It is recommended to use np.asarray instead of np.array to avoid unne…

experiments_mask_generation.py

@@ -6,7 +6,7 @@
 from extras.inpaint_mask import SAMOptions, generate_mask_from_image
 
 original_image = Image.open('cat.webp')
-image = np.array(original_image, dtype=np.uint8)
+image = np.asarray(original_image, dtype=np.uint8)
 
 sam_options = SAMOptions(
     dino_prompt='eye',

extras/facexlib/detection/align_trans.py

@@ -49,12 +49,12 @@ def get_reference_facial_points(output_size=None, inner_padding_factor=0.0, oute
                 inner_padding_factor))
     Returns:
     ----------
-        @reference_5point: 5x2 np.array
+        @reference_5point: 5x2 np.asarray
             each row is a pair of transformed coordinates (x, y)
     """
 
-    tmp_5pts = np.array(REFERENCE_FACIAL_POINTS)
-    tmp_crop_size = np.array(DEFAULT_CROP_SIZE)
+    tmp_5pts = np.asarray(REFERENCE_FACIAL_POINTS)
+    tmp_crop_size = np.asarray(DEFAULT_CROP_SIZE)
 
     # 0) make the inner region a square
     if default_square:
@@ -79,7 +79,7 @@ def get_reference_facial_points(output_size=None, inner_padding_factor=0.0, oute
     if ((inner_padding_factor > 0 or outer_padding[0] > 0 or outer_padding[1] > 0) and output_size is None):
         output_size = tmp_crop_size * \
             (1 + inner_padding_factor * 2).astype(np.int32)
-        output_size += np.array(outer_padding)
+        output_size += np.asarray(outer_padding)
     if not (outer_padding[0] < output_size[0] and outer_padding[1] < output_size[1]):
         raise FaceWarpException('Not (outer_padding[0] < output_size[0] and outer_padding[1] < output_size[1])')
 
@@ -90,7 +90,7 @@ def get_reference_facial_points(output_size=None, inner_padding_factor=0.0, oute
         tmp_crop_size += np.round(size_diff).astype(np.int32)
 
     # 2) resize the padded inner region
-    size_bf_outer_pad = np.array(output_size) - np.array(outer_padding) * 2
+    size_bf_outer_pad = np.asarray(output_size) - np.asarray(outer_padding) * 2
 
     if size_bf_outer_pad[0] * tmp_crop_size[1] != size_bf_outer_pad[1] * tmp_crop_size[0]:
         raise FaceWarpException('Must have (output_size - outer_padding)'
@@ -103,7 +103,7 @@ def get_reference_facial_points(output_size=None, inner_padding_factor=0.0, oute
     tmp_crop_size = size_bf_outer_pad
 
     # 3) add outer_padding to make output_size
-    reference_5point = tmp_5pts + np.array(outer_padding)
+    reference_5point = tmp_5pts + np.asarray(outer_padding)
     tmp_crop_size = output_size
 
     return reference_5point
@@ -116,13 +116,13 @@ def get_affine_transform_matrix(src_pts, dst_pts):
         get affine transform matrix 'tfm' from src_pts to dst_pts
     Parameters:
     ----------
-        @src_pts: Kx2 np.array
+        @src_pts: Kx2 np.asarray
             source points matrix, each row is a pair of coordinates (x, y)
-        @dst_pts: Kx2 np.array
+        @dst_pts: Kx2 np.asarray
             destination points matrix, each row is a pair of coordinates (x, y)
     Returns:
     ----------
-        @tfm: 2x3 np.array
+        @tfm: 2x3 np.asarray
             transform matrix from src_pts to dst_pts
     """
 
