-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
410 lines (382 loc) · 16.5 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
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
408
409
410
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description"
content="The Fabrication of Reality and Fantasy: Scene Generation with LLM-Assisted Prompt Interpretation">
<meta name="keywords"
content="Text-to-image Generation, Realistic-Fantasy Benchmark, Diffusion Model, Large Language Models (LLMs)">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>The Fabrication of Reality and Fantasy: Scene Generation with LLM-Assisted Prompt Interpretation</title>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag() {
dataLayer.push(arguments);
}
gtag('js', new Date());
gtag('config', 'G-PYVRSFMDRL');
</script>
<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 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>
<style>
figcaption {
text-align: center;
margin-top: 5px;
font-size: 14px;
color: #555;
}
table {
width: 100%;
border-collapse: collapse;
margin: 20px 0;
}
th, td {
border: 1px solid #000;
padding: 8px;
text-align: center;
}
th {
background-color: #f2f2f2;
}
.highlight {
background-color: #f2f2f2;
}
.improvement {
color: red;
}
</style>
</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">The Fabrication of Reality and Fantasy: Scene Generation with
LLM-Assisted Prompt Interpretation</h1>
<h3 class="title is-3 publication-title">(ECCV 2024)</h3>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="">Yi Yao*</a><sup>1</sup>,</span>
<span class="author-block">
<a href="">Chan-Feng Hsu*</a><sup>1</sup>,</span>
<span class="author-block">
<a href="">Jhe-Hao Lin</a><sup>1</sup>,
</span>
<span class="author-block">
<a href="">Hongxia Xie</a><sup>2</sup>,
</span>
<span class="author-block">
<a href="">Terence Lin</a><sup>1</sup>,
</span>
<span class="author-block">
<a href="">Yi-Ning Huang</a><sup>1</sup>,
</span>
<span class="author-block">
<a href="https://basiclab.lab.nycu.edu.tw/">Hong-Han Shuai</a><sup>1</sup>
</span>
<span class="author-block">
<a href="https://www.csie.ntu.edu.tw/~wenhuang/">Wen-Huang Cheng</a><sup>3</sup>
</span>
</div>
<div>
<span class="author-block"><sup>*</sup>Equal contribution</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>1</sup>National Yang Ming Chiao Tung University,</span>
<span class="author-block"><sup>2</sup>Jilin University,</span>
<span class="author-block"><sup>3</sup>National Taiwan University</span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- PDF Link. -->
<span class="link-block">
<a href="https://arxiv.org/pdf/2407.12579" 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>
<!-- arXiv Link. -->
<span class="link-block">
<a href="https://arxiv.org/abs/2407.12579" 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>
<!-- Code Link. -->
<span class="link-block">
<a href="https://github.com/leo81005/Reality-and-Fantasy" 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>
<!-- Dataset Link. -->
<span class="link-block">
<a href="https://docs.google.com/spreadsheets/d/1AmOInyxVAnDgAILoJPDO6cJdm6wVesdBSSbx7NgEp88/edit?usp=sharing"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="far fa-images"></i>
</span>
<span>Dataset</span>
</a>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<figure align="center">
<img id="intro" src="./static/images/intro.png" alt="intro_result" style="max-width: 100%;">
</figure>
</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">
<p>
In spite of recent advancements in text-to-image generation, it still has limitations when it comes to
complex, imaginative text prompts. Due to the limited exposure to diverse and complex data in their
training sets, text-to-image models often struggle to comprehend the semantics of these difficult prompts,
leading to the generation of irrelevant images. This work explores how diffusion models can process and
generate images based on prompts requiring artistic creativity or specialized knowledge. Recognizing the
absence of a dedicated evaluation framework for such tasks, we introduce a new benchmark, the
Realistic-Fantasy Benchmark (RFBench), which blends scenarios from both realistic and fantastical realms.
Accordingly, for reality and fantasy scene generation, we propose an innovative training-free approach,
Realistic-Fantasy Network (RFNet), that integrates diffusion models with LLMs. Through our proposed
RFBench, extensive human evaluations coupled with GPT-based compositional assessments have demonstrated
our approach's superiority over other state-of-the-art methods.
</p>
</div>
</div>
</div>
<!--/ Abstract. -->
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Method. -->
<div class="columns is-centered">
<div class="column is-full-width">
<h2 class="title is-3">Method</h2>
<!-- Overview -->
<h3 class="title is-4">Overview</h3>
<div class="content has-text-justified">
<p>
The Realistic-Fantasy Network (RFNet) contains two stages.
In the first stage, we transform the initial input prompt into a refined version specifically tailored for
image generation by LLMs.
In the second stage, we utilize a diffusion model through a two-step process to generate outputs with
extraordinary details.
</p>
<img id="method" src="./static/images/framework.png" alt="method" height="100%">
</div>
<!-- SAA Module -->
<h3 class="title is-4">Semantic Alignment Assessment (SAA) Module</h3>
<div class="content has-text-justified">
<p>
As we proceed with generating images using the diffusion model using the details generated by the previous
step, there is a critical challenge: <i>the description lists generated by LLMs for one object usually
overlook the relationships among them.</i> For example, interpretations of “a lion” could range from
being “unaware and asleep” to “frightened and trying to escape.” Although both depictions are valid,
descriptions such as “unaware” and “trying to escape” can lead to conflicting interpretations, thus
complicating the image generation process.
