Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -5,7 +5,6 @@ import torch
|
|
5 |
import gradio as gr
|
6 |
import spaces
|
7 |
|
8 |
-
from glob import glob
|
9 |
from typing import Tuple
|
10 |
|
11 |
from PIL import Image
|
@@ -15,7 +14,7 @@ import requests
|
|
15 |
from io import BytesIO
|
16 |
import zipfile
|
17 |
|
18 |
-
# Fix the HF space permission error
|
19 |
os.environ["HF_MODULES_CACHE"] = os.path.join("/tmp/hf_cache", "modules")
|
20 |
|
21 |
import transformers
|
@@ -26,7 +25,6 @@ torch.jit.script = lambda f: f
|
|
26 |
|
27 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
28 |
|
29 |
-
### image_proc.py
|
30 |
def refine_foreground(image, mask, r=90):
|
31 |
if mask.size != image.size:
|
32 |
mask = mask.resize(image.size)
|
@@ -66,58 +64,26 @@ class ImagePreprocessor():
|
|
66 |
def proc(self, image: Image.Image) -> torch.Tensor:
|
67 |
return self.transform_image(image)
|
68 |
|
69 |
-
|
70 |
-
|
71 |
-
'General-HR': 'BiRefNet_HR',
|
72 |
-
'Matting-HR': 'BiRefNet_HR-matting',
|
73 |
-
'Matting': 'BiRefNet-matting',
|
74 |
-
'Portrait': 'BiRefNet-portrait',
|
75 |
-
'General-reso_512': 'BiRefNet_512x512',
|
76 |
-
'General-Lite': 'BiRefNet_lite',
|
77 |
-
'General-Lite-2K': 'BiRefNet_lite-2K',
|
78 |
-
'Anime-Lite': 'BiRefNet_lite-Anime',
|
79 |
-
'DIS': 'BiRefNet-DIS5K',
|
80 |
-
'HRSOD': 'BiRefNet-HRSOD',
|
81 |
-
'COD': 'BiRefNet-COD',
|
82 |
-
'DIS-TR_TEs': 'BiRefNet-DIS5K-TR_TEs',
|
83 |
-
'General-legacy': 'BiRefNet-legacy',
|
84 |
-
'General-dynamic': 'BiRefNet_dynamic',
|
85 |
-
}
|
86 |
-
|
87 |
birefnet = transformers.AutoModelForImageSegmentation.from_pretrained(
|
88 |
-
'/'.join(('zhengpeng7',
|
89 |
-
trust_remote_code=True
|
90 |
)
|
91 |
birefnet.to(device)
|
92 |
birefnet.eval(); birefnet.half()
|
93 |
|
94 |
@spaces.GPU
|
95 |
-
def predict(images, resolution
|
96 |
assert images is not None, 'AssertionError: images cannot be None.'
|
97 |
|
98 |
-
|
99 |
-
_weights_file = '/'.join(('zhengpeng7', usage_to_weights_file[weights_file] if weights_file is not None else usage_to_weights_file['General']))
|
100 |
print('Using weights: {}.'.format(_weights_file))
|
101 |
-
birefnet = transformers.AutoModelForImageSegmentation.from_pretrained(_weights_file, trust_remote_code=True)
|
102 |
-
birefnet.to(device)
|
103 |
-
birefnet.eval(); birefnet.half()
|
104 |
|
105 |
try:
|
106 |
resolution = [int(int(reso)//32*32) for reso in resolution.strip().split('x')]
|
107 |
except:
|
108 |
-
|
109 |
-
|
110 |
-
elif weights_file in ['General-Lite-2K']:
|
111 |
-
resolution = (2560, 1440)
|
112 |
-
elif weights_file in ['General-reso_512']:
|
113 |
-
resolution = (512, 512)
|
114 |
-
else:
|
115 |
-
if weights_file in ['General-dynamic']:
|
116 |
-
resolution = None
|
117 |
-
print('Using the original size (div by 32) for inference.')
|
118 |
-
else:
|
119 |
-
resolution = (1024, 1024)
|
120 |
-
print('Invalid resolution input. Automatically changed to 1024x1024 / 2048x2048 / 2560x1440.')
|
121 |
|
122 |
if isinstance(images, list):
|
123 |
save_paths = []
|
@@ -143,8 +109,7 @@ def predict(images, resolution, weights_file):
|
|
143 |
image = image_ori.convert('RGB')
|
144 |
if resolution is None:
|
145 |
resolution_div_by_32 = [int(int(reso)//32*32) for reso in image.size]
|
146 |
-
|
147 |
-
resolution = resolution_div_by_32
|
148 |
image_preprocessor = ImagePreprocessor(resolution=tuple(resolution))
|
149 |
image_proc = image_preprocessor.proc(image).unsqueeze(0)
|
150 |
|
@@ -182,7 +147,6 @@ tab_image = gr.Interface(
|
|
182 |
inputs=[
|
183 |
gr.Image(label='Upload an image'),
|
184 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
185 |
-
gr.Radio(list(usage_to_weights_file.keys()), value='General', label="Weights", info="Choose the weights you want.")
