Merge pull request #18 from philipperemy/vgg16

update

Former-commit-id: cf4a055

README.md

@@ -8,7 +8,7 @@ pip install keract
 *You have just found a (easy) way to get the activations (outputs) and gradients for each layer of your Keras model (LSTM, conv nets...).*
 
 <p align="center">
-  <img src="assets/1.png">
+  <img src="assets/intro.png">
 </p>
 
 ## API
@@ -101,28 +101,17 @@ dense_2/Softmax:0
 (128, 10)
 ```
 
-We can even visualise some of them.
+We can also visualise the activations. Here's another example using VGG16:
 
 <p align="center">
-  <img src="assets/0.png" width="50">
-  <br><i>A random seven from MNIST</i>
+  <img src="assets/cat.jpg">
+  <br><i>A cat.</i>
 </p>
 
 
 <p align="center">
-  <img src="assets/1.png">
-  <br><i>Activation map of CONV1 of LeNet</i>
-</p>
-
-<p align="center">
-  <img src="assets/2.png" width="200">
-  <br><i>Activation map of FC1 of LeNet</i>
-</p>
-
-
-<p align="center">
-  <img src="assets/3.png" width="300">
-  <br><i>Activation map of Softmax of LeNet. <b>Yes it's a seven!</b></i>
+  <img src="assets/cat_activations.png" width="600">
+  <br><i>Outputs of the first conv of VGG16.</i>
 </p>
 
 ### Repo views (since 2018/10/31)

examples/vgg16.py

@@ -0,0 +1,51 @@
+from keras.applications.vgg16 import VGG16
+from keras.applications.vgg16 import decode_predictions
+from keras.applications.vgg16 import preprocess_input
+from keras.preprocessing.image import img_to_array
+from keras.preprocessing.image import load_img
+
+model = VGG16()
+
+from PIL import Image
+import requests
+from io import BytesIO
+
+url = 'https://upload.wikimedia.org/wikipedia/commons/thumb/1/14/Gatto_europeo4.jpg/250px-Gatto_europeo4.jpg'
+response = requests.get(url)
+image = Image.open(BytesIO(response.content))
+image = image.crop((0, 0, 244, 244))
+image.save('cat.jpg')
+
+image = load_img('cat.jpg', target_size=(224, 224))
+image = img_to_array(image)
+image = image.reshape((1, image.shape[0], image.shape[1], image.shape[2]))
+image = preprocess_input(image)
+yhat = model.predict(image)
+label = decode_predictions(yhat)
+label = label[0][0]
+print('{} ({})'.format(label[1], label[2] * 100))
+
+import keract
+
+model.compile(optimizer='adam',
+              loss='categorical_crossentropy',
+              metrics=['accuracy'])
+activations = keract.get_activations(model, image)
+first = activations.get('block1_conv1/Relu:0')
+keract.display_activations(activations)
+#
+# import matplotlib.pyplot as plt
+#
+# fig = plt.figure(figsize=(12, 12))
+#
+# rows = 8
+# columns = 8
+#
+# first = activations.get('block1_conv1/Relu:0')
+#
+# for i in range(1, columns * rows + 1):
+#     img = first[0, :, :, i - 1]
+#     fig.add_subplot(rows, columns, i)
+#     plt.imshow(img)
+#     plt.axis('off')
+# plt.show()

keract/keract.py

@@ -44,36 +44,24 @@ def get_activations(model, x, layer_name=None):
 
 
 def display_activations(activations):
-    import numpy as np
     import matplotlib.pyplot as plt
-    """
-    (1, 26, 26, 32)
-    (1, 24, 24, 64)
-    (1, 12, 12, 64)
-    (1, 12, 12, 64)
-    (1, 9216)
-    (1, 128)
-    (1, 128)
-    (1, 10)
-    """
-    for name, activation_map in activations.items():
-        assert activation_map.shape[0] == 1, 'One image at a time to visualize.'
-        print('Displaying activation map [{}]'.format(name))
-        shape = activation_map.shape
-        if len(shape) == 4:
-            activations = np.hstack(np.transpose(activation_map[0], (2, 0, 1)))
-        elif len(shape) == 2:
-            # try to make it square as much as possible. we can skip some activations.
-            activations = activation_map[0]
-            num_activations = len(activations)
-            if num_activations > 1024:  # too hard to display it on the screen.
-                square_param = int(np.floor(np.sqrt(num_activations)))
-                activations = activations[0: square_param * square_param]
-                activations = np.reshape(activations, (square_param, square_param))
-            else:
-                activations = np.expand_dims(activations, axis=0)
-        else:
-            raise Exception('len(shape) = 3 has not been implemented.')
-        plt.title(name)
-        plt.imshow(activations, interpolation='None', cmap='jet')
+    max_rows = 8
+    max_columns = 8
+    for layer_name, first in activations.items():
+        print(layer_name, first.shape, end=' ')
+        if first.shape[0] != 1:
+            print('-> Skipped. First dimension is not 1.')
+            continue
+        if len(first.shape) <= 2:
+            print('-> Skipped. 2D Activations.')
+            continue
+        print('')
+        fig = plt.figure(figsize=(12, 12))
+        plt.axis('off')
+        plt.title(layer_name)
+        for i in range(1, min(max_columns * max_rows + 1, first.shape[-1] + 1)):
+            img = first[0, :, :, i - 1]
+            fig.add_subplot(max_rows, max_columns, i)
+            plt.imshow(img)
+            plt.axis('off')
         plt.show()