Spaces:
Running
on
Zero
Running
on
Zero
Update gradio_tabs/img_edit.py
Browse files- gradio_tabs/img_edit.py +55 -18
gradio_tabs/img_edit.py
CHANGED
@@ -42,6 +42,7 @@ def load_image(img, size):
|
|
42 |
w, h = img.size
|
43 |
img = img.resize((size, size))
|
44 |
img = np.asarray(img)
|
|
|
45 |
img = np.transpose(img, (2, 0, 1)) # 3 x 256 x 256
|
46 |
|
47 |
return img / 255.0, w, h
|
@@ -55,40 +56,76 @@ def img_preprocessing(img_path, size):
|
|
55 |
return imgs_norm, w, h
|
56 |
|
57 |
|
58 |
-
def resize(img, size):
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
|
63 |
-
|
64 |
|
65 |
|
66 |
-
def resize_back(img, w, h):
|
67 |
-
|
68 |
-
|
69 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
70 |
|
|
|
|
|
|
|
71 |
return transform(img)
|
72 |
|
73 |
|
|
|
|
|
|
|
|
|
|
|
|
|
74 |
def img_denorm(img):
|
75 |
-
img = img.clamp(-1, 1)
|
76 |
img = (img - img.min()) / (img.max() - img.min())
|
77 |
|
78 |
return img
|
79 |
|
80 |
|
81 |
-
def img_postprocessing(image, w, h):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
|
83 |
-
|
84 |
-
image = image.permute(0, 2, 3, 1)
|
85 |
-
edited_image = img_denorm(image)
|
86 |
-
img_output = (edited_image[0].numpy() * 255).astype(np.uint8)
|
87 |
|
88 |
-
|
89 |
-
|
90 |
-
|
|
|
91 |
|
|
|
|
|
92 |
|
93 |
def img_edit(gen, device):
|
94 |
|
|
|
42 |
w, h = img.size
|
43 |
img = img.resize((size, size))
|
44 |
img = np.asarray(img)
|
45 |
+
img = np.copy(img)
|
46 |
img = np.transpose(img, (2, 0, 1)) # 3 x 256 x 256
|
47 |
|
48 |
return img / 255.0, w, h
|
|
|
56 |
return imgs_norm, w, h
|
57 |
|
58 |
|
59 |
+
# def resize(img, size):
|
60 |
+
# transform = torchvision.transforms.Compose([
|
61 |
+
# torchvision.transforms.Resize((size,size), antialias=True),
|
62 |
+
# ])
|
63 |
|
64 |
+
# return transform(img)
|
65 |
|
66 |
|
67 |
+
# def resize_back(img, w, h):
|
68 |
+
# transform = torchvision.transforms.Compose([
|
69 |
+
# torchvision.transforms.Resize((h, w), antialias=True),
|
70 |
+
# ])
|
71 |
+
|
72 |
+
# return transform(img)
|
73 |
+
|
74 |
+
# Pre-compile resize transforms for better performance
|
75 |
+
resize_transform_cache = {}
|
76 |
+
|
77 |
+
def get_resize_transform(size):
|
78 |
+
"""Get cached resize transform - creates once, reuses many times"""
|
79 |
+
if size not in resize_transform_cache:
|
80 |
+
# Only create the transform if it doesn't exist in cache
|
81 |
+
resize_transform_cache[size] = torchvision.transforms.Resize(
|
82 |
+
size,
|
83 |
+
interpolation=torchvision.transforms.InterpolationMode.BILINEAR,
|
84 |
+
antialias=True
|
85 |
+
)
|
86 |
+
return resize_transform_cache[size]
|
87 |
+
|
88 |
|
89 |
+
def resize(img, size):
|
90 |
+
"""Use cached resize transform"""
|
91 |
+
transform = get_resize_transform((size, size))
|
92 |
return transform(img)
|
93 |
|
94 |
|
95 |
+
def resize_back(img, w, h):
|
96 |
+
"""Use cached resize transform for back operation"""
|
97 |
+
transform = get_resize_transform((h, w))
|
98 |
+
return transform(img)
|
99 |
+
|
100 |
+
|
101 |
def img_denorm(img):
|
102 |
+
img = img.clamp(-1, 1)
|
103 |
img = (img - img.min()) / (img.max() - img.min())
|
104 |
|
105 |
return img
|
106 |
|
107 |
|
108 |
+
# def img_postprocessing(image, w, h):
|
109 |
+
|
110 |
+
# image = resize_back(image, w, h)
|
111 |
+
# image = image.permute(0, 2, 3, 1)
|
112 |
+
# edited_image = img_denorm(image)
|
113 |
+
# img_output = (edited_image[0].numpy() * 255).astype(np.uint8)
|
114 |
+
|
115 |
+
# with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as temp_file:
|
116 |
+
# imageio.imwrite(temp_file.name, img_output, quality=8)
|
117 |
+
# return temp_file.name
|
118 |
+
|
119 |
|
120 |
+
def img_postprocessing(img, w, h):
|
|
|
|
|
|
|
121 |
|
122 |
+
img = resize_back(img, w, h)
|
123 |
+
img = img_denorm(img)
|
124 |
+
img = img.squeeze(0).permute(1, 2, 0).contiguous() # contiguous() for fast transfer
|
125 |
+
img_output = (img.cpu().numpy() * 255).astype(np.uint8)
|
126 |
|
127 |
+
return img_output
|
128 |
+
|
129 |
|
130 |
def img_edit(gen, device):
|
131 |
|