Spaces:
Paused
Paused
Added genre selection + inpainting via notebook (#1)
Browse files- Added genre selection + inpainting via notebook (d56e77f57a166ee70088a054b4df03ec76461393)
Co-authored-by: Carlos Marí Noguera <[email protected]>
- app.py +38 -4
- diffusion.py +25 -3
- inference.py +22 -16
- inpainting.ipynb +160 -0
app.py
CHANGED
@@ -1,22 +1,56 @@
|
|
1 |
import streamlit as st
|
2 |
from PIL import Image
|
3 |
-
from inference import inference
|
|
|
4 |
import io
|
5 |
|
6 |
def main():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
7 |
st.title("Image Display App")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
|
9 |
# Button to trigger image generation
|
10 |
if st.button('Generate Image'):
|
11 |
-
|
12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
13 |
|
14 |
# Convert Pillow image to bytes for display in Streamlit
|
15 |
img_buffer = io.BytesIO()
|
|
|
16 |
image.save(img_buffer, format="PNG")
|
17 |
img_buffer.seek(0)
|
18 |
|
19 |
-
# Display the image
|
20 |
st.image(img_buffer, caption='Generated Image', use_column_width=True)
|
21 |
|
22 |
if __name__ == "__main__":
|
|
|
1 |
import streamlit as st
|
2 |
from PIL import Image
|
3 |
+
from inference import inference
|
4 |
+
import torch
|
5 |
import io
|
6 |
|
7 |
def main():
|
8 |
+
|
9 |
+
genres_dict = {
|
10 |
+
'Action': 1,
|
11 |
+
'Adventure': 2,
|
12 |
+
'Animation': 3,
|
13 |
+
'Comedy': 4,
|
14 |
+
'Drama': 5,
|
15 |
+
'Family': 6,
|
16 |
+
'Horror': 7,
|
17 |
+
'Music': 8,
|
18 |
+
'Romance': 9,
|
19 |
+
'Science Fiction': 10,
|
20 |
+
'Western': 11,
|
21 |
+
'Fantasy': 12,
|
22 |
+
'Thriller': 13
|
23 |
+
}
|
24 |
+
|
25 |
st.title("Image Display App")
|
26 |
+
cond = torch.tensor([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0])
|
27 |
+
|
28 |
+
# Add a sidebar for genre selection
|
29 |
+
#genre = st.sidebar.selectbox("Select Genre", list(genres_dict.keys()))
|
30 |
+
|
31 |
+
|
32 |
+
selected_genres = st.sidebar.multiselect('Select Genres', list(genres_dict.keys()))
|
33 |
+
|
34 |
+
|
35 |
|
36 |
# Button to trigger image generation
|
37 |
if st.button('Generate Image'):
|
38 |
+
for genre in selected_genres:
|
39 |
+
code = genres_dict[genre]
|
40 |
+
cond[code-1] = code
|
41 |
+
# Display loading sign while generating image
|
42 |
+
with st.spinner('Generating Image...'):
|
43 |
+
# Call the function from inference.py with selected genre
|
44 |
+
image = inference(cond)
|
45 |
+
#image = inference(genre)
|
46 |
|
47 |
# Convert Pillow image to bytes for display in Streamlit
|
48 |
img_buffer = io.BytesIO()
|
49 |
+
#"""0,0,0,0,0,0,0,1, 2, 7, 4, 0, 0, 0"""
|
50 |
image.save(img_buffer, format="PNG")
|
51 |
img_buffer.seek(0)
|
52 |
|
53 |
+
# Display the generated image
|
54 |
st.image(img_buffer, caption='Generated Image', use_column_width=True)
|
55 |
|
56 |
if __name__ == "__main__":
|
diffusion.py
CHANGED
@@ -160,26 +160,48 @@ class GaussianDiffusion:
|
|
160 |
|
161 |
return x_t_minus_1
|
162 |
|
163 |