@@ -149,17 +149,17 @@ def warp_and_crop_face(src_img, facial_pts, reference_pts=None, crop_size=(96, 1
         apply affine transform 'trans' to uv
     Parameters:
     ----------
-        @src_img: 3x3 np.array
+        @src_img: 3x3 np.asarray
             input image
         @facial_pts: could be
             1)a list of K coordinates (x,y)
         or
-            2) Kx2 or 2xK np.array
+            2) Kx2 or 2xK np.asarray
             each row or col is a pair of coordinates (x, y)
         @reference_pts: could be
             1) a list of K coordinates (x,y)
         or
-            2) Kx2 or 2xK np.array
+            2) Kx2 or 2xK np.asarray
             each row or col is a pair of coordinates (x, y)
         or
             3) None

extras/facexlib/detection/matlab_cp2tform.py

@@ -18,14 +18,14 @@ def tformfwd(trans, uv):
 
     Parameters:
     ----------
-        @trans: 3x3 np.array
+        @trans: 3x3 np.asarray
             transform matrix
-        @uv: Kx2 np.array
+        @uv: Kx2 np.asarray
             each row is a pair of coordinates (x, y)
 
     Returns:
     ----------
-        @xy: Kx2 np.array
+        @xy: Kx2 np.asarray
             each row is a pair of transformed coordinates (x, y)
     """
     uv = np.hstack((uv, np.ones((uv.shape[0], 1))))
@@ -42,14 +42,14 @@ def tforminv(trans, uv):
 
     Parameters:
     ----------
-        @trans: 3x3 np.array
+        @trans: 3x3 np.asarray
             transform matrix
-        @uv: Kx2 np.array
+        @uv: Kx2 np.asarray
             each row is a pair of coordinates (x, y)
 
     Returns:
     ----------
-        @xy: Kx2 np.array
+        @xy: Kx2 np.asarray
             each row is a pair of inverse-transformed coordinates (x, y)
     """
     Tinv = inv(trans)
@@ -84,18 +84,18 @@ def findNonreflectiveSimilarity(uv, xy, options=None):
     tx = r[2]
     ty = r[3]
 
-    Tinv = np.array([[sc, -ss, 0], [ss, sc, 0], [tx, ty, 1]])
+    Tinv = np.asarray([[sc, -ss, 0], [ss, sc, 0], [tx, ty, 1]])
     T = inv(Tinv)
-    T[:, 2] = np.array([0, 0, 1])
+    T[:, 2] = np.asarray([0, 0, 1])
 
     return T, Tinv
 
 
 def findSimilarity(uv, xy, options=None):
     options = {'K': 2}
 
-    #    uv = np.array(uv)
-    #    xy = np.array(xy)
+    #    uv = np.asarray(uv)
+    #    xy = np.asarray(xy)
 
     # Solve for trans1
     trans1, trans1_inv = findNonreflectiveSimilarity(uv, xy, options)
@@ -109,7 +109,7 @@ def findSimilarity(uv, xy, options=None):
     trans2r, trans2r_inv = findNonreflectiveSimilarity(uv, xyR, options)
 
     # manually reflect the tform to undo the reflection done on xyR
-    TreflectY = np.array([[-1, 0, 0], [0, 1, 0], [0, 0, 1]])
+    TreflectY = np.asarray([[-1, 0, 0], [0, 1, 0], [0, 0, 1]])
 
     trans2 = np.dot(trans2r, TreflectY)
 
@@ -140,9 +140,9 @@ def get_similarity_transform(src_pts, dst_pts, reflective=True):
 
     Parameters:
     ----------
-        @src_pts: Kx2 np.array
+        @src_pts: Kx2 np.asarray
             source points, each row is a pair of coordinates (x, y)
-        @dst_pts: Kx2 np.array
+        @dst_pts: Kx2 np.asarray
             destination points, each row is a pair of transformed
             coordinates (x, y)
         @reflective: True or False
@@ -153,9 +153,9 @@ def get_similarity_transform(src_pts, dst_pts, reflective=True):
 
     Returns:
     ----------
-       @trans: 3x3 np.array
+       @trans: 3x3 np.asarray
             transform matrix from uv to xy
-        trans_inv: 3x3 np.array
+        trans_inv: 3x3 np.asarray
             inverse of trans, transform matrix from xy to uv
     """
 
