diff --git a/deepface/detectors/FaceDetector.py b/deepface/detectors/FaceDetector.py
index ff30f76..e12500d 100644
--- a/deepface/detectors/FaceDetector.py
+++ b/deepface/detectors/FaceDetector.py
@@ -6,21 +6,18 @@ from deepface.commons import distance
 
 def build_model(detector_backend):
 
-    if detector_backend == 'opencv':
-        face_detector = OpenCvWrapper.build_model()
+    backends = {
+        'opencv': OpenCvWrapper.build_model,
+        'ssd': SsdWrapper.build_model,
+        'dlib': DlibWrapper.build_model,
+        'mtcnn': MtcnnWrapper.build_model,
+        'retinaface': RetinaFaceWrapper.build_model
+    }
 
-    elif detector_backend == 'ssd':
-        face_detector = SsdWrapper.build_model()
-
-    elif detector_backend == 'dlib':
-        face_detector = DlibWrapper.build_model()
-
-    elif detector_backend == 'mtcnn':
-        face_detector = MtcnnWrapper.build_model()
-
-    elif detector_backend == 'retinaface':
-        face_detector = RetinaFaceWrapper.build_model()
+    face_detector = backends.get(detector_backend)
 
+    if face_detector:
+        face_detector = face_detector()
     else:
         raise ValueError("invalid detector_backend passed - " + detector_backend)
 
@@ -28,21 +25,18 @@ def build_model(detector_backend):
 
 def detect_face(face_detector, detector_backend, img):
 
-    if detector_backend == 'opencv':
-        face, region = OpenCvWrapper.detect_face(face_detector, img)
+    backends = {
+        'opencv': OpenCvWrapper.detect_face,
+        'ssd': SsdWrapper.detect_face,
+        'dlib': DlibWrapper.detect_face,
+        'mtcnn': MtcnnWrapper.detect_face,
+        'retinaface': RetinaFaceWrapper.detect_face
+    }
 
-    elif detector_backend == 'ssd':
-        face, region = SsdWrapper.detect_face(face_detector, img)
-
-    elif detector_backend == 'dlib':
-        face, region = DlibWrapper.detect_face(face_detector, img)
-
-    elif detector_backend == 'mtcnn':
-        face, region = MtcnnWrapper.detect_face(face_detector, img)
-
-    elif detector_backend == 'retinaface':
-        face, region = RetinaFaceWrapper.detect_face(face_detector, img)
+    detect_face = backends.get(detector_backend)
 
+    if detect_face:
+        face, region = detect_face(face_detector, img)
     else:
         raise ValueError("invalid detector_backend passed - " + detector_backend)
 
diff --git a/deepface/detectors/__init__.py b/deepface/detectors/__init__.py
new file mode 100644
index 0000000..e69de29