Log images in autoencoder

Author: vanpelt

keras-cnn/cnn.py

@@ -33,7 +33,6 @@
 model.add(MaxPooling2D(pool_size=(2, 2)))
 model.add(Flatten())
 model.add(Dense(config.dense_layer_size, activation='relu'))
-model.add(Dropout(0.4))
 model.add(Dense(num_classes, activation='softmax'))
 
 model.compile(loss='categorical_crossentropy', optimizer='adam',

keras-cnn/conv-demo.py

@@ -3,9 +3,9 @@
 from scipy.signal import convolve2d
 
 
-kernel = [[0.0,0.0,0.0],
-          [0.0,1.0,0.0],
-          [0.0,0.0,0.0]]
+kernel = [[0.0,-1.0,0.0],
+          [-1.0,4.0,-1.0],
+          [0.0,-1.0,0.0]]
 
 backgroundColor = (0,)*3
 pixelSize = 10

keras-fashion/perceptron-linear.py

@@ -32,7 +32,7 @@
 
 # Fit the model
 model.fit(X_train, y_train, epochs=config.epochs, validation_data=(X_test, y_test),
-                    callbacks=[WandbKerasCallback()])
+                    callbacks=[WandbKerasCallback(validation_data=X_test, labels=y_test)])
 
 
 # Output some predictions
@@ -53,4 +53,4 @@
 
     draw = ImageDraw.Draw(img)
     draw.text((10, 10),labels[prediction],(255))
-    img.save(str(i)+".jpg")
+    img.save(str(i)+".jpg")

keras-mlp/dropout.py

@@ -33,12 +33,6 @@
 model.add(Dropout(config.dropout))
 model.add(Dense(config.hidden_nodes, activation='relu'))
 model.add(Dropout(config.dropout))
-model.add(Dense(config.hidden_nodes, activation='relu'))
-model.add(Dropout(config.dropout))
-model.add(Dense(config.hidden_nodes, activation='relu'))
-model.add(Dropout(config.dropout))
-model.add(Dense(config.hidden_nodes, activation='relu'))
-model.add(Dropout(config.dropout))
 model.add(Dense(num_classes, activation='softmax'))
 model.compile(loss='categorical_crossentropy', optimizer=config.optimizer,
                     metrics=['accuracy'])

keras-perceptron/digits.py

@@ -2,7 +2,7 @@
 import numpy as np
 (X_train, y_train), (X_test, y_test) = mnist.load_data()
 
-idx = 150 # index of the digit I want to look at
+idx = 200 # index of the digit I want to look at
 digit = X_train[idx]
 
 # make an ascii-art drawing of the digit I'm looking at
@@ -14,7 +14,7 @@
         elif digit[i][j] < 128:
             str += "."
         else:
-            str += "X"
+            str += "*"
     str += "\n"
 
 print(str)

keras-transfer/inception-inspect.py

@@ -7,7 +7,7 @@
 
 model = InceptionV3(weights='imagenet')
 
-img_path = 'toaster.jpg'
+img_path = 'elephant.jpg'
 img = image.load_img(img_path, target_size=(299, 299))
 x = image.img_to_array(img)
 x = np.expand_dims(x, axis=0)

keras-transfer/vgg-inspect.py

@@ -61,12 +61,12 @@ def VGG_16(weights_path=None):
     return model
 
 if __name__ == "__main__":
-    # im = cv2.resize(cv2.imread('elephant.jpg'), (224, 224)).astype(np.float32)
-    # im[:,:,0] -= 103.939
-    # im[:,:,1] -= 116.779
-    # im[:,:,2] -= 123.68
-    # im = im.transpose((2,0,1))
-    # im = np.expand_dims(im, axis=0)
+    #im = cv2.resize(cv2.imread('elephant.jpg'), (224, 224)).astype(np.float32)
+    #im[:,:,0] -= 103.939
+    #im[:,:,1] -= 116.779
+    #im[:,:,2] -= 123.68
+    #im = im.transpose((2,0,1))
+    #im = np.expand_dims(im, axis=0)
     model = VGG_16()
     model.summary()
     # Test pretrained model