diff --git a/bob/bio/face/annotator/bobipmtcnn.py b/bob/bio/face/annotator/bobipmtcnn.py
index b207c685fcf1796cc75727afffee62af69b7d290..9e61a06e823587d776b942d9971ce12355fa7ed6 100644
--- a/bob/bio/face/annotator/bobipmtcnn.py
+++ b/bob/bio/face/annotator/bobipmtcnn.py
@@ -1,16 +1,16 @@
-from . import Base, bounding_box_to_annotations
+from . import Base
 
 
 class BobIpMTCNN(Base):
-    """Annotator using bob.ip.mtcnn"""
+    """Annotator using mtcnn in bob.ip.tensorflow_extractor"""
 
     def __init__(self, **kwargs):
         super(BobIpMTCNN, self).__init__(**kwargs)
-        from bob.ip.mtcnn import FaceDetector
-        self.detector = FaceDetector()
+        from bob.ip.tensorflow_extractor import MTCNN
+        self.detector = MTCNN()
 
     def annotate(self, image, **kwargs):
-        """Annotates an image using bob.ip.mtcnn
+        """Annotates an image using mtcnn
 
         Parameters
         ----------
@@ -23,10 +23,7 @@ class BobIpMTCNN(Base):
         -------
         dict
             Annotations contain: (topleft, bottomright, leye, reye, nose,
-            mouthleft, mouthright).
+            mouthleft, mouthright, quality).
         """
-        bounding_box, landmarks = self.detector.detect_single_face(image)
-        if not landmarks:
-            return None
-        landmarks.update(bounding_box_to_annotations(bounding_box))
-        return landmarks
+        # return the annotations for the first/largest face.
+        return self.detector.annotations(image)[0]