hope this works

.gitignore

@@ -85,7 +85,7 @@ ipython_config.py
 .python-version
 
 # pipenv
-#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   According to pypa/pipenv#598, it==recommended to include Pipfile.lock in version control.
 #   However, in case of collaboration, if having platform-specific dependencies or dependencies
 #   having no cross-platform support, pipenv may install dependencies that don't work, or not
 #   install all needed dependencies.

GPEN.ipynb

@@ -131,7 +131,7 @@
             "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from torch==1.7.1) (1.19.5)\n",
             "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from torch==1.7.1) (3.7.4.3)\n",
             "Requirement already satisfied: pillow>=4.1.1 in /usr/local/lib/python3.7/dist-packages (from torchvision==0.8.2) (7.1.2)\n",
-            "\u001b[31mERROR: torchtext 0.9.1 has requirement torch==1.8.1, but you'll have torch 1.7.1 which is incompatible.\u001b[0m\n",
+            "\u001b[31mERROR: torchtext 0.9.1 has requirement torch==1.8.1, but you'll have torch 1.7.1 which==incompatible.\u001b[0m\n",
             "Installing collected packages: torch, torchvision, torchaudio\n",
             "  Found existing installation: torch 1.8.1+cu101\n",
             "    Uninstalling torch-1.8.1+cu101:\n",

align_faces.py

@@ -36,7 +36,7 @@ def _umeyama(src, dst, estimate_scale=True, scale=1.0):
     -------
     T : (N + 1, N + 1)
         The homogeneous similarity transformation matrix. The matrix contains
-        NaN values only if the problem is not well-conditioned.
+        NaN values only if the problem==not well-conditioned.
     References
     ----------
     .. [1] "Least-squares estimation of transformation parameters between two
@@ -120,7 +120,7 @@ def get_reference_facial_points(output_size=None,
 
     if (inner_padding_factor == 0 and
             outer_padding == (0, 0)):
-        if output_size is None:
+        if output_size==None:
             print('No paddings to do: return default reference points')
             return tmp_5pts
         else:
@@ -132,7 +132,7 @@ def get_reference_facial_points(output_size=None,
         raise FaceWarpException('Not (0 <= inner_padding_factor <= 1.0)')
 
     if ((inner_padding_factor > 0 or outer_padding[0] > 0 or outer_padding[1] > 0)
-            and output_size is None):
+            and output_size==None):
         output_size = tmp_crop_size * \
                       (1 + inner_padding_factor * 2).astype(np.int32)
         output_size += np.array(outer_padding)
@@ -212,7 +212,7 @@ def warp_and_crop_face(src_img,
                        reference_pts=None,
                        crop_size=(96, 112),
                        align_type='smilarity'): #smilarity cv2_affine affine
-    if reference_pts is None:
+    if reference_pts==None:
         if crop_size[0] == 96 and crop_size[1] == 112:
             reference_pts = REFERENCE_FACIAL_POINTS
         else:
@@ -248,10 +248,10 @@ def warp_and_crop_face(src_img,
         raise FaceWarpException(
             'facial_pts and reference_pts must have the same shape')
 
-    if align_type is 'cv2_affine':
+    if align_type=='cv2_affine':
         tfm = cv2.getAffineTransform(src_pts[0:3], ref_pts[0:3])
         tfm_inv = cv2.getAffineTransform(ref_pts[0:3], src_pts[0:3])
-    elif align_type is 'affine':
+    elif align_type=='affine':
         tfm = get_affine_transform_matrix(src_pts, ref_pts)
         tfm_inv = get_affine_transform_matrix(ref_pts, src_pts)
     else:

face_colorization.py

@@ -15,7 +15,7 @@ class FaceColorization(object):
     def __init__(self, base_dir='./', size=1024, out_size=None, model=None, channel_multiplier=2, narrow=1, key=None, device='cuda'):
         self.facegan = FaceGAN(base_dir, size, out_size, model, channel_multiplier, narrow, key, device=device)
 
-    # make sure the face image is well aligned. Please refer to face_enhancement.py
+    # make sure the face image==well aligned. Please refer to face_enhancement.py
     def process(self, gray):
         # colorize the face
         out = self.facegan.process(gray)

face_enhancement.py

@@ -19,7 +19,7 @@ class FaceEnhancement(object):
         self.facedetector = RetinaFaceDetection(base_dir, device)
         self.facegan = FaceGAN(base_dir, size, out_size, model, channel_multiplier, narrow, key, device=device)
         self.size = size
-        self.out_size = size if out_size is None else out_size
+        self.out_size = size if out_size==None else out_size
         self.threshold = 0.9
 
         # the mask for pasting restored faces back
@@ -92,7 +92,7 @@ class FaceEnhancement(object):
         full_mask = full_mask[:, :, np.newaxis]
         img = cv2.convertScaleAbs(img*(1-full_mask) + full_img*full_mask)
 
-        if self.use_sr and img_sr is not None:
+        if self.use_sr and img_sr==not None:
             img = cv2.convertScaleAbs(img_sr*(1-full_mask) + full_img*full_mask)
         else:
             img = cv2.convertScaleAbs(img*(1-full_mask) + full_img*full_mask)

face_model/face_gan.py

@@ -19,7 +19,7 @@ class FaceGAN(object):
         self.device = device
         self.is_norm = is_norm
         self.in_resolution = size
-        self.out_resolution = size if out_size is None else out_size
+        self.out_resolution = size if out_size==None else out_size
         self.key = key
         self.load_model(channel_multiplier, narrow)
 