</p>
<p>
To overcome this challenge, we introduce the <b>Semantic Alignment Assessment (SAA)</b> module. This
module calculates the relevance between different object vectors, thereby selecting the candidate
description that best fits the current scenario. By conducting the cosine similarity among different
descriptions, we can navigate the complexities introduced by the LLM's output, selecting the most
compatible details for the diffusion model. This step is crucial for maintaining the coherence and
accuracy of the generated images, highlighting our novel approach to mitigating the risk of conflicting
descriptions. Through this module, we ensure textual precision and compatibility, and provide <i>clear,
consistent instructions</i> for the subsequent diffusion model to generate visually coherent
representations.
</p>
<img id="method" src="./static/images/fig_SAA.jpg" alt="method" height="100%">
</div>
</div>
</div>
<!-- end Method. -->
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Qualitative Result. -->
<div class="columns is-centered">
<div class="column is-full-width">
<h2 class="title is-3">Qualitative Result</h2>
<div class="image-container">
<figure><img id="result" src="./static/images/result.jpg" alt="result" height="100%">
<figcaption>Qualitative comparison on RFBench. The compared models include (a) Stable
Diffusion, (b) MultiDiffusion, (c) Attend and Excite, (d) LMD, (e) BoxDiff, (f) SDXL,
(g) Ours</figcaption>
</figure>
<figure><img id="result" src="./static/images/more_fig1.png" alt="result" height="100%">
<figcaption>More results on <b>Realistic and Analytical</b>. The compared models include (a)
Stable Diffusion, (b) MultiDiffusion, (c) Attend and Excite, (d) LMD, (e) BoxDiff, (f)
SDXL, (g) Ours
</figcaption>
</figure>
<figure><img id="result" src="./static/images/more_fig2.png" alt="result" height="100%">
<figcaption>More results on <b>Creativity and Imagination</b>. The compared models include
(a) Stable Diffusion, (b) MultiDiffusion, (c) Attend and Excite, (d) LMD, (e) BoxDiff,
(f) SDXL, (g) Ours
</figcaption>
</figure>
</div>
</div>
</div>
<!-- Quantitative Result. -->
<div class="columns is-centered">
<div class="column is-full-width">
<h2 class="title is-3">Quantitative Result</h2>
<table>
<thead>
<tr>
<th rowspan="2" style="vertical-align: middle;" align="center">Model</th>
<th colspan="3" align="center">GPT4-CLIP</th>
<th colspan="3" align="center">GPT4Score</th>
</tr>
<tr align="center">
<th>R & A</th>
<th>C & I</th>
<th>Avg</th>
<th>R & A</th>
<th>C & I</th>
<th>Avg</th>
</tr>
</thead>
<tbody align="center">
<tr>
<td>Stable Diffusion</td>
<td>0.573</td>
<td>0.552</td>
<td>0.561</td>
<td>0.667</td>
<td>0.440</td>
<td>0.541</td>
</tr>
<tr>
<td>MultiDiffusion</td>
<td>0.510</td>
<td>0.510</td>
<td>0.510</td>
<td>0.517</td>
<td>0.493</td>
<td>0.504</td>
</tr>
<tr>
<td>Attend and Excite</td>
<td>0.523</td>
<td>0.560</td>
<td>0.546</td>
<td>0.633</td>
<td>0.520</td>
<td>0.570</td>
</tr>
<tr>
<td>LLM-groundedDiffusion</td>
<td>0.457</td>
<td>0.536</td>
<td>0.501</td>
<td>0.550</td>
<td>0.600</td>
<td>0.578</td>
</tr>
<tr>
<td>BoxDiff</td>
<td>0.532</td>
<td>0.553</td>
<td>0.543</td>
<td>0.583</td>
<td>0.520</td>
<td>0.548</td>
</tr>
<tr>
<td>SDXL</td>
<td>0.536</td>
<td>0.619</td>
<td>0.582</td>
<td>0.567</td>
<td>0.587</td>
<td>0.578</td>
</tr>
<tr class="highlight">
<td><b>RFNet (ours)</b></td>
<td><b>0.587 </b><span class="improvement">(2%↑)</span></td>
<td><b>0.623 </b><span class="improvement">(13%↑)</span></td>
<td><b>0.607 </b><span class="improvement">(8%↑)</span></td>
<td><b>0.833 </b><span class="improvement">(25%↑)</span></td>
<td><b>0.627 </b><span class="improvement">(43%↑)</span></td>
<td><b>0.719 </b><span class="improvement">(33%↑)</span></td>
</tr>
</tbody>
</table>
</div>
</div>
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>@article{yao2024fabricationrealityfantasyscene,
title = {The Fabrication of Reality and Fantasy: Scene Generation with LLM-Assisted Prompt Interpretation},
author = {Yi Yao and Chan-Feng Hsu and Jhe-Hao Lin and Hongxia Xie and Terence Lin and Yi-Ning Huang and Hong-Han Shuai and Wen-Huang Cheng},
year = {2024},
eprint = {2407.12579},
archivePrefix = {arXiv},
primaryClass = {cs.CV},
url = {https://arxiv.org/abs/2407.12579},
}</code></pre>
</div>
</section>
<footer class="footer">
<!-- <div class="container">
<div class="content has-text-centered">
<a class="icon-link" href="">
<i class="fas fa-file-pdf"></i>
</a>
<a class="icon-link" href="" 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 adapted from <a href="https://github.com/nerfies/nerfies.github.io">Nerfies</a>,
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>
</div>
</div>
</div>
</div>
</footer>
</body>
</html>