diff --git a/onnx_tf/backend_rep.py b/onnx_tf/backend_rep.py
index bee1ec49e..71ea39064 100644
--- a/onnx_tf/backend_rep.py
+++ b/onnx_tf/backend_rep.py
@@ -49,40 +49,43 @@ def tensor_dict(self):
   def tensor_dict(self, tensor_dict):
     self._tensor_dict = tensor_dict
 
-  def run(self, inputs, **kwargs):
+  def run(self, inputs, sess=None, **kwargs):
     """ Run TensorflowRep.
 
     :param inputs: Given inputs.
+    :param sess: tf.Session. The environment in which Operation objects are executed,
+      and Tensor objects are evaluated.
     :param kwargs: Other args.
     :return: Outputs.
     """
     super(TensorflowRep, self).run(inputs, **kwargs)
 
+    should_close_sess = sess is None
     # TODO: handle name scope if necessary
     with self.graph.as_default():
-      with tf.Session() as sess:
-        if isinstance(inputs, dict):
-          feed_dict = inputs
-        elif isinstance(inputs, list) or isinstance(inputs, tuple):
-          if len(self.inputs) != len(inputs):
-            raise RuntimeError('Expected {} values for uninitialized '
-                               'graph inputs ({}), but got {}.'.format(
-                                   len(self.inputs), ', '.join(self.inputs),
-                                   len(inputs)))
-          feed_dict = dict(zip(self.inputs, inputs))
-        else:
-          # single input
-          feed_dict = dict([(self.inputs[0], inputs)])
-
-        feed_dict = {
-            self.tensor_dict[key]: feed_dict[key] for key in self.inputs
-        }
-
-        sess.run(tf.global_variables_initializer())
-        outputs = [self.tensor_dict[output] for output in self.outputs]
-
-        output_values = sess.run(outputs, feed_dict=feed_dict)
-        return namedtupledict('Outputs', self.outputs)(*output_values)
+      sess = sess or tf.Session()
+      if isinstance(inputs, dict):
+        feed_dict = inputs
+      elif isinstance(inputs, list) or isinstance(inputs, tuple):
+        if len(self.inputs) != len(inputs):
+          raise RuntimeError('Expected {} values for uninitialized '
+                             'graph inputs ({}), but got {}.'.format(
+                                 len(self.inputs), ', '.join(self.inputs),
+                                 len(inputs)))
+        feed_dict = dict(zip(self.inputs, inputs))
+      else:
+        # single input
+        feed_dict = dict([(self.inputs[0], inputs)])
+
+      feed_dict = {self.tensor_dict[key]: feed_dict[key] for key in self.inputs}
+
+      sess.run(tf.global_variables_initializer())
+      outputs = [self.tensor_dict[output] for output in self.outputs]
+
+      output_values = sess.run(outputs, feed_dict=feed_dict)
+      if should_close_sess:
+        sess.close()
+      return namedtupledict('Outputs', self.outputs)(*output_values)
 
   def export_graph(self, path):
     """Export backend representation to a Tensorflow proto file.
@@ -99,3 +102,12 @@ def export_graph(self, path):
     file = open(path, "wb")
     file.write(graph_proto.SerializeToString())
     file.close()
+
+  def create_session(self):
+    """ Create tf.Session object by using current graph.
+    Pass it to `run` function could reduce the overhead of initialization
+    when doing inference consecutively.
+
+    :returns: A Session object.
+    """
+    return tf.Session(graph=self.graph)