model.summary())
 
diff --git a/migrate.py b/migrate.py
index 40faf4a..71eeb3c 100644
--- a/migrate.py
+++ b/migrate.py
@@ -2,7 +2,7 @@
 import numpy as np
 
 from config import channel
-from model import build_encoder_decoder_net
+from model import build_encoder_decoder
 from vgg16 import vgg16_model
 
 
@@ -43,7 +43,7 @@ def migrate_model(new_model):
 
 
 if __name__ == '__main__':
-    model = build_encoder_decoder_net()
+    model = build_encoder_decoder()
     migrate_model(model)
     print(model.summary())
     model.save_weights('models/model_weights.h5')
diff --git a/model.py b/model.py
index da47e06..a396263 100644
--- a/model.py
+++ b/model.py
@@ -6,7 +6,7 @@
 from custom_layers.unpooling_layer import Unpooling
 
 
-def build_encoder_decoder_net():
+def build_encoder_decoder():
     # Encoder
     input_tensor = Input(shape=(320, 320, 4))
     x = ZeroPadding2D((1, 1))(input_tensor)
@@ -96,7 +96,7 @@ def build_encoder_decoder_net():
     return model
 
 
-def build_refinement_net(encoder_decoder):
+def build_refinement(encoder_decoder):
     input_tensor = encoder_decoder.input
     x = encoder_decoder.output
     x = Concatenate(axis=1)([input_tensor, x])
@@ -111,12 +111,12 @@ def build_refinement_net(encoder_decoder):
 
 
 if __name__ == '__main__':
-    encoder_decoder = build_encoder_decoder_net(320, 320, 4)
+    encoder_decoder = build_encoder_decoder()
     # input_layer = model.get_layer('input')
     print(encoder_decoder.summary())
     plot_model(encoder_decoder, to_file='encoder_decoder.svg', show_layer_names=True, show_shapes=True)
 
-    refinement = build_refinement_net(encoder_decoder)
+    refinement = build_refinement(encoder_decoder)
     print(refinement.summary())
     plot_model(refinement, to_file='refinement.svg', show_layer_names=True, show_shapes=True)
 
diff --git a/plot_model.py b/plot_model.py
index 414c679..8f5fd52 100644
--- a/plot_model.py
+++ b/plot_model.py
@@ -1,9 +1,9 @@
 # dependency: pip install pydot & brew install graphviz
-from model import build_encoder_decoder_net
+from model import build_encoder_decoder
 from keras.utils import plot_model
 
 if __name__ == '__main__':
     img_rows, img_cols = 320, 320
     channel = 3
-    model = build_encoder_decoder_net(img_rows, img_cols, channel)
+    model = build_encoder_decoder(img_rows, img_cols, channel)
     plot_model(model, to_file='model.svg', show_layer_names=True, show_shapes=True)
diff --git a/test.py b/test.py
index 2bbc17c..558d949 100644
--- a/test.py
+++ b/test.py
@@ -3,7 +3,7 @@
 import cv2 as cv
 import numpy as np
 
-from model import build_encoder_decoder_net
+from model import build_encoder_decoder
 
 # python test.py -i "images/image.png" -t "images/trimap.png"
 if __name__ == '__main__':
@@ -11,7 +11,7 @@
     channel = 4
 
     model_weights_path = 'models/model.35-0.03.hdf5'
-    model = build_encoder_decoder_net()
+    model = build_encoder_decoder()
     model.load_weights(model_weights_path)
     print(model.summary())
 
diff --git a/train.py b/train.py
index a7836f2..b33395e 100644
--- a/train.py
+++ b/train.py
@@ -8,7 +8,7 @@
 import migrate
 from config import patience, batch_size, epochs, num_train_samples, num_valid_samples
 from data_generator import train_gen, valid_gen
-from model import build_encoder_decoder_net
+from model import build_encoder_decoder
 from utils import custom_loss_wrapper, get_available_cpus, get_available_gpus
 
 if __name__ == '__main__':
@@ -57,10 +57,10 @@ def on_epoch_end(self, epoch, logs=None):
     #     model_checkpoint = MyCbk(model)
     # else:
     if pretrained_path is not None:
-        new_model = build_encoder_decoder_net()
+        new_model = build_encoder_decoder()
         new_model.load_weights(pretrained_path)
     else:
-        new_model = build_encoder_decoder_net()
+        new_model = build_encoder_decoder()
         migrate.migrate_model(new_model)
 
     # sgd = SGD(lr=1e-3, decay=1e-6, momentum=0.9, nesterov=True)