Fixed Facenet

bob/learn/tensorflow/network/FaceNet.py

@@ -184,13 +184,13 @@ class FaceNet(SequenceNetwork):
         self.add(MaxPooling(name="pooling6", shape=pool6_shape))
 
         self.add(FullyConnected(name="fc1", output_dim=fc1_output,
-                                activation=tf.nn.tanh,
+                                activation=tf.nn.relu,
                                 weights_initialization=Xavier(seed=seed, use_gpu=self.use_gpu),
                                 bias_initialization=Constant(use_gpu=self.use_gpu)
                                 ))
 
         self.add(FullyConnected(name="fc2", output_dim=fc2_output,
-                                activation=tf.nn.tanh,
+                                activation=tf.nn.relu,
                                 weights_initialization=Xavier(seed=seed, use_gpu=self.use_gpu),
                                 bias_initialization=Constant(use_gpu=self.use_gpu)
                                 ))

bob/learn/tensorflow/network/SequenceNetwork.py

@@ -220,7 +220,12 @@ class SequenceNetwork(six.with_metaclass(abc.ABCMeta, object)):
         hdf5.set('input_divide', self.input_divide)
         hdf5.set('input_subtract', self.input_subtract)
 
-    def load(self, hdf5, shape=None, session=None):
+    def turn_gpu_onoff(self, state=True):
+        for k in self.sequence_net:
+            self.sequence_net[k].weights_initialization.use_gpu = state
+            self.sequence_net[k].bias_initialization.use_gpu = state
+
+    def load(self, hdf5, shape=None, session=None, batch=1):
         """
         Load the network
 
@@ -240,12 +245,15 @@ class SequenceNetwork(six.with_metaclass(abc.ABCMeta, object)):
         self.input_divide = hdf5.read('input_divide')
         self.input_subtract = hdf5.read('input_subtract')
 
-        # Saving the architecture
+        # Loading architecture
         self.sequence_net = pickle.loads(hdf5.read('architecture'))
         self.deployment_shape = hdf5.read('deployment_shape')
 
+        self.turn_gpu_onoff(False)
+
         if shape is None:
             shape = self.deployment_shape
+            shape[0] = batch
 
         # Loading variables
         place_holder = tf.placeholder(tf.float32, shape=shape, name="load")

bob/learn/tensorflow/script/train_siamese_casia_webface.py

@@ -60,11 +60,16 @@ def main():
     #                                               input_shape=[125, 125, 3],
     #                                               batch_size=BATCH_SIZE)
 
-    train_data_shuffler = TripletWithFastSelectionDisk(train_file_names, train_labels,
+    #train_data_shuffler = TripletWithFastSelectionDisk(train_file_names, train_labels,
+    #                                                   input_shape=[112, 112, 3],
+    #                                                   batch_size=BATCH_SIZE)
+
+    train_data_shuffler = TripletDisk(train_file_names, train_labels,
                                                        input_shape=[112, 112, 3],
                                                        batch_size=BATCH_SIZE)
 
 
+
     # Preparing train set
     directory = "/idiap/temp/tpereira/DEEP_FACE/CASIA_WEBFACE/mobio/preprocessed"
     validation_objects = sorted(db_mobio.objects(protocol="male", groups="dev"), key=lambda x: x.id)
@@ -91,13 +96,13 @@ def main():
     #                         snapshot=VALIDATION_TEST,
     #                         optimizer=optimizer)
 
-    loss = TripletLoss(margin=0.5)
+    loss = TripletLoss(margin=0.2)
     trainer = TripletTrainer(architecture=architecture, loss=loss,
                              iterations=ITERATIONS,
-                             base_learning_rate=0.1,
+                             base_learning_rate=0.05,
                              prefetch=False,
-                             temp_dir="./LOGS_CASIA/triplet-cnn-fast-selection")
+                             temp_dir="/idiap/temp/tpereira/CNN_MODELS/triplet-cnn-RANDOM-selection-gpu")
 
 
-    trainer.train(train_data_shuffler, validation_data_shuffler)
-    #trainer.train(train_data_shuffler)
+    #trainer.train(train_data_shuffler, validation_data_shuffler)
+    trainer.train(train_data_shuffler)

bob/learn/tensorflow/trainers/Trainer.py

@@ -14,8 +14,8 @@ from tensorflow.core.framework import summary_pb2
 import time
 from bob.learn.tensorflow.datashuffler.OnlineSampling import OnLineSampling
 
-os.environ["CUDA_VISIBLE_DEVICES"] = "3,2,0,1"
-#os.environ["CUDA_VISIBLE_DEVICES"] = ""
+#os.environ["CUDA_VISIBLE_DEVICES"] = "1,3,0,2"
+os.environ["CUDA_VISIBLE_DEVICES"] = ""
 
 logger = bob.core.log.setup("bob.learn.tensorflow")