-
def sample(self, num_samples, show_progress=True):
|
164 |
"""
|
165 |
Sample from the model
|
166 |
"""
|
167 |
-
cond = None
|
168 |
-
if
|
169 |
# cond is arange()
|
170 |
assert num_samples <= self.model.num_classes, "num_samples must be less than or equal to the number of classes"
|
171 |
cond = torch.arange(self.model.num_classes)[:num_samples].to(self.device)
|
172 |
cond = rearrange(cond, 'i -> i ()')
|
173 |
|
|
|
|
|
174 |
self.model.eval()
|
175 |
image_versions = []
|
176 |
with torch.no_grad():
|
177 |
x = torch.randn(num_samples, self.channels, *self.image_size).to(self.device)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
178 |
it = reversed(range(1, self.noise_steps))
|
179 |
if show_progress:
|
180 |
it = tqdm(it)
|
181 |
for t in it:
|
182 |
image_versions.append(self.denormalize_image(torch.clip(x, -1, 1)).clone().squeeze(0))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
183 |
x = self.sample_step(x, t, cond)
|
184 |
self.model.train()
|
185 |
x = torch.clip(x, -1.0, 1.0)
|
|
|
160 |
|
161 |
return x_t_minus_1
|
162 |
|
163 |
+
def sample(self, num_samples, show_progress=True, cond=None, x0=None):
|
164 |
"""
|
165 |
Sample from the model
|
166 |
"""
|
167 |
+
#cond = None
|
168 |
+
if cond == None:
|
169 |
# cond is arange()
|
170 |
assert num_samples <= self.model.num_classes, "num_samples must be less than or equal to the number of classes"
|
171 |
cond = torch.arange(self.model.num_classes)[:num_samples].to(self.device)
|
172 |
cond = rearrange(cond, 'i -> i ()')
|
173 |
|
174 |
+
|
175 |
+
# Inpainting
|
176 |
self.model.eval()
|
177 |
image_versions = []
|
178 |
with torch.no_grad():
|
179 |
x = torch.randn(num_samples, self.channels, *self.image_size).to(self.device)
|
180 |
+
|
181 |
+
|
182 |
+
if x0 is not None:
|
183 |
+
x0 = x0.to(self.device)
|
184 |
+
mask = x0 != -1
|
185 |
+
x_noised = self.apply_noise(x0,self.noise_steps -1)[0].to(self.device)
|
186 |
+
new_x = x
|
187 |
+
new_x[mask] = x_noised[mask]
|
188 |
+
|
189 |
+
x = new_x
|
190 |
+
|
191 |
+
|
192 |
it = reversed(range(1, self.noise_steps))
|
193 |
if show_progress:
|
194 |
it = tqdm(it)
|
195 |
for t in it:
|
196 |
image_versions.append(self.denormalize_image(torch.clip(x, -1, 1)).clone().squeeze(0))
|
197 |
+
|
198 |
+
if x0 is not None and t > 80:
|
199 |
+
x_noised = self.apply_noise(x0,t)[0]
|
200 |
+
new_x = x
|
201 |
+
new_x[mask] = x_noised[mask]
|
202 |
+
|
203 |
+
x = new_x
|
204 |
+
|
205 |
x = self.sample_step(x, t, cond)
|
206 |
self.model.train()
|
207 |
x = torch.clip(x, -1.0, 1.0)
|
inference.py
CHANGED
@@ -13,12 +13,13 @@ from diffusion import GaussianDiffusion, DiffusionImageAPI
|
|
13 |
|
14 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
15 |
|
|
|
16 |
def inference1():
|
17 |
# new image from web page
|
18 |
image = requests.get("https://picsum.photos/120/80").content
|
19 |
return Image.open(io.BytesIO(image))
|
20 |
|
21 |
-
def inference():
|
22 |
model = Unet(
|
23 |