@@ -181,12 +181,12 @@ def cvt_tform_mat_for_cv2(trans):
 
     Parameters:
     ----------
-        @trans: 3x3 np.array
+        @trans: 3x3 np.asarray
             transform matrix from uv to xy
 
     Returns:
     ----------
-        @cv2_trans: 2x3 np.array
+        @cv2_trans: 2x3 np.asarray
             transform matrix from src_pts to dst_pts, could be directly used
             for cv2.warpAffine()
     """
@@ -209,9 +209,9 @@ def get_similarity_transform_for_cv2(src_pts, dst_pts, reflective=True):
 
     Parameters:
     ----------
-        @src_pts: Kx2 np.array
+        @src_pts: Kx2 np.asarray
             source points, each row is a pair of coordinates (x, y)
-        @dst_pts: Kx2 np.array
+        @dst_pts: Kx2 np.asarray
             destination points, each row is a pair of transformed
             coordinates (x, y)
         reflective: True or False
@@ -222,7 +222,7 @@ def get_similarity_transform_for_cv2(src_pts, dst_pts, reflective=True):
 
     Returns:
     ----------
-        @cv2_trans: 2x3 np.array
+        @cv2_trans: 2x3 np.asarray
             transform matrix from src_pts to dst_pts, could be directly used
             for cv2.warpAffine()
     """
@@ -276,8 +276,8 @@ def get_similarity_transform_for_cv2(src_pts, dst_pts, reflective=True):
     x = [-1, 0, 4]
     y = [-1, -10, 4]
 
-    uv = np.array((u, v)).T
-    xy = np.array((x, y)).T
+    uv = np.asarray((u, v)).T
+    xy = np.asarray((x, y)).T
 
     print('\n--->uv:')
     print(uv)

extras/facexlib/detection/retinaface.py

@@ -245,7 +245,7 @@ def __align_multi(self, image, boxes, landmarks, limit=None):
         for landmark in landmarks:
             facial5points = [[landmark[2 * j], landmark[2 * j + 1]] for j in range(5)]
 
-            warped_face = warp_and_crop_face(np.array(image), facial5points, self.reference, crop_size=(112, 112))
+            warped_face = warp_and_crop_face(np.asarray(image), facial5points, self.reference, crop_size=(112, 112))
             faces.append(warped_face)
 
         return np.concatenate((boxes, landmarks), axis=1), faces
@@ -304,15 +304,15 @@ def batched_transform(self, frames, use_origin_size):
     def batched_detect_faces(self, frames, conf_threshold=0.8, nms_threshold=0.4, use_origin_size=True):
         """
         Arguments:
-            frames: a list of PIL.Image, or np.array(shape=[n, h, w, c],
+            frames: a list of PIL.Image, or np.asarray(shape=[n, h, w, c],
                 type=np.uint8, BGR format).
             conf_threshold: confidence threshold.
             nms_threshold: nms threshold.
             use_origin_size: whether to use origin size.
         Returns:
-            final_bounding_boxes: list of np.array ([n_boxes, 5],
+            final_bounding_boxes: list of np.asarray ([n_boxes, 5],
                 type=np.float32).
-            final_landmarks: list of np.array ([n_boxes, 10], type=np.float32).
+            final_landmarks: list of np.asarray ([n_boxes, 10], type=np.float32).
         """
         # self.t['forward_pass'].tic()
         frames, self.resize = self.batched_transform(frames, use_origin_size)
@@ -340,8 +340,8 @@ def batched_detect_faces(self, frames, conf_threshold=0.8, nms_threshold=0.4, us
             # ignore low scores
             pred, landm = pred[inds, :], landm[inds, :]
             if pred.shape[0] == 0:
-                final_bounding_boxes.append(np.array([], dtype=np.float32))
-                final_landmarks.append(np.array([], dtype=np.float32))
+                final_bounding_boxes.append(np.asarray([], dtype=np.float32))
+                final_landmarks.append(np.asarray([], dtype=np.float32))
                 continue
 