|
186 |
],
|
187 |
outputs=gr.ImageSlider(label="BiRefNet's prediction", type="pil", format='png'),
|
188 |
api_name="image",
|
@@ -194,21 +158,21 @@ tab_text = gr.Interface(
|
|
194 |
inputs=[
|
195 |
gr.Textbox(label="Paste an image URL"),
|
196 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
197 |
-
gr.Radio(list(usage_to_weights_file.keys()), value='General', label="Weights", info="Choose the weights you want.")
|
198 |
],
|
199 |
outputs=gr.ImageSlider(label="BiRefNet's prediction", type="pil", format='png'),
|
200 |
api_name="URL",
|
201 |
-
|
|
|
202 |
|
203 |
tab_batch = gr.Interface(
|
204 |
fn=predict,
|
205 |
inputs=[
|
206 |
gr.File(label="Upload multiple images", type="filepath", file_count="multiple"),
|
207 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
208 |
-
gr.Radio(list(usage_to_weights_file.keys()), value='General', label="Weights", info="Choose the weights you want.")
|
209 |
],
|
210 |
outputs=[gr.Gallery(label="BiRefNet's predictions"), gr.File(label="Download masked images.")],
|
211 |
api_name="batch",
|
|
|
212 |
)
|
213 |
|
214 |
demo = gr.TabbedInterface(
|
|
|
5 |
import gradio as gr
|
6 |
import spaces
|
7 |
|
|
|
8 |
from typing import Tuple
|
9 |
|
10 |
from PIL import Image
|
|
|
14 |
from io import BytesIO
|
15 |
import zipfile
|
16 |
|
17 |
+
# Fix the HF space permission error
|
18 |
os.environ["HF_MODULES_CACHE"] = os.path.join("/tmp/hf_cache", "modules")
|
19 |
|
20 |
import transformers
|
|
|
25 |
|
26 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
27 |
|
|
|
28 |
def refine_foreground(image, mask, r=90):
|
29 |
if mask.size != image.size:
|
30 |
mask = mask.resize(image.size)
|
|
|
64 |
def proc(self, image: Image.Image) -> torch.Tensor:
|
65 |
return self.transform_image(image)
|
66 |
|
67 |
+
# Fixed weights
|
68 |
+
weights_file = 'BiRefNet'
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
69 |
birefnet = transformers.AutoModelForImageSegmentation.from_pretrained(
|
70 |
+
'/'.join(('zhengpeng7', weights_file)), trust_remote_code=True
|
|
|
71 |
)
|
72 |
birefnet.to(device)
|
73 |
birefnet.eval(); birefnet.half()
|
74 |
|
75 |
@spaces.GPU
|
76 |
+
def predict(images, resolution):
|
77 |
assert images is not None, 'AssertionError: images cannot be None.'
|
78 |
|
79 |
+
_weights_file = '/'.join(('zhengpeng7', weights_file))
|
|
|
80 |
print('Using weights: {}.'.format(_weights_file))
|
|
|
|
|
|
|
81 |
|
82 |
try:
|
83 |
resolution = [int(int(reso)//32*32) for reso in resolution.strip().split('x')]
|
84 |
except:
|
85 |
+
resolution = (1024, 1024)
|
86 |
+
print('Invalid resolution input. Automatically changed to 1024x1024.')
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
87 |
|
88 |
if isinstance(images, list):
|
89 |
save_paths = []
|
|
|
109 |
image = image_ori.convert('RGB')
|
110 |
if resolution is None:
|
111 |
resolution_div_by_32 = [int(int(reso)//32*32) for reso in image.size]
|
112 |
+
resolution = resolution_div_by_32
|
|
|
113 |
image_preprocessor = ImagePreprocessor(resolution=tuple(resolution))
|
114 |
image_proc = image_preprocessor.proc(image).unsqueeze(0)
|
115 |
|
|
|
147 |
inputs=[
|
148 |
gr.Image(label='Upload an image'),
|
149 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
|
|
150 |
],
|
151 |
outputs=gr.ImageSlider(label="BiRefNet's prediction", type="pil", format='png'),
|
152 |
api_name="image",
|
|
|
158 |
inputs=[
|
159 |
gr.Textbox(label="Paste an image URL"),
|
160 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
|
|
161 |
],
|
162 |
outputs=gr.ImageSlider(label="BiRefNet's prediction", type="pil", format='png'),
|
163 |
api_name="URL",
|
164 |
+
description=descriptions + '\nTab-URL is partially modified from https://huggingface.co/spaces/not-lain/background-removal, thanks to this great work!',
|
165 |
+
)
|
166 |
|
167 |
tab_batch = gr.Interface(
|
168 |
fn=predict,
|
169 |
inputs=[
|
170 |
gr.File(label="Upload multiple images", type="filepath", file_count="multiple"),
|
171 |
gr.Textbox(lines=1, placeholder="Type the resolution (`WxH`) you want, e.g., `1024x1024`.", label="Resolution"),
|
|
|
172 |
],
|
173 |
outputs=[gr.Gallery(label="BiRefNet's predictions"), gr.File(label="Download masked images.")],
|
174 |
api_name="batch",
|
175 |
+
description=descriptions + '\nTab-batch is partially modified from https://huggingface.co/spaces/NegiTurkey/Multi_Birefnetfor_Background_Removal, thanks to this great work!',
|
176 |
)
|
177 |
|
178 |
demo = gr.TabbedInterface(
|