Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -10,28 +10,28 @@ import torchvision.transforms as transforms
|
|
10 |
|
11 |
photos_folder = "Photos"
|
12 |
|
13 |
-
#
|
14 |
repo_id = "Kiwinicki/sat2map-generator"
|
15 |
generator_path = hf_hub_download(repo_id=repo_id, filename="generator.pth")
|
16 |
config_path = hf_hub_download(repo_id=repo_id, filename="config.json")
|
17 |
model_path = hf_hub_download(repo_id=repo_id, filename="model.py")
|
18 |
|
19 |
-
#
|
20 |
sys.path.append(os.path.dirname(model_path))
|
21 |
from model import Generator
|
22 |
|
23 |
-
#
|
24 |
with open(config_path, "r") as f:
|
25 |
config_dict = json.load(f)
|
26 |
cfg = OmegaConf.create(config_dict)
|
27 |
|
28 |
-
#
|
29 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
30 |
generator = Generator(cfg).to(device)
|
31 |
generator.load_state_dict(torch.load(generator_path, map_location=device))
|
32 |
generator.eval()
|
33 |
|
34 |
-
#
|
35 |
transform = transforms.Compose([
|
36 |
transforms.Resize((256, 256)),
|
37 |
transforms.ToTensor(),
|
@@ -39,86 +39,86 @@ transform = transforms.Compose([
|
|
39 |
])
|
40 |
|
41 |
def process_image(image):
|
42 |
-
|
|
|
|
|
|
|
43 |
image_tensor = transform(image).unsqueeze(0).to(device)
|
44 |
|
45 |
-
#
|
46 |
with torch.no_grad():
|
47 |
output_tensor = generator(image_tensor)
|
48 |
|
49 |
-
#
|
50 |
output_image = output_tensor.squeeze(0).cpu()
|
51 |
-
output_image = output_image * 0.5 + 0.5 #
|
52 |
output_image = transforms.ToPILImage()(output_image)
|
53 |
|
54 |
return output_image
|
55 |
|
56 |
-
|
57 |
def load_images_from_folder(folder):
|
58 |
images = []
|
|
|
|
|
|
|
|
|
59 |
for filename in os.listdir(folder):
|
60 |
if filename.lower().endswith(('.png', '.jpg', '.jpeg')):
|
61 |
img_path = os.path.join(folder, filename)
|
62 |
-
|
63 |
-
|
|
|
|
|
|
|
64 |
return images
|
65 |
|
66 |
-
|
67 |
-
|
68 |
-
def load_image_from_gallery(images, index):
|
69 |
-
if images and 0 <= index < len(images):
|
70 |
-
image = images[index]
|
71 |
-
if isinstance(image, tuple):
|
72 |
-
image = image[0]
|
73 |
-
return image
|
74 |
-
return None
|
75 |
-
|
76 |
-
|
77 |
-
def gallery_click_event(images, evt: gr.SelectData):
|
78 |
-
index = evt.index
|
79 |
-
selected_img = load_image_from_gallery(images, index)
|
80 |
-
return selected_img
|
81 |
-
|
82 |
-
|
83 |
-
def clear_image():
|
84 |
-
return None
|
85 |
-
|
86 |
-
|
87 |
def app():
|
88 |
images = load_images_from_folder(photos_folder)
|
|
|
89 |
|
90 |
-
with gr.Blocks(
|
91 |
with gr.Row():
|
92 |
with gr.Column():
|
93 |
-
|
94 |
clear_button = gr.Button("Clear")
|
95 |
|
96 |
with gr.Column():
|
97 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
98 |
|
99 |
with gr.Column():
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
|
104 |
-
|
105 |
-
|
|
|
|
|
|
|
|
|
|
|
106 |
)
|
107 |
-
|
108 |
-
|
|
|
109 |
fn=process_image,
|
110 |
-
inputs=
|
111 |
-
outputs=
|
112 |
)
|
113 |
-
|
|
|
114 |
clear_button.click(
|
115 |
-
fn=
|
116 |
-
|
117 |
-
outputs=selected_image
|
118 |
)
|
119 |
|
120 |
demo.launch()
|
121 |
|
122 |
-
|
123 |
if __name__ == "__main__":
|
124 |
-
app()
|
|
|
10 |
|
11 |
photos_folder = "Photos"
|
12 |
|
13 |
+
# Download model and config
|
14 |
repo_id = "Kiwinicki/sat2map-generator"
|
15 |
generator_path = hf_hub_download(repo_id=repo_id, filename="generator.pth")
