Merge pull request #1103 from serengil/feat-task-1103-bugs-and-enhanc…

…ements

Feat task 1103 bugs and enhancements

deepface/detectors/YuNet.py

@@ -22,8 +22,15 @@ def build_model(self) -> Any:
         """
 
         opencv_version = cv2.__version__.split(".")
+        if not len(opencv_version) >= 2:
+            raise ValueError(
+                f"OpenCv's version must have major and minor values but it is {opencv_version}"
+            )
+
+        opencv_version_major = int(opencv_version[0])
+        opencv_version_minor = int(opencv_version[1])
 
-        if len(opencv_version) > 2 and int(opencv_version[0]) == 4 and int(opencv_version[1]) < 8:
+        if opencv_version_major < 4 or (opencv_version_major == 4 and opencv_version_minor < 8):
             # min requirement: https://github.com/opencv/opencv_zoo/issues/172
             raise ValueError(f"YuNet requires opencv-python >= 4.8 but you have {cv2.__version__}")
 
@@ -67,9 +74,9 @@ def detect_faces(self, img: np.ndarray) -> List[FacialAreaRegion]:
         # resize image if it is too large (Yunet fails to detect faces on large input sometimes)
         # I picked 640 as a threshold because it is the default value of max_size in Yunet.
         resized = False
+        r = 1  # resize factor
         if height > 640 or width > 640:
             r = 640.0 / max(height, width)
-            original_image = img.copy()
             img = cv2.resize(img, (int(width * r), int(height * r)))
             height, width = img.shape[0], img.shape[1]
             resized = True
@@ -93,16 +100,12 @@ def detect_faces(self, img: np.ndarray) -> List[FacialAreaRegion]:
             left eye, nose tip, the right corner and left corner of the mouth respectively.
             """
             (x, y, w, h, x_re, y_re, x_le, y_le) = list(map(int, face[:8]))
-            left_eye = (x_re, y_re)
-            right_eye = (x_le, y_le)
 
-            # Yunet returns negative coordinates if it thinks part of
-            # the detected face is outside the frame.
-            # We set the coordinate to 0 if they are negative.
+            # YuNet returns negative coordinates if it thinks part of the detected face
+            # is outside the frame.
             x = max(x, 0)
             y = max(y, 0)
             if resized:
-                img = original_image
                 x, y, w, h = int(x / r), int(y / r), int(w / r), int(h / r)
                 x_re, y_re, x_le, y_le = (
                     int(x_re / r),
@@ -118,8 +121,8 @@ def detect_faces(self, img: np.ndarray) -> List[FacialAreaRegion]:
                 w=w,
                 h=h,
                 confidence=confidence,
-                left_eye=left_eye,
-                right_eye=right_eye,
+                left_eye=(x_re, y_re),
+                right_eye=(x_le, y_le),
             )
             resp.append(facial_area)
         return resp

deepface/modules/verification.py

@@ -93,6 +93,7 @@ def verify(
     model: FacialRecognition = modeling.build_model(model_name)
     dims = model.output_shape
 
+    # extract faces from img1
     if isinstance(img1_path, list):
         # given image is already pre-calculated embedding
         if not all(isinstance(dim, float) for dim in img1_path):
@@ -115,16 +116,20 @@ def verify(
         img1_embeddings = [img1_path]
         img1_facial_areas = [None]
     else:
-        img1_embeddings, img1_facial_areas = __extract_faces_and_embeddings(
-            img_path=img1_path,
-            model_name=model_name,
-            detector_backend=detector_backend,
-            enforce_detection=enforce_detection,
-            align=align,
-            expand_percentage=expand_percentage,
-            normalization=normalization,
-        )
+        try:
+            img1_embeddings, img1_facial_areas = __extract_faces_and_embeddings(
+                img_path=img1_path,
+                model_name=model_name,
+                detector_backend=detector_backend,
+                enforce_detection=enforce_detection,
+                align=align,
+                expand_percentage=expand_percentage,
+                normalization=normalization,
+            )
+        except ValueError as err:
+            raise ValueError("Exception while processing img1_path") from err
 
+    # extract faces from img2
     if isinstance(img2_path, list):
         # given image is already pre-calculated embedding
         if not all(isinstance(dim, float) for dim in img2_path):
@@ -147,15 +152,18 @@ def verify(
         img2_embeddings = [img2_path]
         img2_facial_areas = [None]
     else:
-        img2_embeddings, img2_facial_areas = __extract_faces_and_embeddings(
-            img_path=img2_path,
-            model_name=model_name,
-            detector_backend=detector_backend,
-            enforce_detection=enforce_detection,
-            align=align,
-            expand_percentage=expand_percentage,
-            normalization=normalization,
-        )
+        try:
+            img2_embeddings, img2_facial_areas = __extract_faces_and_embeddings(
+                img_path=img2_path,
+                model_name=model_name,
+                detector_backend=detector_backend,
+                enforce_detection=enforce_detection,
+                align=align,
+                expand_percentage=expand_percentage,
+                normalization=normalization,
+            )
+        except ValueError as err:
+            raise ValueError("Exception while processing img2_path") from err
 
     no_facial_area = {
         "x": None,
@@ -218,18 +226,15 @@ def __extract_faces_and_embeddings(
     model: FacialRecognition = modeling.build_model(model_name)
     target_size = model.input_shape
 
-    try:
-        img_objs = detection.extract_faces(
-            img_path=img_path,
-            target_size=target_size,
-            detector_backend=detector_backend,
-            grayscale=False,
-            enforce_detection=enforce_detection,
-            align=align,
-            expand_percentage=expand_percentage,
-        )
-    except ValueError as err:
-        raise ValueError("Exception while processing img1_path") from err
+    img_objs = detection.extract_faces(
+        img_path=img_path,
+        target_size=target_size,
+        detector_backend=detector_backend,
+        grayscale=False,
+        enforce_detection=enforce_detection,
+        align=align,
+        expand_percentage=expand_percentage,
+    )
 
     # find embeddings for each face
     for img_obj in img_objs: