Upload 6 files
Browse filesDatasets, network summaries and notebooks to generate them
- .gitattributes +1 -0
- HIGH.npz +3 -0
- LOW.npz +3 -0
- MEDIUM.npz +3 -0
- characterize_networks.ipynb +3 -0
- data_gen.ipynb +226 -0
- summaries.csv +0 -0
.gitattributes
CHANGED
@@ -53,3 +53,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
53 |
*.jpg filter=lfs diff=lfs merge=lfs -text
|
54 |
*.jpeg filter=lfs diff=lfs merge=lfs -text
|
55 |
*.webp filter=lfs diff=lfs merge=lfs -text
|
|
|
|
53 |
*.jpg filter=lfs diff=lfs merge=lfs -text
|
54 |
*.jpeg filter=lfs diff=lfs merge=lfs -text
|
55 |
*.webp filter=lfs diff=lfs merge=lfs -text
|
56 |
+
characterize_networks.ipynb filter=lfs diff=lfs merge=lfs -text
|
HIGH.npz
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c28d51927e829601f2c9bb420e5c4789c26b63d0100bd1ce8134118621093c1b
|
3 |
+
size 44042186
|
LOW.npz
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2b3f4af0734079675fcd49b63e2ef6bdfea903ba11e32240e6095f640fb15ae9
|
3 |
+
size 44042186
|
MEDIUM.npz
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:82f97a0bcce9e706c5f6d3e2914e5fb3b8f924fc3a284803bc5297ff87a6360c
|
3 |
+
size 44042186
|
characterize_networks.ipynb
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e4be8e5e258da2be04a31421e0eb57e9af944f729a4a981fbea51b1bc75a3adc
|
3 |
+
size 11253056
|
data_gen.ipynb
ADDED
@@ -0,0 +1,226 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cells": [
|
3 |
+
{
|
4 |
+
"cell_type": "markdown",
|
5 |
+
"id": "jewish-bhutan",
|
6 |
+
"metadata": {},
|
7 |
+
"source": [
|
8 |
+
"This notebook shows a simple implementation of supervised denoising with associated estimation of standard deviation and subsequent calibration of the errors via conformal prediction.\n",
|
9 |
+
"\n",
|
10 |
+
"The notebook is a work in progress, and more comments will be needed."
|
11 |
+
]
|
12 |
+
},
|
13 |
+
{
|
14 |
+
"cell_type": "code",
|
15 |
+
"execution_count": 1,
|
16 |
+
"id": "vulnerable-enforcement",
|
17 |
+
"metadata": {},
|
18 |
+
"outputs": [],
|
19 |
+
"source": [
|
20 |
+
"import numpy as np\n",
|
21 |
+
"import random\n",
|
22 |
+
"import torch\n",
|
23 |
+
"from torch import nn\n",
|
24 |
+
"from dlsia.core.networks import tunet\n",
|
25 |
+
"import matplotlib.pyplot as plt\n",
|
26 |
+
"import einops\n",
|
27 |
+
"\n",
|
28 |
+
"import torch\n",
|
29 |
+
"from torch.utils.data import TensorDataset, DataLoader\n",
|
30 |
+
"\n",
|
31 |
+
"from dlsia.core.networks import smsnet, construct_sms_ensemble, three_way, baggins\n",
|
32 |
+
"from dlsia.core.train_scripts import train_regression\n",
|
33 |
+
"from dlsia.core.helpers import get_device\n",
|
34 |
+
"\n",
|
35 |
+
"\n",
|
36 |
+
"SEED = 2003\n",
|
37 |
+
"random.seed(SEED)\n",
|
38 |
+
"np.random.seed(SEED)\n",
|
39 |
+
"torch.manual_seed(0)\n",
|
40 |
+
"torch.cuda.manual_seed(0)\n",
|
41 |
+
"torch.backends.cudnn.deterministic = True\n",
|
42 |
+
"torch.backends.cudnn.benchmark = False\n",
|
43 |
+
"\n",
|
44 |
+
"\n",
|
45 |
+
"DATASET = \"LOW\"\n",
|
46 |
+
"\n",
|
