vggnet19 for extracting features

core/vggnet.py

@@ -9,52 +9,52 @@
               'conv5_1', 'relu5_1', 'conv5_2', 'relu5_2', 'conv5_3', 'relu5_3', 'conv5_4', 'relu5_4']
 
 class Vgg19(object):
-  def __init__(self, vgg_path):
-    self.vgg_path = vgg_path
-
-  def build_inputs(self):
-    self.images = tf.placeholder(tf.float32, [None, 224, 224, 3], 'images')
-
-  def build_params(self):
-    model = scipy.io.loadmat(self.vgg_path)
-    layers = model['layers'][0]
-    self.params = {}
-    with tf.variable_scope('encoder'):
-      for i, layer in enumerate(layers):
-        layer_name = layer[0][0][0][0]
-        layer_type = layer[0][0][1][0]
-        if layer_type == 'conv':
-          w = layer[0][0][2][0][0].transpose(1, 0, 2, 3)
-          b = layer[0][0][2][0][1].reshape(-1)
-          self.params[layer_name] = {}
-          self.params[layer_name]['w'] = tf.get_variable(layer_name+'/w', initializer=tf.constant(w))
-          self.params[layer_name]['b'] = tf.get_variable(layer_name+'/b',initializer=tf.constant(b))
-
-  def _conv(self, x, w, b):
-    return tf.nn.bias_add(tf.nn.conv2d(x, w, strides=[1, 1, 1, 1], padding='SAME'), b)
-
-  def _relu(self, x):
-    return tf.nn.relu(x)
-
-  def _pool(self, x):
-    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')
-
-  def build_model(self):
-    for i, layer in enumerate(vgg_layers):
-      layer_type = layer[:4]
-      if layer_type == 'conv':
-        if layer == 'conv1_1':
-          h = self.images
-        h = self._conv(h, self.params[layer]['w'], self.params[layer]['b'])
-      elif layer_type == 'relu':
-        h = self._relu(h)
-      elif layer_type == 'pool':
-        h = self._pool(h)
-      if layer == 'conv5_3':
-        self.features = tf.reshape(h, [-1, 196, 512])
-
-
-  def build(self):
-    self.build_inputs()
-    self.build_params()
-    self.build_model()
+    def __init__(self, vgg_path):
+        self.vgg_path = vgg_path
+
+    def build_inputs(self):
+        self.images = tf.placeholder(tf.float32, [None, 224, 224, 3], 'images')
+
+    def build_params(self):
+        model = scipy.io.loadmat(self.vgg_path)
+        layers = model['layers'][0]
+        self.params = {}
+        with tf.variable_scope('encoder'):
+            for i, layer in enumerate(layers):
+                layer_name = layer[0][0][0][0]
+                layer_type = layer[0][0][1][0]
+                if layer_type == 'conv':
+                    w = layer[0][0][2][0][0].transpose(1, 0, 2, 3)
+                    b = layer[0][0][2][0][1].reshape(-1)
+                    self.params[layer_name] = {}
+                    self.params[layer_name]['w'] = tf.get_variable(layer_name+'/w', initializer=tf.constant(w))
+                    self.params[layer_name]['b'] = tf.get_variable(layer_name+'/b',initializer=tf.constant(b))
+
+    def _conv(self, x, w, b):
+        return tf.nn.bias_add(tf.nn.conv2d(x, w, strides=[1, 1, 1, 1], padding='SAME'), b)
+
+    def _relu(self, x):
+        return tf.nn.relu(x)
+
+    def _pool(self, x):
+        return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')
+
+    def build_model(self):
+        for i, layer in enumerate(vgg_layers):
+            layer_type = layer[:4]
+            if layer_type == 'conv':
+                if layer == 'conv1_1':
+                    h = self.images
+                h = self._conv(h, self.params[layer]['w'], self.params[layer]['b'])
+            elif layer_type == 'relu':
+                h = self._relu(h)
+            elif layer_type == 'pool':
+                h = self._pool(h)
+            if layer == 'conv5_3':
+                self.features = tf.reshape(h, [-1, 196, 512])
+            
+
+    def build(self):
+        self.build_inputs()
+        self.build_params()
+        self.build_model()