Spaces:
Running
on
L40S
Running
on
L40S
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,105 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
import shutil
|
3 |
+
import tempfile
|
4 |
+
import gradio as gr
|
5 |
+
from PIL import Image
|
6 |
+
from rembg import remove
|
7 |
+
import subprocess
|
8 |
+
from glob import glob
|
9 |
+
|
10 |
+
def remove_background(input_url):
|
11 |
+
# Create a temporary folder for downloaded and processed images
|
12 |
+
temp_dir = tempfile.mkdtemp()
|
13 |
+
|
14 |
+
# Download the image from the URL
|
15 |
+
image_path = os.path.join(temp_dir, 'input_image.png')
|
16 |
+
try:
|
17 |
+
image = Image.open(requests.get(input_url, stream=True).raw)
|
18 |
+
image.save(image_path)
|
19 |
+
except Exception as e:
|
20 |
+
shutil.rmtree(temp_dir)
|
21 |
+
return f"Error downloading or saving the image: {str(e)}"
|
22 |
+
|
23 |
+
# Run background removal
|
24 |
+
try:
|
25 |
+
removed_bg_path = os.path.join(temp_dir, 'output_image_rmbg.png')
|
26 |
+
img = Image.open(image_path)
|
27 |
+
result = remove(img)
|
28 |
+
result.save(removed_bg_path)
|
29 |
+
except Exception as e:
|
30 |
+
shutil.rmtree(temp_dir)
|
31 |
+
return f"Error removing background: {str(e)}"
|
32 |
+
|
33 |
+
return removed_bg_path, temp_dir
|
34 |
+
|
35 |
+
def run_inference(temp_dir):
|
36 |
+
# Define the inference configuration
|
37 |
+
inference_config = "configs/inference-768-6view.yaml"
|
38 |
+
pretrained_model = "pengHTYX/PSHuman_Unclip_768_6views"
|
39 |
+
crop_size = 740
|
40 |
+
seed = 600
|
41 |
+
num_views = 7
|
42 |
+
save_mode = "rgb"
|
43 |
+
|
44 |
+
try:
|
45 |
+
# Run the inference command
|
46 |
+
subprocess.run(
|
47 |
+
[
|
48 |
+
"python", "inference.py",
|
49 |
+
"--config", inference_config,
|
50 |
+
f"pretrained_model_name_or_path={pretrained_model}",
|
51 |
+
f"validation_dataset.crop_size={crop_size}",
|
52 |
+
f"with_smpl=false",
|
53 |
+
f"validation_dataset.root_dir={temp_dir}",
|
54 |
+
f"seed={seed}",
|
55 |
+
f"num_views={num_views}",
|
56 |
+
f"save_mode={save_mode}"
|
57 |
+
],
|
58 |
+
check=True
|
59 |
+
)
|
60 |
+
|
61 |
+
# Collect the output images
|
62 |
+
output_images = glob(os.path.join(temp_dir, "*.png"))
|
63 |
+
return output_images
|
64 |
+
except subprocess.CalledProcessError as e:
|
65 |
+
return f"Error during inference: {str(e)}"
|
66 |
+
|
67 |
+
def process_image(input_url):
|
68 |
+
# Remove background
|
69 |
+
removed_bg_path, temp_dir = remove_background(input_url)
|
70 |
+
|
71 |
+
if isinstance(removed_bg_path, str) and removed_bg_path.startswith("Error"):
|
72 |
+
return removed_bg_path
|
73 |
+
|
74 |
+
# Run inference
|
75 |
+
output_images = run_inference(temp_dir)
|
76 |
+
|
77 |
+
if isinstance(output_images, str) and output_images.startswith("Error"):
|
78 |
+
shutil.rmtree(temp_dir)
|
79 |
+
return output_images
|
80 |
+
|
81 |
+
# Prepare outputs for display
|
82 |
+
results = []
|
83 |
+
for img_path in output_images:
|
84 |
+
results.append((img_path, img_path))
|
85 |
+
|
86 |
+
shutil.rmtree(temp_dir) # Cleanup temporary folder
|
87 |
+
return results
|
88 |
+
|
89 |
+
def gradio_interface():
|
90 |
+
with gr.Blocks() as app:
|
91 |
+
gr.Markdown("# Background Removal and Inference Pipeline")
|
92 |
+
|
93 |
+
with gr.Row():
|
94 |
+
input_url = gr.Textbox(label="Image URL", placeholder="Enter the URL of the image")
|
95 |
+
submit_button = gr.Button("Process")
|
96 |
+
|
97 |
+
output_gallery = gr.Gallery(label="Output Images").style(grid=[2], height="300px")
|
98 |
+
|
99 |
+
submit_button.click(process_image, inputs=[input_url], outputs=[output_gallery])
|
100 |
+
|
101 |
+
return app
|
102 |
+
|
103 |
+
# Launch the Gradio app
|
104 |
+
app = gradio_interface()
|
105 |
+
app.launch()
|