47 |
+
"high_noise = 10\n",
|
48 |
+
"medium_noise = 1.0\n",
|
49 |
+
"low_noise = 0.1\n",
|
50 |
+
"\n",
|
51 |
+
"params = {\"HIGH\":high_noise, \"MEDIUM\":medium_noise, \"LOW\":low_noise}\n",
|
52 |
+
"noise = params[DATASET]\n"
|
53 |
+
]
|
54 |
+
},
|
55 |
+
{
|
56 |
+
"cell_type": "markdown",
|
57 |
+
"id": "robust-involvement",
|
58 |
+
"metadata": {},
|
59 |
+
"source": [
|
60 |
+
"Some basic tools for generating random 2D data"
|
61 |
+
]
|
62 |
+
},
|
63 |
+
{
|
64 |
+
"cell_type": "code",
|
65 |
+
"execution_count": 2,
|
66 |
+
"id": "committed-fireplace",
|
67 |
+
"metadata": {},
|
68 |
+
"outputs": [],
|
69 |
+
"source": [
|
70 |
+
"def generate_random_surface(grid_size, order):\n",
|
71 |
+
" # Initialize the frequency domain representation with zeros\n",
|
72 |
+
" freq_domain = np.zeros(grid_size, dtype=complex)\n",
|
73 |
+
" coefs = np.random.normal(0,1, grid_size) + 1j*np.random.normal(0,1,grid_size)\n",
|
74 |
+
"\n",
|
75 |
+
" # limit coeffiecients\n",
|
76 |
+
" for i in range(grid_size[0]):\n",
|
77 |
+
" for j in range(grid_size[1]):\n",
|
78 |
+
" if i*i+j*j > order*order:\n",
|
79 |
+
" coefs[i,j] = 0.0\n",
|
80 |
+
" \n",
|
81 |
+
" for i in range(grid_size[0]):\n",
|
82 |
+
" for j in range(grid_size[1]): \n",
|
83 |
+
" coefs[-i, -j] = np.conj(coefs[i, j]) \n",
|
84 |
+
" \n",
|
85 |
+
" surface = np.fft.ifft2(coefs, s=grid_size).real\n",
|
86 |
+
" surface = (surface-np.mean(surface))/np.std(surface)\n",
|
87 |
+
" return surface\n",
|
88 |
+
"\n",
|
89 |
+
"def modifier(img, tau=0.0, kappa=2.0):\n",
|
90 |
+
" tmp = kappa*(img-tau)\n",
|
91 |
+
" tmp = 1/(1+np.exp(-tmp))\n",
|
92 |
+
" #tmp = (tmp-np.mean(tmp))/(np.std(tmp)+1e-3)\n",
|
93 |
+
" return tmp\n",
|
94 |
+
" \n",
|
95 |
+
" \n",
|
96 |
+
"\n",
|
97 |
+
"\n",
|
98 |
+
"def build_data_set(N_images, grid_size, order, noise_level):\n",
|
99 |
+
" gt = []\n",
|
100 |
+
" obs = [] \n",
|
101 |
+
" for i in range(N_images):\n",
|
102 |
+
" noise1 = np.random.normal(0,noise_level, grid_size)\n",
|
103 |
+
" noise2 = np.random.normal(0,noise_level, grid_size)\n",
|
104 |
+
" \n",
|
105 |
+
" surf = generate_random_surface(grid_size, order)\n",
|
106 |
+
" surf = surf*modifier(surf)\n",
|
107 |
+
" noise = noise1 + np.abs(surf)*np.sqrt(noise2*noise2)*modifier(surf)\n",
|
108 |
+
" gt.append(surf) \n",
|
109 |
+
" noise = noise+surf\n",
|
110 |
+
" obs.append(noise)\n",
|
111 |
+
" \n",
|
112 |
+
" return np.array(gt), np.array(obs)\n",
|
113 |
+
"\n",
|
114 |
+
"def compute_noise_level(gt,obs):\n",
|
115 |
+
" msignal = np.mean(gt)\n",
|
116 |
+
" stdnoise = np.std(gt-obs)\n",
|
117 |
+
" return 20*np.log10(msignal/stdnoise)\n",
|
118 |
+
" "
|
119 |
+
]
|
120 |
+
},
|
121 |
+
{
|
122 |
+
"cell_type": "markdown",
|
123 |
+
"id": "based-consequence",
|
124 |
+
"metadata": {},
|
125 |
+
"source": [
|
126 |
+
"Define the noise level, number of data points and build data loaders"
|
127 |
+
]
|
128 |
+
},
|
129 |
+
{
|
130 |
+
"cell_type": "code",
|
131 |
+
"execution_count": 3,
|
132 |
+
