-
Notifications
You must be signed in to change notification settings - Fork 7
/
Copy pathindex.html
407 lines (356 loc) · 17.9 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<!-- Meta tags for social media banners, these should be filled in appropriatly as they are your "business card" -->
<!-- Replace the content tag with appropriate information -->
<meta name="description" content="DESCRIPTION META TAG">
<meta property="og:title" content="SOCIAL MEDIA TITLE TAG" />
<meta property="og:description" content="SOCIAL MEDIA DESCRIPTION TAG TAG" />
<meta property="og:url" content="URL OF THE WEBSITE" />
<!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X630-->
<meta property="og:image" content="static/image/your_banner_image.png" />
<meta property="og:image:width" content="1200" />
<meta property="og:image:height" content="630" />
<meta name="twitter:title" content="TWITTER BANNER TITLE META TAG">
<meta name="twitter:description" content="TWITTER BANNER DESCRIPTION META TAG">
<!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X600-->
<meta name="twitter:image" content="static/images/your_twitter_banner_image.png">
<meta name="twitter:card" content="summary_large_image">
<!-- Keywords for your paper to be indexed by-->
<meta name="keywords" content="KEYWORDS SHOULD BE PLACED HERE">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>SDSeg</title>
<link rel="icon" type="image/x-icon" href="static/images/favicon.ico">
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
<link rel="stylesheet" href="static/css/bulma.min.css">
<link rel="stylesheet" href="static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="static/css/bulma-slider.min.css">
<link rel="stylesheet" href="static/css/fontawesome.all.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="static/css/index.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://documentcloud.adobe.com/view-sdk/main.js"></script>
<script defer src="static/js/fontawesome.all.min.js"></script>
<script src="static/js/bulma-carousel.min.js"></script>
<script src="static/js/bulma-slider.min.js"></script>
<script src="static/js/index.js"></script>
<!-- Google tag (gtag.js) -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-MPD6C77MDF"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'G-MPD6C77MDF');
</script>
</head>
<body>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title">
<span
style="background: linear-gradient(to bottom, indigo, skyblue, violet, indigo, violet); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Stable
</span> Diffusion Segmentation <br>
for Biomedical Images with
<span
style="background: linear-gradient(to left, indigo, skyblue, violet, indigo, violet); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Single-step
</span> Reverse Process</h1>
<div class="is-size-5 publication-authors">
<!-- Paper authors -->
<span class="author-block">
<a href="https://lin-tianyu.github.io/" target="_blank">Tianyu Lin</a><sup>1</sup>,</span>
<span class="author-block">
<a target="_blank">Zhiguang Chen</a><sup>2</sup>,</span>
<span class="author-block">
<a href="https://github.com/zzzyzh" target="_blank">Zhonghao Yan</a><sup>3</sup>,</span>
<span class="author-block">
<a href="https://github.com/yuweijiang" target="_blank">Weijiang Yu</a><sup>2*</sup>,</span>
<span class="author-block">
<a target="_blank">Fudan Zheng</a><sup>2*</sup></span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block">
{<sup>1</sup>School of Biomedical Engineering,
<sup>2</sup>School of Computer Science and Engineering}, Sun Yat-sen University<br>
<sup>3</sup>International School, Beijing University of Posts and Telecommunications
</span>
<span class="eql-cntrb"><small><br><sup>*</sup>Corresponding Author</small></span><br>
<span class="author-block" style="background: linear-gradient(to right, rgb(255, 0, 0), rgb(0, 255, 0), rgb(0, 0, 0)); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;"><b>MICCAI 2024</b></span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- ArXiv abstract Link -->
<span class="link-block">
<a href="https://arxiv.org/abs/2406.18361" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>ArXiv</span>
</a>
</span>
<!-- Arxiv PDF link -->
<span class="link-block">
<a href="https://link.springer.com/chapter/10.1007/978-3-031-72111-3_62" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Paper</span>
</a>
</span>
<!-- Github link -->
<span class="link-block">
<a href="https://github.com/lin-tianyu/Stable-Diffusion-Seg" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Teaser-->
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/framework-v2.jpg" alt="MY ALT TEXT" , width="700" />
<h4 class="subtitle has-text-centered">The overview of <b>SDSeg</b>.</h4>
</div>
</div>
</section>
<!-- End teaser -->
<section class="section hero is-light">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-2" style="background: linear-gradient(to right, indigo, indigo, skyblue, indigo, indigo); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Highlights</h2>
</div>
</div>
<p>
🌟 <b><u>Stable</u></b> not only shows that SDSeg is built on Stable Diffusion but also indicates its remarkable stability.<br>
🌟 SDSeg only requires a <b>single-step reverse process</b> to generate segmentation results.<br>
🌟 SDSeg has remarkable <b>stability</b> and doesn't need to sample multiple times for average.
