Spaces:
Sleeping
Sleeping
VenkateshRoshan
commited on
Commit
·
4aaf04f
1
Parent(s):
474e221
App updated.
Browse files- app.py +84 -0
- butterfly.png +0 -0
- dockerfile +0 -0
- requirements.txt +13 -0
app.py
ADDED
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
from transformers import pipeline, AutoModelForImageSegmentation
|
3 |
+
from gradio_imageslider import ImageSlider
|
4 |
+
import torch
|
5 |
+
from torchvision import transforms
|
6 |
+
import spaces
|
7 |
+
from PIL import Image
|
8 |
+
|
9 |
+
import numpy as np
|
10 |
+
import time
|
11 |
+
|
12 |
+
birefnet = AutoModelForImageSegmentation.from_pretrained(
|
13 |
+
"ZhengPeng7/BiRefNet", trust_remote_code=True
|
14 |
+
)
|
15 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
16 |
+
|
17 |
+
print("Using device:", device)
|
18 |
+
|
19 |
+
birefnet.to(device)
|
20 |
+
transform_image = transforms.Compose(
|
21 |
+
[
|
22 |
+
transforms.Resize((1024, 1024)),
|
23 |
+
transforms.ToTensor(),
|
24 |
+
transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225]),
|
25 |
+
]
|
26 |
+
)
|
27 |
+
|
28 |
+
# @spaces.GPU
|
29 |
+
# def PreProcess(image):
|
30 |
+
# size = image.size
|
31 |
+
# image = transform_image(image).unsqueeze(0).to(device)
|
32 |
+
|
33 |
+
# with torch.no_grad():
|
34 |
+
# preds = birefnet(image)[-1].sigmoid().cpu()
|
35 |
+
# pred = preds[0].squeeze()
|
36 |
+
# pred = transforms.ToPILImage()(pred)
|
37 |
+
# mask = pred.resize(size)
|
38 |
+
# # image.putalpha(mask)
|
39 |
+
# return image
|
40 |
+
|
41 |
+
@spaces.GPU
|
42 |
+
def PreProcess(image):
|
43 |
+
size = image.size # Save original size
|
44 |
+
image_tensor = transform_image(image).unsqueeze(0).to(device) # Transform the image into a tensor
|
45 |
+
|
46 |
+
with torch.no_grad():
|
47 |
+
preds = birefnet(image_tensor)[-1].sigmoid().cpu() # Get predictions
|
48 |
+
pred = preds[0].squeeze()
|
49 |
+
|
50 |
+
# Convert the prediction tensor to a PIL image
|
51 |
+
pred_pil = transforms.ToPILImage()(pred)
|
52 |
+
|
53 |
+
# Resize the mask to match the original image size
|
54 |
+
mask = pred_pil.resize(size)
|
55 |
+
|
56 |
+
# Convert the original image (passed as input) to a PIL image
|
57 |
+
image_pil = image.convert("RGBA") # Ensure the image has an alpha channel
|
58 |
+
|
59 |
+
# Apply the alpha mask to the image
|
60 |
+
image_pil.putalpha(mask)
|
61 |
+
|
62 |
+
return image_pil
|
63 |
+
|
64 |
+
def segment_image(image):
|
65 |
+
start = time.time()
|
66 |
+
image = Image.fromarray(image)
|
67 |
+
image = image.convert("RGB")
|
68 |
+
org = image.copy()
|
69 |
+
image = PreProcess(image)
|
70 |
+
time_taken = np.round((time.time() - start),2)
|
71 |
+
return (image, org), time_taken
|
72 |
+
|
73 |
+
slider = ImageSlider(label='birefnet', type="pil")
|
74 |
+
image = gr.Image(label="Upload an Image")
|
75 |
+
|
76 |
+
butterfly = Image.open("butterfly.png")
|
77 |
+
|
78 |
+
time_taken = gr.Textbox(label="Time taken", type="text")
|
79 |
+
|
80 |
+
demo = gr.Interface(
|
81 |
+
segment_image, inputs=image, outputs=[slider,time_taken], examples=[butterfly], api_name="BiRefNet")
|
82 |
+
|
83 |
+
if __name__ == '__main__' :
|
84 |
+
demo.launch()
|
butterfly.png
ADDED
![]() |
dockerfile
ADDED
File without changes
|
requirements.txt
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
torch
|
2 |
+
accelerate
|
3 |
+
opencv-python
|
4 |
+
spaces
|
5 |
+
torchvision
|
6 |
+
pillow
|
7 |
+
numpy
|
8 |
+
huggingface-hub
|
9 |
+
gradio
|
10 |
+
gradio-imageslider
|
11 |
+
transformers
|
12 |
+
timm
|
13 |
+
kornia
|