-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
360 lines (320 loc) · 15.5 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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description" content="xT: Nested Tokenization for Larger Context in Large Images">
<meta name="keywords" content="">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>🌐 xT - BAIR Climate Initiative</title>
<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">
<link rel="apple-touch-icon" sizes="180x180" href="/apple-touch-icon.png">
<link rel="icon" type="image/png" sizes="32x32" href="/favicon-32x32.png">
<link rel="icon" type="image/png" sizes="16x16" href="/favicon-16x16.png">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.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> -->
<script src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script>
<script id="MathJax-script" src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
<meta property="og:url" content="https://ai-climate.berkeley.edu/xt-website/">
<meta property="og:type" content="website">
<meta property="og:title" content="xT: Nested Tokenization for Larger Context in Large Images">
<meta property="og:description" content="xT enables powerful, yet myopic vision backbones to model extremely large images on contemporary GPUs.">
<meta property="og:image" content="https://ai-climate.berkeley.edu/xt-website/static/images/xt.png">
<meta name="twitter:card" content="summary_large_image">
<meta property="twitter:domain" content="ai-climate.berkeley.edu">
<meta property="twitter:url" content="https://ai-climate.berkeley.edu/xt-website/">
<meta name="twitter:title" content="xT: Nested Tokenization for Larger Context in Large Images">
<meta name="twitter:description" content="xT enables powerful, yet myopic vision backbones to model extremely large images on contemporary GPUs.">
<meta name="twitter:image" content="https://ai-climate.berkeley.edu/xt-website/static/images/xt.png">
</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">\(x\)T</h1>
<h2 class="title is-2 publication-title">Nested Tokenization for Larger Context in Large Images</h2>
<div class="is-size-5 publication-authors">
<a href="https://ritwikgupta.me">Ritwik Gupta</a>*<sup>1</sup>,</span>
<span class="author-block">
<a href="https://homepage.jackli.org/">Shufan Li</a>*<sup>2</sup>,
</span>
<span class="author-block">
<a href="https://tylerzhu.com/research/">Tyler Zhu</a>*<sup>1,3</sup>,
</span>
<br>
<span class="author-block">
<a href="https://people.eecs.berkeley.edu/~malik/">Jitendra Malik</a><sup>1</sup>,
</span>
<span class="author-block">
<a href="https://people.eecs.berkeley.edu/~trevor">Trevor Darrell</a><sup>1</sup>,
</span>
<span class="author-block">
<a href="https://karttikeya.github.io/">Karttikeya Mangalam</a><sup>1</sup>
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>1</sup>Berkeley AI Research, UC Berkeley,</span>
<span class="author-block"><sup>2</sup>UCLA</span>,
<span class="author-block"><sup>3</sup>Princeton University</span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- PDF Link. -->
<span class="link-block">
<a href="https://arxiv.org/abs/2403.01915" 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>
<!-- Code Link. -->
<span class="link-block">
<a href="https://github.com/bair-climate-initiative/xT/"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code and Models</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<video id="teaser" autoplay muted loop playsinline>
<source src="./static/videos/xT-Teaser.mp4" type="video/mp4">
<!-- <source src="./static/videos/xT-Teaser.webm" type="video/webm"> -->
</video>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Abstract</h2>
<!-- <div class="content has-text-justified"> -->
<div class="content sans-serif has-text-justified is-6">
<p>
Modern computer vision pipelines handle large images in one of two sub-optimal ways: down-sampling or
cropping.
These two methods incur significant losses in the amount of information and context present in an image.
There are many downstream applications in which global context matters as much as high frequency details,
such as in real-world satellite imagery; in such cases researchers have to make the uncomfortable choice
of which information to discard.
We introduce \(x\)T, a simple framework for vision transformers which effectively
aggregates global context with local details and can model large images end-to-end on contemporary GPUs.
We select a set of benchmark datasets across classic vision tasks which accurately reflect a vision
model's ability to understand truly large images and incorporate fine details over large scales and assess
our method's improvement on them.
By introducing a nested tokenization scheme for large images in conjunction with long-sequence length
models normally used for natural language processing, we are able to increase accuracy by up to 8.6% on
challenging classification tasks and \(F_1\) score by 11.6 on context-dependent segmentation in large
images.
</p>
</div>
</div>
</div>
<!--/ Abstract. -->
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<h2 class="title is-3">Images are Getting Bigger</h2>
<div class="column has-text-centered">
<img src="./static/images/football.png"
width="100%"
center
alt="A photo of a football field."/>
</div>
<!-- <img src="./static/images/football.png"> -->
<div class="content has-text-justified">
<!-- <h3 class="subtitle sans-serif has-text-justified is-6"> -->
<p>
Images have been getting increasingly larger over the past decade. For example, consider a video feed of
a
football game which is captured natively in 8K resolution. We would like to understand where the player
in
the middle of the screen is passing the ball to. However, today's leading models would not be able to
reason over the entire image in one pass.