|
16 |
config_path = hf_hub_download(repo_id=repo_id, filename="config.json")
|
17 |
model_path = hf_hub_download(repo_id=repo_id, filename="model.py")
|
18 |
|
19 |
+
# Add path to model
|
20 |
sys.path.append(os.path.dirname(model_path))
|
21 |
from model import Generator
|
22 |
|
23 |
+
# Load configuration
|
24 |
with open(config_path, "r") as f:
|
25 |
config_dict = json.load(f)
|
26 |
cfg = OmegaConf.create(config_dict)
|
27 |
|
28 |
+
# Initialize model
|
29 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
30 |
generator = Generator(cfg).to(device)
|
31 |
generator.load_state_dict(torch.load(generator_path, map_location=device))
|
32 |
generator.eval()
|
33 |
|
34 |
+
# Transformations
|
35 |
transform = transforms.Compose([
|
36 |
transforms.Resize((256, 256)),
|
37 |
transforms.ToTensor(),
|
|
|
39 |
])
|
40 |
|
41 |
def process_image(image):
|
42 |
+
if image is None:
|
43 |
+
return None
|
44 |
+
|
45 |
+
# Convert to tensor
|
46 |
image_tensor = transform(image).unsqueeze(0).to(device)
|
47 |
|
48 |
+
# Inference
|
49 |
with torch.no_grad():
|
50 |
output_tensor = generator(image_tensor)
|
51 |
|
52 |
+
# Prepare output
|
53 |
output_image = output_tensor.squeeze(0).cpu()
|
54 |
+
output_image = output_image * 0.5 + 0.5 # Denormalization
|
55 |
output_image = transforms.ToPILImage()(output_image)
|
56 |
|
57 |
return output_image
|
58 |
|
|
|
59 |
def load_images_from_folder(folder):
|
60 |
images = []
|
61 |
+
if not os.path.exists(folder):
|
62 |
+
os.makedirs(folder)
|
63 |
+
return images
|
64 |
+
|
65 |
for filename in os.listdir(folder):
|
66 |
if filename.lower().endswith(('.png', '.jpg', '.jpeg')):
|
67 |
img_path = os.path.join(folder, filename)
|
68 |
+
try:
|
69 |
+
img = Image.open(img_path)
|
70 |
+
images.append((img, filename))
|
71 |
+
except Exception as e:
|
72 |
+
print(f"Error loading {filename}: {e}")
|
73 |
return images
|
74 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
75 |
def app():
|
76 |
images = load_images_from_folder(photos_folder)
|
77 |
+
gallery_images = [img[0] for img in images] if images else []
|
78 |
|
79 |
+
with gr.Blocks() as demo:
|
80 |
with gr.Row():
|
81 |
with gr.Column():
|
82 |
+
input_image = gr.Image(label="Input Image", type="pil")
|
83 |
clear_button = gr.Button("Clear")
|
84 |
|
85 |
with gr.Column():
|
86 |
+
gallery = gr.Gallery(
|
87 |
+
label="Image Gallery",
|
88 |
+
value=gallery_images,
|
89 |
+
columns=3,
|
90 |
+
rows=2,
|
91 |
+
height="auto"
|
92 |
+
).style(grid=3)
|
93 |
|
94 |
with gr.Column():
|
95 |
+
output_image = gr.Image(label="Result Image", type="pil")
|
96 |
+
|
97 |
+
# Handle gallery selection
|
98 |
+
def on_select(evt: gr.SelectData):
|
99 |
+
if 0 <= evt.index < len(images):
|
100 |
+
return images[evt.index][0]
|
101 |
+
return None
|
102 |
+
|
103 |
+
gallery.select(
|
104 |
+
fn=on_select,
|
105 |
+
outputs=input_image
|
106 |
)
|
107 |
+
|
108 |
+
# Process image when input changes
|
109 |
+
input_image.change(
|
110 |
fn=process_image,
|
111 |
+
inputs=input_image,
|
112 |
+
outputs=output_image
|
113 |
)
|
114 |
+
|
115 |
+
# Clear button functionality
|
116 |
clear_button.click(
|
117 |
+
fn=lambda: None,
|
118 |
+
outputs=input_image
|
|
|
119 |
)
|
120 |
|
121 |
demo.launch()
|
122 |
|
|
|
123 |
if __name__ == "__main__":
|
124 |
+
app()
|