Debugging

bob/learn/tensorflow/analyzers/ExperimentAnalizer.py

@@ -52,10 +52,10 @@ class ExperimentAnalizer:
 
     def __call__(self, data_shuffler, network, session):
 
-        if self.data_shuffler is None:
-            self.data_shuffler = data_shuffler
-            self.network = network
-            self.session = session
+        #if self.data_shuffler is None:
+        #    self.data_shuffler = data_shuffler
+        #    self.network = network
+        #    self.session = session
 
         # Getting the base class. Recipe extracted from
         # http://stackoverflow.com/questions/5516263/creating-an-object-from-a-base-class-object-in-python/5516330#5516330
@@ -68,18 +68,18 @@ class ExperimentAnalizer:
 
         # Extracting features for enrollment
         enroll_data, enroll_labels = base_data_shuffler.get_batch()
-        enroll_features = self.network(enroll_data, session=self.session)
+        enroll_features = network(enroll_data, session=session)
         del enroll_data
 
         # Extracting features for probing
         probe_data, probe_labels = base_data_shuffler.get_batch()
-        probe_features = self.network(probe_data, session=self.session)
+        probe_features = network(probe_data, session=session)
         del probe_data
 
         # Creating models
         models = []
         for i in range(len(base_data_shuffler.possible_labels)):
-            indexes_model = numpy.where(enroll_labels == self.data_shuffler.possible_labels[i])[0]
+            indexes_model = numpy.where(enroll_labels == data_shuffler.possible_labels[i])[0]
             models.append(numpy.mean(enroll_features[indexes_model, :], axis=0))
 
         # Probing

bob/learn/tensorflow/network/Chopra.py

@@ -83,20 +83,20 @@ class Chopra(SequenceNetwork):
 
         self.add(Conv2D(name="conv1", kernel_size=conv1_kernel_size,
                         filters=conv1_output,
-                        activation=tf.nn.relu,
+                        activation=None,
                         weights_initialization=Xavier(seed=seed, use_gpu=self.use_gpu),
                         bias_initialization=Constant(use_gpu=self.use_gpu),
                         batch_norm=batch_norm
                         ))
-        self.add(MaxPooling(name="pooling1", shape=pooling1_size, activation=tf.nn.relu, batch_norm=False))
+        self.add(MaxPooling(name="pooling1", shape=pooling1_size, activation=tf.nn.tanh, batch_norm=False))
 
         self.add(Conv2D(name="conv2", kernel_size=conv2_kernel_size,
                         filters=conv2_output,
-                        activation=tf.nn.relu,
+                        activation=None,
                         weights_initialization=Xavier(seed=seed,  use_gpu=self.use_gpu),
                         bias_initialization=Constant(use_gpu=self.use_gpu),
                         batch_norm=batch_norm))
-        self.add(MaxPooling(name="pooling2", shape=pooling2_size, activation=tf.nn.relu, batch_norm=False))
+        self.add(MaxPooling(name="pooling2", shape=pooling2_size, activation=tf.nn.tanh, batch_norm=False))
 
         self.add(FullyConnected(name="fc1", output_dim=fc1_output,
                                 activation=None,

bob/learn/tensorflow/script/train_mnist.py

@@ -67,7 +67,7 @@ def main():
         trainer = Trainer(architecture=architecture,
                           loss=loss,
                           iterations=ITERATIONS,
-                          prefetch=False, temp_dir="./temp/cnn/no-batch-norm-all-relu")
+                          prefetch=False, temp_dir="./temp/cnn/no-batch-norm")
 
         #prefetch = False, temp_dir = "./temp/cnn/batch-norm-2convs-all-relu")
 

bob/learn/tensorflow/test/test_dnn.py

@@ -55,7 +55,6 @@ def test_dnn_trainer():
     trainer.train(train_data_shuffler)
     del trainer# Just to clean the variables
 
-    import ipdb; ipdb.set_trace();
     with tf.Session() as session:
         # Testing
         mlp = MLP(10, hidden_layers=[15, 20])

bob/learn/tensorflow/trainers/Trainer.py

@@ -252,6 +252,12 @@ class Trainer(object):
             session.run(self.enqueue_op, feed_dict=feed_dict)
 
     def create_graphs(self, train_data_shuffler, validation_data_shuffler):
+        """
+
+        :param train_data_shuffler:
+        :param validation_data_shuffler:
+        :return:
+        """
 
         # Creating train graph
         self.training_graph = self.compute_graph(train_data_shuffler, prefetch=self.prefetch, name="train")