Add an option to make auxilary logits optional in inception v3. (#1862)

slim/nets/inception_v3.py

@@ -425,6 +425,7 @@ def inception_v3(inputs,
                  prediction_fn=slim.softmax,
                  spatial_squeeze=True,
                  reuse=None,
+                 create_aux_logits=True,
                  scope='InceptionV3'):
   """Inception model from http://arxiv.org/abs/1512.00567.
 
@@ -457,6 +458,7 @@ def inception_v3(inputs,
         of shape [B, 1, 1, C], where B is batch_size and C is number of classes.
     reuse: whether or not the network and its variables should be reused. To be
       able to reuse 'scope' must be given.
+    create_aux_logits: Whether to create the auxiliary logits.
     scope: Optional variable_scope.
 
   Returns:
@@ -481,30 +483,31 @@ def inception_v3(inputs,
           depth_multiplier=depth_multiplier)
 
       # Auxiliary Head logits
-      with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],
-                          stride=1, padding='SAME'):
-        aux_logits = end_points['Mixed_6e']
-        with tf.variable_scope('AuxLogits'):
-          aux_logits = slim.avg_pool2d(
-              aux_logits, [5, 5], stride=3, padding='VALID',
-              scope='AvgPool_1a_5x5')
-          aux_logits = slim.conv2d(aux_logits, depth(128), [1, 1],
-                                   scope='Conv2d_1b_1x1')
-
-          # Shape of feature map before the final layer.
-          kernel_size = _reduced_kernel_size_for_small_input(
-              aux_logits, [5, 5])
-          aux_logits = slim.conv2d(
-              aux_logits, depth(768), kernel_size,
-              weights_initializer=trunc_normal(0.01),
-              padding='VALID', scope='Conv2d_2a_{}x{}'.format(*kernel_size))
-          aux_logits = slim.conv2d(
-              aux_logits, num_classes, [1, 1], activation_fn=None,
-              normalizer_fn=None, weights_initializer=trunc_normal(0.001),
-              scope='Conv2d_2b_1x1')
-          if spatial_squeeze:
-            aux_logits = tf.squeeze(aux_logits, [1, 2], name='SpatialSqueeze')
-          end_points['AuxLogits'] = aux_logits
+      if create_aux_logits:
+        with slim.arg_scope([slim.conv2d, slim.max_pool2d, slim.avg_pool2d],
+                            stride=1, padding='SAME'):
+          aux_logits = end_points['Mixed_6e']
+          with tf.variable_scope('AuxLogits'):
+            aux_logits = slim.avg_pool2d(
+                aux_logits, [5, 5], stride=3, padding='VALID',
+                scope='AvgPool_1a_5x5')
+            aux_logits = slim.conv2d(aux_logits, depth(128), [1, 1],
+                                     scope='Conv2d_1b_1x1')
+
+            # Shape of feature map before the final layer.
+            kernel_size = _reduced_kernel_size_for_small_input(
+                aux_logits, [5, 5])
+            aux_logits = slim.conv2d(
+                aux_logits, depth(768), kernel_size,
+                weights_initializer=trunc_normal(0.01),
+                padding='VALID', scope='Conv2d_2a_{}x{}'.format(*kernel_size))
+            aux_logits = slim.conv2d(
+                aux_logits, num_classes, [1, 1], activation_fn=None,
+                normalizer_fn=None, weights_initializer=trunc_normal(0.001),
+                scope='Conv2d_2b_1x1')
+            if spatial_squeeze:
+              aux_logits = tf.squeeze(aux_logits, [1, 2], name='SpatialSqueeze')
+            end_points['AuxLogits'] = aux_logits
 
       # Final pooling and prediction
       with tf.variable_scope('Logits'):