image_channels=3,
|
24 |
dropout=0.1,
|
@@ -37,26 +38,31 @@ def inference():
|
|
37 |
image_size=(192, 128),
|
38 |
)
|
39 |
|
|
|
|
|
|
|
|
|
|
|
40 |
model.to(device)
|
41 |
diffusion.to(device)
|
42 |
|
43 |
imageAPI = DiffusionImageAPI(diffusion)
|
44 |
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
58 |
-
|
59 |
-
return imageAPI.tensor_to_image(
|
60 |
|
61 |
if __name__ == "__main__":
|
62 |
inference().show()
|
|
|
13 |
|
14 |
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
15 |
|
16 |
+
|
17 |
def inference1():
|
18 |
# new image from web page
|
19 |
image = requests.get("https://picsum.photos/120/80").content
|
20 |
return Image.open(io.BytesIO(image))
|
21 |
|
22 |
+
def inference(cond, x0=None, gif=False):
|
23 |
model = Unet(
|
24 |
image_channels=3,
|
25 |
dropout=0.1,
|
|
|
38 |
image_size=(192, 128),
|
39 |
)
|
40 |
|
41 |
+
if x0 is not None:
|
42 |
+
x0 = diffusion.normalize_image(x0)
|
43 |
+
x0 = x0.permute(2, 0, 1)
|
44 |
+
x0 = x0.unsqueeze(0)
|
45 |
+
|
46 |
model.to(device)
|
47 |
diffusion.to(device)
|
48 |
|
49 |
imageAPI = DiffusionImageAPI(diffusion)
|
50 |
|
51 |
+
new_images, versions = diffusion.sample(1,cond=cond,x0=x0)
|
52 |
+
if gif:
|
53 |
+
images = []
|
54 |
+
for image in versions:
|
55 |
+
images.append(imageAPI.tensor_to_image(image.squeeze(0)))
|
56 |
+
|
57 |
+
print(len(images))
|
58 |
+
print(images[0])
|
59 |
+
# make gif out of pillow images
|
60 |
+
images[0].save('./gif_output/versions.gif',
|
61 |
+
save_all=True,
|
62 |
+
append_images=images[1:],
|
63 |
+
duration=100,
|
64 |
+
loop=0)
|
65 |
+
return imageAPI.tensor_to_image(new_images.squeeze(0))
|
66 |
|
67 |
if __name__ == "__main__":
|
68 |
inference().show()
|
inpainting.ipynb
ADDED
@@ -0,0 +1,160 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "code",
|
5 |
+
"execution_count": 29,
|
6 |
+
"metadata": {},
|
7 |
+
"outputs": [
|
8 |
+
{
|
9 |
+
"name": "stdout",
|
10 |
+
"output_type": "stream",
|
11 |
+
"text": [
|
12 |
+
"The autoreload extension is already loaded. To reload it, use:\n",
|
13 |
+
" %reload_ext autoreload\n"
|
14 |
+
]
|
15 |
+
}
|
16 |
+
],
|
17 |
+
"source": [
|
18 |
+
"%load_ext autoreload\n",
|
19 |
+
"%autoreload 2\n",
|
20 |
+
"from PIL import Image\n",
|
21 |
+
"import torch \n",
|
22 |
+
"from diffusion import GaussianDiffusion, DiffusionImageAPI\n",
|
23 |
+
"from unet import Unet\n",
|
24 |
+
"from inference import inference\n",
|
25 |
+
"import numpy as np"
|
26 |
+
]
|
27 |
+
},
|
28 |
+
{
|
29 |
+
"cell_type": "code",
|
30 |
+
"execution_count": 30,
|
31 |
+
"metadata": {},
|
32 |
+
"outputs": [
|
33 |
+
{
|
34 |
+
"data": {
|
35 |
+
"text/plain": [
|
36 |
+
"True"
|
37 |
+
]
|
38 |
+
},
|
39 |
+
"execution_count": 30,
|
40 |
+
"metadata": {},
|
41 |
+
"output_type": "execute_result"
|
42 |
+
}
|
43 |
+
],
|
44 |
+
"source": [
|
45 |
+
"torch.cuda.is_available()"
|
46 |
+
]
|
47 |
+
},
|
48 |
+
{
|
49 |
+
"cell_type": "code",
|
50 |
+
"execution_count": 31,
|
51 |
+
