optimized imports

cnnModel/model.py

@@ -1,46 +1,42 @@
-from os.path import exists
-
-import numpy as np
-import tensorflow as tf
-import os
-import tflearn
-from tflearn.layers.conv import conv_2d, max_pool_2d
-from tflearn.layers.core import input_data, dropout, fully_connected
-from tflearn.layers.estimator import regression
-
-class cnnModel:
-    def __init__(self):
-        self.model = None
-        cnn = input_data(shape=[150, 150, 3])
-
-        #32 filters and stride is 5, the filter will move 5 pixels at a time
-        cnn = conv_2d(cnn, 32, filter_size=5, activation='relu')
-        cnn = max_pool_2d(cnn, 5)
-        cnn = conv_2d(cnn, 64, 5, activation='relu')
-        cnn = max_pool_2d(cnn, 5)
-        cnn = conv_2d(cnn, 128, 5, activation='relu')
-        cnn = max_pool_2d(cnn, 5)
-        cnn = conv_2d(cnn, 256, 5, activation='relu')
-
-        #input layer
-        cnn = fully_connected(cnn, 1024, activation='relu')
-        self.cnn = dropout(cnn, 0.5)
-        # input layer
-        cnn = fully_connected(cnn, 256, activation='relu')
-        self.cnn = dropout(cnn, 0.5)
-
-    def set_output_layer(self, labels):
-        # output layer
-        self.labels = labels
-        self.cnn = fully_connected(self.cnn, len(labels.keys()), activation='softmax')
-        self.cnn = regression(self.cnn, optimizer='adam', learning_rate=0.001, loss='categorical_crossentropy')
-
-        # final model initializer
-        self.model = tflearn.DNN(self.cnn, tensorboard_verbose=1)
-
-    def train(self, x_train, y_train, x_test, y_test):
-        self.model.fit(x_train, y_train, n_epoch=5, validation_set=(x_test, y_test), show_metric=True)
-        self.model.save('savedModels/cnn.tfl')
-
-    def predict(self, current_face):
+import numpy as np
+import tflearn
+from tflearn.layers.conv import conv_2d, max_pool_2d
+from tflearn.layers.core import input_data, dropout, fully_connected
+from tflearn.layers.estimator import regression
+
+class cnnModel:
+    def __init__(self):
+        self.model = None
+        cnn = input_data(shape=[150, 150, 3])
+
+        #32 filters and stride is 5, the filter will move 5 pixels at a time
+        cnn = conv_2d(cnn, 32, filter_size=5, activation='relu')
+        cnn = max_pool_2d(cnn, 5)
+        cnn = conv_2d(cnn, 64, 5, activation='relu')
+        cnn = max_pool_2d(cnn, 5)
+        cnn = conv_2d(cnn, 128, 5, activation='relu')
+        cnn = max_pool_2d(cnn, 5)
+        cnn = conv_2d(cnn, 256, 5, activation='relu')
+
+        #input layer
+        cnn = fully_connected(cnn, 1024, activation='relu')
+        self.cnn = dropout(cnn, 0.5)
+        # input layer
+        cnn = fully_connected(cnn, 256, activation='relu')
+        self.cnn = dropout(cnn, 0.5)
+
+    def set_output_layer(self, labels):
+        # output layer
+        self.labels = labels
+        self.cnn = fully_connected(self.cnn, len(labels.keys()), activation='softmax')
+        self.cnn = regression(self.cnn, optimizer='adam', learning_rate=0.001, loss='categorical_crossentropy')
+
+        # final model initializer
+        self.model = tflearn.DNN(self.cnn, tensorboard_verbose=1)
+
+    def train(self, x_train, y_train, x_test, y_test):
+        self.model.fit(x_train, y_train, n_epoch=5, validation_set=(x_test, y_test), show_metric=True)
+        self.model.save('savedModels/cnn.tfl')
+
+    def predict(self, current_face):
         return self.model.predict(np.array(current_face).reshape(-1, 150, 150, 3))

engine.py

@@ -1,53 +1,53 @@
-from cnnModel import model
-from getWebCamImages import take_pictures
-from lbphModel import lbph
-from trainModels import train
-from runDetector import detectFace
-
-lbph_model = lbph.lbphModel()
-cnn_model = model.cnnModel()
-
-
-def run():
-    cur_command = input("Enter Command (type 'help' to see list of commands): ")
-
-    while cur_command != 'exit':
-        if cur_command == 'take pictures':
-            take_pictures.takeImages()
-            print("Pictures added to database!")
-
-        elif cur_command == 'train lbph':
-            train.train_lbph(lbph_model)
-            print("LBPH model trained successfully!")
-
-        elif cur_command == 'train cnn':
-            train.train_cnn(cnn_model)
-            print("CNN model trained successfully!")
-
-        elif cur_command == 'run predictor lbph':
-            detectFace.runDetector(lbph_model.name_dict, lbph_model, 'lbph')
-            print()
-            print("Predictor closed successfully")
-
-        elif cur_command == 'run predictor cnn':
-            detectFace.runDetector(cnn_model.labels, cnn_model, 'cnn')
-            print()
-            print("Predictor closed successfully")
-
-        elif cur_command == 'help':
-            print()
-            print("Commands:")
-            print("take pictures: take pictures used for training models")
-            print('train lbph: train the local binary pattern histogram model')
-            print('train cnn: train convolutional neural network model (Will take a lot longer, if a model has already been trained previously, you '
-                  'will be given the option to load that model)')
-            print('run predictor: run the real-time prediction algorithm')
-            print()
-
-        else:
-            print("No such command!")
-
-        cur_command = input("Enter Command (type 'help' to see list of commands): ")
-
-
-run()
+from cnnModel import model
+from getWebCamImages import take_pictures
+from lbphModel import lbph
+from trainModels import train
+from runDetector import detectFace
+
+lbph_model = lbph.lbphModel()
+cnn_model = model.cnnModel()
+
+
+def run():
+    cur_command = input("Enter Command (type 'help' to see list of commands): ")
+
+    while cur_command != 'exit':
+        if cur_command == 'take pictures':
+            take_pictures.takeImages()
+            print("Pictures added to database!")
+
+        elif cur_command == 'train lbph':
+            train.train_lbph(lbph_model)
+            print("LBPH model trained successfully!")
+
+        elif cur_command == 'train cnn':
+            train.train_cnn(cnn_model)
+            print("CNN model trained successfully!")
+
+        elif cur_command == 'run predictor lbph':
+            detectFace.runDetector(lbph_model.name_dict, lbph_model, 'lbph')
+            print()
+            print("Predictor closed successfully")
+
+        elif cur_command == 'run predictor cnn':
+            detectFace.runDetector(cnn_model.labels, cnn_model, 'cnn')
+            print()
+            print("Predictor closed successfully")
+
+        elif cur_command == 'help':
+            print()
+            print("Commands:")
+            print("take pictures: take pictures used for training models")
+            print('train lbph: train the local binary pattern histogram model')
+            print('train cnn: train convolutional neural network model (Will take a lot longer, if a model has already been trained previously, you '
+                  'will be given the option to load that model)')
+            print('run predictor: run the real-time prediction algorithm')
+            print()
+
+        else:
+            print("No such command!")
+
+        cur_command = input("Enter Command (type 'help' to see list of commands): ")
+
+
+run()