</p>
<!-- </h3> -->
<!-- <p></p> -->
<!-- <h3 class="subtitle sans-serif has-text-justified is-6"> -->
<p>
Modern computer vision pipelines are limited by the memory in the systems they are trained upon,
resulting
in the creation of models that only operate on small images. Computer vision practitioners limit the
size
of images in two less-than-ideal ways: down-sampling or cropping. While these simple operations produce
powerful models when measured against typical computer vision benchmarks, the loss of high frequency
information or global context is limited for many real-world tasks.
</p>
<!-- </h3> -->
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<h2 class="title is-3">Using \(x\)T to Model Large Images</h2>
<div class="columns is-centered is-full-width">
<div class="column">
<img src="./static/images/xt.png">
<h3 class="has-text-centered">Figure 1: Architecture for the \(x\)T framework.</h2>
</div>
</div>
<div class="content has-text-justified">
<p>
\(x\)T is framework that allows existing vision backbones to process large images in a memory efficient and
contextual manner.
We achieve this through an iterative, two-stage design.
</p>
<p>
First, images are tokenized hierarchically (Nested Tokenization) before being independently
featurized by a region encoder with a limited context window (Independent Region Encoding).
Then, a lightweight context encoder incorporates context globally across this sequence of features
(Context-Aware Encoding), which then gets passed to the task-specific decoders.
</p>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<h2 class="title is-3">Results</h2>
<div class="columns is-centered is-full-width">
<div class="column has-text-centered">
<img src="./static/images/model_performance_plot_mamba.png"
width="80%"
center
alt="xT sets a new frontier on downstream tasks."/>
<h3 class="has-text-centered">Figure 2: Powerful vision models used with \(x\)T set a new frontier on
downstream tasks.
</h3>
</div>
</div>
<div class="content has-text-justified">
<p>
The use of \(x\)T allows myopic, memory-hungry vision backbones to effectively "see" across the entire large
image at once. On tasks such as classification (iNaturalist-Reptilia shown in the figure), \(x\)T can achieve
higher accuracy with fewer parameters due to its ability to incorporate global context across local regions of
the image.
</p>
</div>
<div class="columns is-centered is-full-width">
<div class="column has-text-centered">
<img src="./static/images/ERF.png"
width="70%"
center/>
<h3 class="has-text-centered">Figure 3: \(x\)T increases the receptive field of vision backbones.</h3>
</div>
</div>
<div class="content has-text-justified">
<p>
This is best visualized through Figure 3, which demonstrates the effective receptive field of Swin-B and
Swin-B <\(x\)T> XL as the input image gets larger. Swin-B cannot model an image that is >2,800 x
2,800
pixels large, while it can modeled with \(x\)T properly.
</p>
</div>
<div class="columns is-centered is-full-width">
<div class="column has-text-centered">
<img src="./static/images/mem.png"
width="70%"
center/>
<h3 class="has-text-centered">Figure 4: \(x\)T increases the receptive field of vision backbones.</h3>
</div>
</div>
<div class="content has-text-justified">
<p>
Critically, as inputs get larger, backbones such as Swin scale memory usage quadratically, whereas \(x\)T
memory
usage stays near-constant per region. This enables entirely new classes of applications not possible before,
such as the effective processing of images captured from large-format sensors such as satellites and
microscopes.
</p>
</div>
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>@article{xTLargeImageModeling,
title={xT: Nested Tokenization for Larger Context in Large Images},
author={Gupta, Ritwik and Li, Shufan and Zhu, Tyler and Malik, Jitendra and Darrell, Trevor and Mangalam, Karttikeya},
journal={arXiv preprint arXiv:2403.01915},
year={2024}
}</code></pre>
</div>
</section>
<footer class="footer">
<div class="container">
<div class="content has-text-centered">
<a class="icon-link" href="https://arxiv.org/abs/2403.01915">
<i class="fas fa-file-pdf"></i>
</a>
<a class="icon-link" href="https://github.com/bair-climate-initiative/xT/" class="external-link" disabled>
<i class="fab fa-github"></i>
</a>
</div>
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This website is licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
<p>
This website came from the <a href="https://github.com/nerfies/nerfies.github.io">Nerfies project
website
template</a>.
</p>
</div>
</div>
</div>
</div>
</footer>
</body>
<script>
const container = document.querySelector('.container');
function updateContainerWidth() {
const windowWidth = window.innerWidth;
const newContainerWidth = windowWidth < 800 ? '90%' : '65%';
document.documentElement.style.setProperty('--container-width', newContainerWidth);
}
updateContainerWidth();
// Listen for the resize event on the window
window.addEventListener('resize', updateContainerWidth);
// // Initial setup when the page loads
// document.addEventListener('DOMContentLoaded', updateContainerWidth);
// Initial setup when the page loads
document.addEventListener('DOMContentLoaded', function() {
var teaserVideo = document.getElementById('teaser');
// Check if the video element exists
if (teaserVideo) {
// Set the playbackRate to 1.25 to make it play 25% faster
teaserVideo.playbackRate = 1.25;
}
});
</script>
</html>