diff --git a/bob/example/faceverify/dct_ubm.py b/bob/example/faceverify/dct_ubm.py
index e78a37b7bf3bbff6b417b3dc25ea4f6ca667406c..44bfd22cc0f17e34a61bc1fd67506fc4cc7097d1 100644
--- a/bob/example/faceverify/dct_ubm.py
+++ b/bob/example/faceverify/dct_ubm.py
@@ -157,7 +157,7 @@ def main():
 
   print("Extracting training features")
   training_features = {}
-  for key, image in training_images.iteritems():
+  for key, image in training_images.items():
     training_features[key] = extract_feature(image)
 
   print("Training UBM model")
@@ -178,7 +178,7 @@ def main():
     model_images = load_images(atnt_db, group = 'dev', purpose = 'enrol', client_id = model_id, database_directory = image_directory)
     models_for_current_id = {}
     # extract model features
-    for key, image in model_images.iteritems():
+    for key, image in model_images.items():
       models_for_current_id[key] = extract_feature(image)
     # enroll model for the current identity from these features
     model = enroll(models_for_current_id, ubm, gmm_trainer)
@@ -190,7 +190,7 @@ def main():
   print("Computing probe statistics")
   probe_images = load_images(atnt_db, group = 'dev', purpose = 'probe', database_directory = image_directory)
   probes = {}
-  for key, image in probe_images.iteritems():
+  for key, image in probe_images.items():
     # extract probe features
     probe_feature = extract_feature(image)
     # compute GMM statistics
@@ -205,8 +205,8 @@ def main():
   distance_function = bob.learn.misc.linear_scoring
 
   # iterate through models and probes and compute scores
-  for model_id, model_gmm in models.iteritems():
-    for probe_key, probe_stats in probes.iteritems():
+  for model_id, model_gmm in models.items():
+    for probe_key, probe_stats in probes.items():
       # compute score
       score = distance_function([model_gmm], ubm, [probe_stats])[0,0]
 
diff --git a/bob/example/faceverify/eigenface.py b/bob/example/faceverify/eigenface.py
index 7f8951437a25825543272b3c3ee4d154a4f753f0..075ec841e3d44fefc294c30973338726c684d906 100644
--- a/bob/example/faceverify/eigenface.py
+++ b/bob/example/faceverify/eigenface.py
@@ -109,21 +109,21 @@ def main():
 
   print("Extracting models")
   model_features = {}
-  for key, image in model_images.iteritems():
+  for key, image in model_images.items():
     model_features[key] = extract_feature(image, pca_machine)
 
   # enroll models from 5 features by simply storing all features
   model_ids = [client.id for client in atnt_db.clients(groups = 'dev')]
   models = dict((model_id, []) for model_id in model_ids) # note: py26 compat.
   # iterate over model features
-  for file_id, image in model_features.iteritems():
+  for file_id, image in model_features.items():
     model_id = atnt_db.get_client_id_from_file_id(file_id)
     # "enroll" model by collecting all model features of this client
     models[model_id].append(model_features[key])
 
   print("Extracting probes")
   probe_features = {}
-  for key, image in probe_images.iteritems():
+  for key, image in probe_images.items():
     probe_features[key] = extract_feature(image, pca_machine)
 
 
@@ -135,8 +135,8 @@ def main():
   print("Computing scores")
 
   # iterate through models and probes and compute scores
-  for model_id, model in models.iteritems():
-    for probe_key, probe_feature in probe_features.iteritems():
+  for model_id, model in models.items():
+    for probe_key, probe_feature in probe_features.items():
       # compute scores for all model features
       scores = [- DISTANCE_FUNCTION(model_feature, probe_feature) for model_feature in model]
       # the final score is the minimum distance (i.e., the maximum negative distance)
diff --git a/bob/example/faceverify/gabor_graph.py b/bob/example/faceverify/gabor_graph.py
index a7fd622275cfbd09003cf2ed2a867d3ad1a0cbe0..e4168c103b2c276cd80953f246ef62c17641b22a 100644
--- a/bob/example/faceverify/gabor_graph.py
+++ b/bob/example/faceverify/gabor_graph.py
@@ -106,21 +106,21 @@ def main():
 
   print("Extracting models")
   model_features = {}
-  for key, image in model_images.iteritems():
+  for key, image in model_images.items():
     model_features[key] = extract_feature(image, graph_machine)
 
   # enroll models from 5 features by simply storing all features
   model_ids = [client.id for client in atnt_db.clients(groups = 'dev')]
   models = dict((model_id, []) for model_id in model_ids) # note: py26 compat.
   # iterate over model features
-  for key, image in model_features.iteritems():
+  for key, image in model_features.items():
     model_id = atnt_db.get_client_id_from_file_id(key)
     # "enroll" model by collecting all model features of this client
     models[model_id].append(model_features[key])
 
   print("Extracting probes")
   probe_features = {}
-  for key, image in probe_images.iteritems():
+  for key, image in probe_images.items():
     probe_features[key] = extract_feature(image, graph_machine)
 
 
@@ -133,11 +133,11 @@ def main():
 
   # iterate through models and probes and compute scores
   model_count = 1
-  for model_id, model in models.iteritems():
+  for model_id, model in models.items():
     print("\rModel", model_count, "of", len(models), end='')
     sys.stdout.flush()
     model_count += 1
-    for probe_key, probe_feature in probe_features.iteritems():
+    for probe_key, probe_feature in probe_features.items():
       # compute local scores for each model gabor jet and each probe jet
       scores = numpy.ndarray((len(model), len(probe_feature)), dtype = numpy.float)
       for model_feature_index in range(len(model)):