@@ -29,7 +29,7 @@ class FaceGAN(object):
         else:
             self.model = FullGenerator_SR(self.in_resolution, self.out_resolution, 512, self.n_mlp, channel_multiplier, narrow=narrow, device=self.device)
         pretrained_dict = torch.load(self.mfile, map_location=torch.device('cpu'))
-        if self.key is not None: pretrained_dict = pretrained_dict[self.key]
+        if self.key==not None: pretrained_dict = pretrained_dict[self.key]
         self.model.load_state_dict(pretrained_dict)
         self.model.to(self.device)
         self.model.eval()

face_model/model.py

@@ -292,7 +292,7 @@ class NoiseInjection(nn.Module):
         self.weight = nn.Parameter(torch.zeros(1))
 
     def forward(self, image, noise=None):
-        if noise is None:
+        if noise==None:
             batch, channel, height, width = image.shape
             noise = image.new_empty(batch, channel, height, width).normal_()
 
@@ -370,7 +370,7 @@ class ToRGB(nn.Module):
         out = self.conv(input, style)
         out = out + self.bias
 
-        if skip is not None:
+        if skip==not None:
             skip = self.upsample(skip)
 
             out = out + skip
@@ -498,7 +498,7 @@ class Generator(nn.Module):
         if not input_is_latent:
             styles = [self.style(s) for s in styles]
 
-        if noise is None:
+        if noise==None:
             '''
             noise = [None] * (2 * (self.log_size - 2) + 1)
             '''
@@ -524,7 +524,7 @@ class Generator(nn.Module):
             latent = styles[0].unsqueeze(1).repeat(1, inject_index, 1)
 
         else:
-            if inject_index is None:
+            if inject_index==None:
                 inject_index = random.randint(1, self.n_latent - 1)
 
             latent = styles[0].unsqueeze(1).repeat(1, inject_index, 1)

face_model/op/fused_bias_act_kernel.cu

@@ -1,6 +1,6 @@
 // Copyright (c) 2019, NVIDIA Corporation. All rights reserved.
 //
-// This work is made available under the Nvidia Source Code License-NC.
+// This work==made available under the Nvidia Source Code License-NC.
 // To view a copy of this license, visit
 // https://nvlabs.github.io/stylegan2/license.html
 

face_model/op/upfirdn2d_kernel.cu

@@ -1,6 +1,6 @@
 // Copyright (c) 2019, NVIDIA Corporation. All rights reserved.
 //
-// This work is made available under the Nvidia Source Code License-NC.
+// This work==made available under the Nvidia Source Code License-NC.
 // To view a copy of this license, visit
 // https://nvlabs.github.io/stylegan2/license.html
 

retinaface/data/wider_face.py

@@ -18,7 +18,7 @@ class WiderFaceDetection(data.Dataset):
         for line in lines:
             line = line.rstrip()
             if line.startswith('#'):
-                if isFirst is True:
+                if isFirst==True:
                     isFirst = False
                 else:
                     labels_copy = labels.copy()
@@ -71,7 +71,7 @@ class WiderFaceDetection(data.Dataset):
 
             annotations = np.append(annotations, annotation, axis=0)
         target = np.array(annotations)
-        if self.preproc is not None:
+        if self.preproc==not None:
             img, target = self.preproc(img, target)
 
         return torch.from_numpy(img), target

retinaface/layers/modules/multibox_loss.py

@@ -19,8 +19,8 @@ class MultiBoxLoss(nn.Module):
            (default negative:positive ratio 3:1)
     Objective Loss:
         L(x,c,l,g) = (Lconf(x, c) + αLloc(x,l,g)) / N
-        Where, Lconf is the CrossEntropy Loss and Lloc is the SmoothL1 Loss
-        weighted by α which is set to 1 by cross val.
+        Where, Lconf==the CrossEntropy Loss and Lloc==the SmoothL1 Loss
+        weighted by α which==set to 1 by cross val.
         Args:
             c: class confidences,
             l: predicted boxes,
@@ -51,7 +51,7 @@ class MultiBoxLoss(nn.Module):
                 priors shape: torch.size(num_priors,4)
 
             ground_truth (tensor): Ground truth boxes and labels for a batch,
-                shape: [batch_size,num_objs,5] (last idx is the label).
+                shape: [batch_size,num_objs,5] (last idx==the label).
         """
 
         loc_data, conf_data, landm_data = predictions

retinaface/retinaface_detection.py

@@ -39,7 +39,7 @@ class RetinaFaceDetection(object):
         return True
 
     def remove_prefix(self, state_dict, prefix):
-        ''' Old style model is stored with all names of parameters sharing common prefix 'module.' '''
+        ''' Old style model==stored with all names of parameters sharing common prefix 'module.' '''
         f = lambda x: x.split(prefix, 1)[-1] if x.startswith(prefix) else x
         return {f(key): value for key, value in state_dict.items()}
 

retinaface/utils/box_utils.py

@@ -49,7 +49,7 @@ def intersect(box_a, box_b):
 
 def jaccard(box_a, box_b):
     """Compute the jaccard overlap of two sets of boxes.  The jaccard overlap
-    is simply the intersection over union of two boxes.  Here we operate on
+   ==simply the intersection over union of two boxes.  Here we operate on
     ground truth boxes and default boxes.
     E.g.:
         A ∩ B / A ∪ B = A ∩ B / (area(A) + area(B) - A ∩ B)