             # sort

extras/facexlib/utils/face_restoration_helper.py

@@ -33,12 +33,12 @@ def get_location(val, length):
 
 def get_center_face(det_faces, h=0, w=0, center=None):
     if center is not None:
-        center = np.array(center)
+        center = np.asarray(center)
     else:
-        center = np.array([w / 2, h / 2])
+        center = np.asarray([w / 2, h / 2])
     center_dist = []
     for det_face in det_faces:
-        face_center = np.array([(det_face[0] + det_face[2]) / 2, (det_face[1] + det_face[3]) / 2])
+        face_center = np.asarray([(det_face[0] + det_face[2]) / 2, (det_face[1] + det_face[3]) / 2])
         dist = np.linalg.norm(face_center - center)
         center_dist.append(dist)
     center_idx = center_dist.index(min(center_dist))
@@ -67,10 +67,10 @@ def __init__(self,
         self.face_size = (int(face_size * self.crop_ratio[1]), int(face_size * self.crop_ratio[0]))
 
         if self.template_3points:
-            self.face_template = np.array([[192, 240], [319, 240], [257, 371]])
+            self.face_template = np.asarray([[192, 240], [319, 240], [257, 371]])
         else:
             # standard 5 landmarks for FFHQ faces with 512 x 512
-            self.face_template = np.array([[192.98138, 239.94708], [318.90277, 240.1936], [256.63416, 314.01935],
+            self.face_template = np.asarray([[192.98138, 239.94708], [318.90277, 240.1936], [256.63416, 314.01935],
                                            [201.26117, 371.41043], [313.08905, 371.15118]])
         self.face_template = self.face_template * (face_size / 512.0)
         if self.crop_ratio[0] > 1:
@@ -144,9 +144,9 @@ def get_face_landmarks_5(self,
                 continue
 
             if self.template_3points:
-                landmark = np.array([[bbox[i], bbox[i + 1]] for i in range(5, 11, 2)])
+                landmark = np.asarray([[bbox[i], bbox[i + 1]] for i in range(5, 11, 2)])
             else:
-                landmark = np.array([[bbox[i], bbox[i + 1]] for i in range(5, 15, 2)])
+                landmark = np.asarray([[bbox[i], bbox[i + 1]] for i in range(5, 15, 2)])
             self.all_landmarks_5.append(landmark)
             self.det_faces.append(bbox[0:5])
         if len(self.det_faces) == 0:

extras/facexlib/utils/face_utils.py

@@ -64,7 +64,7 @@ def align_crop_face_landmarks(img,
         transform_size = output_size * 4
 
     # Parse landmarks
-    lm = np.array(landmarks)
+    lm = np.asarray(landmarks)
     if lm.shape[0] == 5 and lm_type == 'retinaface_5':
         eye_left = lm[0]
         eye_right = lm[1]
@@ -163,7 +163,7 @@ def align_crop_face_landmarks(img,
     # Transform use cv2
     h_ratio = shrink_ratio[0] / shrink_ratio[1]
     dst_h, dst_w = int(transform_size * h_ratio), transform_size
-    template = np.array([[0, 0], [0, dst_h], [dst_w, dst_h], [dst_w, 0]])
+    template = np.asarray([[0, 0], [0, dst_h], [dst_w, dst_h], [dst_w, 0]])
     # use cv2.LMEDS method for the equivalence to skimage transform
     # ref: https://blog.csdn.net/yichxi/article/details/115827338
     affine_matrix = cv2.estimateAffinePartial2D(quad, template, method=cv2.LMEDS)[0]
@@ -176,11 +176,11 @@ def align_crop_face_landmarks(img,
 
