Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -1,74 +1,18 @@
|
|
1 |
-
import
|
2 |
-
from loadimg import load_img
|
3 |
-
from transformers import AutoModelForImageSegmentation
|
4 |
-
import torch
|
5 |
-
from torchvision import transforms
|
6 |
-
from typing import Union, Tuple
|
7 |
from PIL import Image
|
|
|
8 |
|
9 |
-
|
10 |
-
|
11 |
-
|
12 |
-
|
13 |
-
birefnet = AutoModelForImageSegmentation.from_pretrained(
|
14 |
-
"ZhengPeng7/BiRefNet", trust_remote_code=True
|
15 |
-
)
|
16 |
-
birefnet.to(device)
|
17 |
-
|
18 |
-
transform_image = transforms.Compose([
|
19 |
-
transforms.Resize((1024, 1024)),
|
20 |
-
transforms.ToTensor(),
|
21 |
-
transforms.Normalize([0.485, 0.456, 0.406],
|
22 |
-
[0.229, 0.224, 0.225]),
|
23 |
-
])
|
24 |
-
|
25 |
-
def process(image: Image.Image) -> Image.Image:
|
26 |
-
image_size = image.size
|
27 |
-
input_images = transform_image(image).unsqueeze(0).to(device)
|
28 |
-
with torch.no_grad():
|
29 |
-
preds = birefnet(input_images)[-1].sigmoid().cpu()
|
30 |
-
pred = preds[0].squeeze()
|
31 |
-
pred_pil = transforms.ToPILImage()(pred)
|
32 |
-
mask = pred_pil.resize(image_size)
|
33 |
-
image.putalpha(mask)
|
34 |
-
return image
|
35 |
-
|
36 |
-
def fn(image: Union[Image.Image, str]) -> Tuple[Image.Image, Image.Image]:
|
37 |
-
im = load_img(image, output_type="pil").convert("RGB")
|
38 |
-
origin = im.copy()
|
39 |
-
processed_image = process(im)
|
40 |
-
return processed_image, origin
|
41 |
-
|
42 |
-
def process_file(f: str) -> str:
|
43 |
-
name_path = f.rsplit(".", 1)[0] + ".png"
|
44 |
-
im = load_img(f, output_type="pil").convert("RGB")
|
45 |
-
transparent = process(im)
|
46 |
-
transparent.save(name_path)
|
47 |
-
return name_path
|
48 |
-
|
49 |
-
slider1 = gr.ImageSlider(label="Processed Image", type="pil", format="png")
|
50 |
-
slider2 = gr.ImageSlider(label="Processed Image from URL", type="pil", format="png")
|
51 |
-
image_upload = gr.Image(label="Upload an image")
|
52 |
-
image_file_upload = gr.Image(label="Upload an image", type="filepath")
|
53 |
-
url_input = gr.Textbox(label="Paste an image URL")
|
54 |
-
output_file = gr.File(label="Output PNG File")
|
55 |
-
|
56 |
-
chameleon = load_img("butterfly.jpg", output_type="pil")
|
57 |
-
url_example = "https://hips.hearstapps.com/hmg-prod/images/gettyimages-1229892983-square.jpg"
|
58 |
|
59 |
-
|
60 |
-
|
61 |
-
|
62 |
-
examples=[url_example], api_name="text")
|
63 |
-
tab3 = gr.Interface(process_file, inputs=image_file_upload,
|
64 |
-
outputs=output_file, examples=["butterfly.jpg"],
|
65 |
-
api_name="png")
|
66 |
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
|
71 |
-
)
|
72 |
|
73 |
-
|
74 |
-
demo.launch(show_error=True, server_name="0.0.0.0", server_port=7860)
|
|
|
1 |
+
from transformers import AutoProcessor, AutoModelForImageSegmentation
|
|
|
|
|
|
|
|
|
|
|
2 |
from PIL import Image
|
3 |
+
import torch
|
4 |
|
5 |
+
# Carica modello e processor
|
6 |
+
processor = AutoProcessor.from_pretrained("BritishWerewolf/U-2-Netp")
|
7 |
+
model = AutoModelForImageSegmentation.from_pretrained("BritishWerewolf/U-2-Netp")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
|
9 |
+
# Prepara l’immagine
|
10 |
+
img = Image.open("input.jpg").convert("RGB")
|
11 |
+
inputs = processor(images=img, return_tensors="pt")
|
|
|
|
|
|
|
|
|
12 |
|
13 |
+
# Inferenzia maschera
|
14 |
+
with torch.no_grad():
|
15 |
+
outputs = model(**inputs)
|
16 |
+
mask = outputs.logits.argmax(dim=1)[0].cpu().numpy()
|
|
|
17 |
|
18 |
+
# Applica maschera all’immagine originale...
|
|