</p>
</div>
</section>
<section class="section hero">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-2" style="background: linear-gradient(to right, indigo, indigo, skyblue, indigo, indigo); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Motivation</h2>
<h5 class="title is-5" style="text-align: center;">
Inspired by the <a href="https://en.wikipedia.org/wiki/Occam%27s_razor">Occam's razor</a>✂️, we believe that remove redundant techniques in conventional diffusion models can benefit segmentation task.
</h5>
<div class="columns is-centered">
<!-- reenact. -->
<div class="column">
<div class="content">
<h4 class="title is-5" style="text-align: center;">🔒 Challenges</h4>
<p>
(1) <b>Pixel-level Diffusion process</b> is computing resource-consuming during training.<br>
<!-- (1) <b>Diffusion process in pixel-space</b> brings computing resource-consuming issue during training stage.<br> -->
(2) <b>Multi-step reverse process</b> and <b>multiple-sample average scheme</b> are time-consuming during inference.
</p>
</div>
</div>
<!--/ reenact -->
<!-- swap. -->
<div class="column">
<h4 class="title is-5" style="text-align: center;">🔑 Solutions</h4>
<div class="columns is-centered">
<div class="column content">
<p>
<!-- <br> -->
(1) Using <b>Latent-level Diffusion</b> model (Stable Diffusion, SD).<br>
Also, SD's Autoencoder can generalize to segmentation maps. <b> No fintune needed.</b><br>
<!-- (1) Building our model based on Stable Diffusion model, which conduct diffusion process on a much smaller <b>latent space</b>.<br> -->
(2) Designing a <b>single-step reverse process</b> with strong <b>stability against initial noise</b>.
<!-- (2) Constructing a <b>single-step reverse process</b> with strong <b>stability among different initial noise</b> to speed up inference. -->
</p>
</div>
</div>
</div>
<!--/ swap. -->
</div>
<div class="hero-body">
<img src="static/images/LatentRep.jpg" alt="MY ALT TEXT" , width="1000" />
<h4 class="subtitle has-text-centered"><b> visualization of latent representation maps.</b></h4>
The latent representations has high similarity among their corresponding segmentation maps, and segmentation maps have much less semantic knowledge comparing to RGB images. These observations indicate a much simpler diffusion process can be established.
</div>
</div>
</div>
</div>
</section>
<section class="section hero is-light">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-2" style="background: linear-gradient(to right, indigo, indigo, skyblue, indigo, indigo); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Single-step Reverse process</h2>
<h5 class="title is-5" style="text-align: center;">
SDSeg only needs ☝️ step to generate the final segmentation map, which even has better result comparing to dozens of steps of DDIM sampling.
</h5>
<div class="hero-body">
<img src="static/images/ReverseCurve.png" alt="MY ALT TEXT" , width="900" />
<h4 class="subtitle has-text-centered"><b>Comparison of DDIM convergence speed w/ and w/o latent estimation.</b></h4>
</div>
<!-- <div class="hero-body">
<img src="static/images/ReverseProcess.jpeg" alt="MY ALT TEXT" , width="1000" />
<h4 class="subtitle has-text-centered">Visualization of the predicted probability maps in
reverse process.</h4>
</div> -->
</div>
</div>
</div>
<!-- Video carousel -->
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<div id="results-carousel" class="carousel results-carousel">
<div class="item item1" style="text-align:center;">
<img src="static/images/ReverseProcess-1.png" alt="MY ALT TEXT" , width="700" />
</div>
<div class="item item2" style="text-align:center;">
<img src="static/images/ReverseProcess-2.png" alt="MY ALT TEXT" , width="700" />
</div>
</div>
</div>
</div>
<h4 class="subtitle has-text-centered"><b>Visualization of the predicted probability maps in reverse process (DDIM sampler).</b></h4>
</section>
<!-- End video carousel -->
</section>
<section class="section hero">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-2" style="background: linear-gradient(to right, indigo, indigo, skyblue, indigo, indigo); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Stability Evaluation</h2>
<h5 class="title is-5" style="text-align: center;">
We proposed a Stability Evaluation scheme to measure the stability of any diffusion-based segmentation methods:
</h5>
<p style="text-align: left;">
<b>👉 Dataset-level Stability</b>: performs repeated inferences on test data to measure variability.<br>
<b>👉 Instance-level Stability</b>: examines the model’s consistency under varying initial noise.