"id": "wired-associate",
|
133 |
+
"metadata": {
|
134 |
+
"scrolled": false
|
135 |
+
},
|
136 |
+
"outputs": [
|
137 |
+
{
|
138 |
+
"data": {
|
139 |
+
"image/png": "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",
|
140 |
+
"text/plain": [
|
141 |
+
"<Figure size 1000x400 with 4 Axes>"
|
142 |
+
]
|
143 |
+
},
|
144 |
+
"metadata": {},
|
145 |
+
"output_type": "display_data"
|
146 |
+
}
|
147 |
+
],
|
148 |
+
"source": [
|
149 |
+
"level_sigma = [10.0, 1.0, 0.1] \n",
|
150 |
+
"GRID=64\n",
|
151 |
+
"M = 64\n",
|
152 |
+
"M_train = 512\n",
|
153 |
+
"M_cal = 32\n",
|
154 |
+
"\n",
|
155 |
+
"gt_train,obs_train = build_data_set(M_train, (GRID,GRID), 4, noise)\n",
|
156 |
+
"gt_test,obs_test = build_data_set(M, (GRID,GRID), 4, noise)\n",
|
157 |
+
"gt_cal,obs_cal = build_data_set(M_cal, (GRID,GRID), 4, noise)\n",
|
158 |
+
"gt_val,obs_val = build_data_set(M, (GRID,GRID), 4, noise)\n",
|
159 |
+
"\n",
|
160 |
+
"nl = compute_noise_level(gt_train,obs_train).round(0)\n",
|
161 |
+
"\n",
|
162 |
+
"obs1 = obs_val[0]\n",
|
163 |
+
"gt1 = gt_val[0]\n",
|
164 |
+
"\n",
|
165 |
+
"fig, axs = plt.subplots(nrows=1, ncols=2, figsize=(10, 4))\n",
|
166 |
+
"im1 = axs[0].imshow(obs1)#, vmin=0,vmax=1.0)\n",
|
167 |
+
"im2 = axs[1].imshow(gt1)#, vmin=0,vmax=1.0)\n",
|
168 |
+
"\n",
|
169 |
+
"fig.colorbar(im1, ax=axs[0], shrink=0.87)\n",
|
170 |
+
"fig.colorbar(im2, ax=axs[1], shrink=0.87)\n",
|
171 |
+
"\n",
|
172 |
+
"t1 = axs[0].set_title(\"Observed - Noise: %i dB\"%nl)\n",
|
173 |
+
"t4 = axs[1].set_title(\"GT\")\n",
|
174 |
+
"plt.show()\n",
|
175 |
+
"\n"
|
176 |
+
]
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"cell_type": "code",
|
180 |
+
"execution_count": 4,
|
181 |
+
"id": "093c2a19",
|
182 |
+
"metadata": {},
|
183 |
+
"outputs": [],
|
184 |
+
"source": [
|
185 |
+
"np.savez('%s.npz'%DATASET, \n",
|
186 |
+
" gt_train = gt_train,\n",
|
187 |
+
" obs_train = obs_train,\n",
|
188 |
+
" gt_test = gt_test,\n",
|
189 |
+
" obs_test = obs_test,\n",
|
190 |
+
" gt_cal = gt_cal,\n",
|
191 |
+
" obs_cal = obs_cal,\n",
|
192 |
+
" gt_val = gt_val,\n",
|
193 |
+
" obs_val = obs_val)"
|
194 |
+
]
|
195 |
+
},
|
196 |
+
{
|
197 |
+
"cell_type": "code",
|
198 |
+
"execution_count": null,
|
199 |
+
"id": "344571db",
|
200 |
+
"metadata": {},
|
201 |
+
"outputs": [],
|
202 |
+
"source": []
|
203 |
+
}
|
204 |
+
],
|
205 |
+
"metadata": {
|
206 |
+
"kernelspec": {
|
207 |
+
"display_name": "Python 3 (ipykernel)",
|
208 |
+
"language": "python",
|
209 |
+
"name": "python3"
|
210 |
+
},
|
211 |
+
"language_info": {
|
212 |
+
"codemirror_mode": {
|
213 |
+
"name": "ipython",
|
214 |
+
"version": 3
|
215 |
+
},
|
216 |
+
"file_extension": ".py",
|
217 |
+
"mimetype": "text/x-python",
|
218 |
+
"name": "python",
|
219 |
+
"nbconvert_exporter": "python",
|
220 |
+
"pygments_lexer": "ipython3",
|
221 |
+
"version": "3.9.16"
|
222 |
+
}
|
223 |
+
},
|
224 |
+
"nbformat": 4,
|
225 |
+
"nbformat_minor": 5
|
226 |
+
}
|
summaries.csv
ADDED
The diff for this file is too large to render.
See raw diff
|
|