"metadata": {},
|
52 |
+
"outputs": [],
|
53 |
+
"source": [
|
54 |
+
"cond = torch.tensor([2, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]) \n",
|
55 |
+
"genres_dict = {\n",
|
56 |
+
" 'Action': 1,\n",
|
57 |
+
" 'Adventure': 2,\n",
|
58 |
+
" 'Animation': 3,\n",
|
59 |
+
" 'Comedy': 4,\n",
|
60 |
+
" 'Drama': 5,\n",
|
61 |
+
" 'Family': 6,\n",
|
62 |
+
" 'Horror': 7,\n",
|
63 |
+
" 'Music': 8,\n",
|
64 |
+
" 'Romance': 9,\n",
|
65 |
+
" 'Science Fiction': 10,\n",
|
66 |
+
" 'Western': 11,\n",
|
67 |
+
" 'Fantasy': 12,\n",
|
68 |
+
" 'Thriller': 13\n",
|
69 |
+
"}"
|
70 |
+
]
|
71 |
+
},
|
72 |
+
{
|
73 |
+
"cell_type": "code",
|
74 |
+
"execution_count": 45,
|
75 |
+
"metadata": {},
|
76 |
+
"outputs": [
|
77 |
+
{
|
78 |
+
"name": "stderr",
|
79 |
+
"output_type": "stream",
|
80 |
+
"text": [
|
81 |
+
"999it [01:18, 12.69it/s]\n"
|
82 |
+
]
|
83 |
+
}
|
84 |
+
],
|
85 |
+
"source": [
|
86 |
+
"pic = 'IndianaBovik'\n",
|
87 |
+
"image_np = np.array(Image.open(f\"InferenceTests/{pic}.png\").convert('RGB'))\n",
|
88 |
+
"\n",
|
89 |
+
"# Convert the NumPy array to a PyTorch tensor with explicitly specifying the data type\n",
|
90 |
+
"x0 = torch.tensor(image_np, dtype=torch.float32)\n",
|
91 |
+
"\n",
|
92 |
+
"\n",
|
93 |
+
"\n",
|
94 |
+
"image = inference(cond,x0)\n"
|
95 |
+
]
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"cell_type": "code",
|
99 |
+
"execution_count": 46,
|
100 |
+
"metadata": {},
|
101 |
+
"outputs": [
|
102 |
+
{
|
103 |
+
"data": {
|
104 |
+
"image/png": "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",
|
105 |
+
"text/plain": [
|
106 |
+
"<PIL.Image.Image image mode=RGB size=128x192>"
|
107 |
+
]
|
108 |
+
},
|
109 |
+
"execution_count": 46,
|
110 |
+
"metadata": {},
|
111 |
+
"output_type": "execute_result"
|
112 |
+
}
|
113 |
+
],
|
114 |
+
"source": [
|
115 |
+
"image"
|
116 |
+
]
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"cell_type": "code",
|
120 |
+
"execution_count": 47,
|
121 |
+
"metadata": {},
|
122 |
+
"outputs": [],
|
123 |
+
"source": [
|
124 |
+
"i = 0"
|
125 |
+
]
|
126 |
+
},
|
127 |
+
{
|
128 |
+
"cell_type": "code",
|
129 |
+
"execution_count": 48,
|
130 |
+
"metadata": {},
|
131 |
+
"outputs": [],
|
132 |
+
"source": [
|
133 |
+
"route = f'./InferenceOutputs/{pic}{i}.png'\n",
|
134 |
+
"i+=1\n",
|
135 |
+
"image.save(route)"
|
136 |
+
]
|
137 |
+
}
|
138 |
+
],
|
139 |
+
"metadata": {
|
140 |
+
"kernelspec": {
|
141 |
+
"display_name": "DIP_DEMO",
|
142 |
+
"language": "python",
|
143 |
+
"name": "python3"
|
144 |
+
},
|
145 |
+
"language_info": {
|
146 |
+
"codemirror_mode": {
|
147 |
+
"name": "ipython",
|
148 |
+
"version": 3
|
149 |
+
},
|
150 |
+
"file_extension": ".py",
|
151 |
+
"mimetype": "text/x-python",
|
152 |
+
"name": "python",
|
153 |
+
"nbconvert_exporter": "python",
|
154 |
+
"pygments_lexer": "ipython3",
|
155 |
+
"version": "3.10.13"
|
156 |
+
}
|
157 |
+
},
|
158 |
+
"nbformat": 4,
|
159 |
+
"nbformat_minor": 2
|
160 |
+
}
|