</p>
<div class="hero-body">
<img src="static/images/StabilityEvaluationV4.jpeg" alt="MY ALT TEXT" , width="1000" />
<h2 class="subtitle has-text-centered"><b>The proposed Stability Evaluation scheme.</b></h2>
</div>
</div>
</div>
</div>
</section>
<section class="section hero is-light">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-2" style="background: linear-gradient(to right, indigo, indigo, skyblue, indigo, indigo); -webkit-background-clip: text; -webkit-text-fill-color: transparent;line-height: 1.5;">
Other Qualitative Analysis</h2>
<div class="hero-body">
<img src="static/images/Conditioner.png" alt="MY ALT TEXT" , width="900" />
<h2 class="subtitle has-text-centered"><b>Visualization of the latent representations of medical images from the trainable
vision encoder.</b></h2>
At iteration 0, the encoder pre-trained on natural images
couldn’t capture enough meaningful semantic features for segmentation. During training, the conditioning encoder gradually learns to focus on segmentation targets.
</div>
</div>
</div>
</div>
</section>
<!-- Video carousel -->
<!-- <section class="hero is-small">
<div class="hero-body">
<div class="container">
<h2 class="title is-3">Another Carousel</h2>
<div id="results-carousel" class="carousel results-carousel">
<div class="item item-video1">
<video poster="" id="video1" autoplay controls muted loop height="100%">
<source src="static/videos/carousel1.mp4" type="video/mp4">
</video>
</div>
<div class="item item-video2">
<video poster="" id="video2" autoplay controls muted loop height="100%">
<source src="static/videos/carousel2.mp4" type="video/mp4">
</video>
</div>
<div class="item item-video3">
<video poster="" id="video3" autoplay controls muted loop height="100%">\
<source src="static/videos/carousel3.mp4" type="video/mp4">
</video>
</div>
</div>
</div>
</div>
</section> -->
<!-- End video carousel -->
<!-- Paper poster -->
<!-- <section class="hero is-small is-light">
<div class="hero-body">
<div class="container">
<h2 class="title">Poster</h2>
<iframe src="static/pdfs/sample.pdf" width="100%" height="550"></iframe>
</div>
</div>
</section> -->
<!--End paper poster -->
<!--BibTex citation -->
<section class="section hero" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTex</h2>
<pre><code>
@InProceedings{lin2024stable,
author="Lin, Tianyu
and Chen, Zhiguang
and Yan, Zhonghao
and Yu, Weijiang
and Zheng, Fudan",
title="Stable Diffusion Segmentation for Biomedical Images with Single-Step Reverse Process",
booktitle="Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024",
year="2024",
publisher="Springer Nature Switzerland",
address="Cham",
pages="656--666",
isbn="978-3-031-72111-3"
}
</code></pre>
</div>
</section>
<!--End BibTex citation -->
<footer class="footer">
<div class="container">
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This page was built using the <a href="https://github.com/eliahuhorwitz/Academic-project-page-template"
target="_blank">Academic Project Page Template</a> which was adopted from the <a
href="https://nerfies.github.io" target="_blank">Nerfies</a> project page.
<br> This website is licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/" target="_blank">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
</div>
</div>
</div>
</div>
</footer>
<!-- Statcounter tracking code -->
<!-- You can add a tracker to track page visits by creating an account at statcounter.com -->
<!-- End of Statcounter Code -->
</body>
</html>