asvspoo2017 corrections, tensorflow gmm for hdf5

bob/pad/voice/algorithm/tensorflow_algorithm.py

@@ -32,7 +32,7 @@ class TensorflowAlgo(Algorithm):
         )
 
         self.data_reader = DiskAudio([0], [0])
-#        self.session = tf.Session()
+        self.session = tf.Session()
         self.dnn_model = None
 
 #    def __del__(self):
@@ -47,8 +47,8 @@ class TensorflowAlgo(Algorithm):
     def load_projector(self, projector_file):
         logger.info("Loading pretrained model from {0}".format(projector_file))
         self.dnn_model = SequenceNetwork()
-#        self.dnn_model.load(bob.io.base.HDF5File(projector_file), session=self.session)
-        self.dnn_model.load(projector_file, True)
+        self.dnn_model.load_hdf5(bob.io.base.HDF5File(projector_file), shape=[1, 6560, 1])
+#        self.dnn_model.load(projector_file, True)
 
     def project_feature(self, feature):
 
@@ -58,8 +58,8 @@ class TensorflowAlgo(Algorithm):
         logger.debug(" .... And %d frames are extracted to pass into DNN model" % frames.shape[0])
         frames = numpy.reshape(frames, (frames.shape[0], -1, 1))
         forward_output = self.dnn_model(frames)
-#        return tf.nn.log_softmax(tf.nn.log_softmax(forward_output)).eval(session=self.session)
-        return forward_output
+        return tf.nn.log_softmax(tf.nn.log_softmax(forward_output)).eval(session=self.session)
+#        return forward_output
 
     def project(self, feature):
         """project(feature) -> projected

bob/pad/voice/database/asvspoof2017.py

@@ -48,7 +48,7 @@ class ASVspoof2017PadDatabase(PadDatabase):
 
         if names is None:
             return None
-        mapping = dict(zip(low_level_names, high_level_names))
+        mapping = dict(zip(high_level_names, low_level_names))
         if isinstance(names, str):
             return mapping.get(names)
         return [mapping[g] for g in names]
@@ -78,5 +78,6 @@ class ASVspoof2017PadDatabase(PadDatabase):
         Returns: A set of Files with the specified properties.
         """
         purposes = self.convert_purposes(purposes, ('genuine', 'spoof'), ('real', 'attack'))
+
         objects = self.__db.objects(protocol=protocol, groups=groups, purposes=purposes, **kwargs)
         return [ASVspoof2017PadFile(f) for f in objects]