     if return_inverse_affine:
         dst_h, dst_w = int(output_size * h_ratio), output_size
-        template = np.array([[0, 0], [0, dst_h], [dst_w, dst_h], [dst_w, 0]])
+        template = np.asarray([[0, 0], [0, dst_h], [dst_w, dst_h], [dst_w, 0]])
         # use cv2.LMEDS method for the equivalence to skimage transform
         # ref: https://blog.csdn.net/yichxi/article/details/115827338
         affine_matrix = cv2.estimateAffinePartial2D(
-            quad_ori, np.array([[0, 0], [0, output_size], [dst_w, dst_h], [dst_w, 0]]), method=cv2.LMEDS)[0]
+            quad_ori, np.asarray([[0, 0], [0, output_size], [dst_w, dst_h], [dst_w, 0]]), method=cv2.LMEDS)[0]
         inverse_affine = cv2.invertAffineTransform(affine_matrix)
     else:
         inverse_affine = None
@@ -234,7 +234,7 @@ def paste_face_back(img, face, inverse_affine):
     bboxes = get_largest_face(bboxes, h, w)[0]
     visualize_detection(img_ori, [bboxes], f'tmp/{img_name}_det.png')
 
-    landmarks = np.array([[bboxes[i], bboxes[i + 1]] for i in range(5, 15, 2)])
+    landmarks = np.asarray([[bboxes[i], bboxes[i + 1]] for i in range(5, 15, 2)])
 
     cropped_face, inverse_affine = align_crop_face_landmarks(
         img_ori,

extras/inpaint_mask.py

@@ -104,7 +104,7 @@ def generate_mask_from_image(image: np.ndarray, mask_model: str = 'sam', extras=
             draw = ImageDraw.Draw(debug_dino_image)
             for box in boxes.numpy():
                 draw.rectangle(box.tolist(), fill="white")
-            return np.array(debug_dino_image), dino_detection_count, sam_detection_count, sam_detection_on_mask_count
+            return np.asarray(debug_dino_image), dino_detection_count, sam_detection_count, sam_detection_on_mask_count
 
         transformed_boxes = sam_predictor.transform.apply_boxes_torch(boxes, image.shape[:2])
         masks, _, _ = sam_predictor.predict_torch(
@@ -126,5 +126,5 @@ def generate_mask_from_image(image: np.ndarray, mask_model: str = 'sam', extras=
 
     final_mask_tensor = (final_mask_tensor > 0).to('cpu').numpy()
     mask_image = np.dstack((final_mask_tensor, final_mask_tensor, final_mask_tensor)) * 255
-    mask_image = np.array(mask_image, dtype=np.uint8)
+    mask_image = np.asarray(mask_image, dtype=np.uint8)
     return mask_image, dino_detection_count, sam_detection_count, sam_detection_on_mask_count

extras/wd14tagger.py

@@ -57,7 +57,7 @@ def default_interrogator(image_rgb, threshold=0.35, character_threshold=0.85, ex
     square = Image.new("RGB", (height, height), (255, 255, 255))
     square.paste(image, ((height-new_size[0])//2, (height-new_size[1])//2))
 
-    image = np.array(square).astype(np.float32)
+    image = np.asarray(square).astype(np.float32)
     image = image[:, :, ::-1]  # RGB -> BGR
     image = np.expand_dims(image, 0)
 

ldm_patched/contrib/external.py

@@ -1476,10 +1476,10 @@ def load_image(self, image):
             if i.mode == 'I':
                 i = i.point(lambda i: i * (1 / 255))
             image = i.convert("RGB")
-            image = np.array(image).astype(np.float32) / 255.0
+            image = np.asarray(image).astype(np.float32) / 255.0
             image = torch.from_numpy(image)[None,]
             if 'A' in i.getbands():
-                mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0
+                mask = np.asarray(i.getchannel('A')).astype(np.float32) / 255.0
                 mask = 1. - torch.from_numpy(mask)
             else:
                 mask = torch.zeros((64,64), dtype=torch.float32, device="cpu")
@@ -1536,7 +1536,7 @@ def load_image(self, image, channel):
         mask = None
         c = channel[0].upper()
         if c in i.getbands():
-            mask = np.array(i.getchannel(c)).astype(np.float32) / 255.0
+            mask = np.asarray(i.getchannel(c)).astype(np.float32) / 255.0
             mask = torch.from_numpy(mask)
             if c == 'A':
                 mask = 1. - mask