-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
604 lines (530 loc) · 24.2 KB
/
Copy pathindex.html
File metadata and controls
604 lines (530 loc) · 24.2 KB
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
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
<!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>RAAG</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>
</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">RAAG: Ratio Aware Adaptive Guidance</h1>
<div class="is-size-5 publication-authors">
<!-- Paper authors -->
<span class="author-block">
<a href="https://github.com/zhushangwen" target="_blank">Shangwen Zhu</a><sup>1*</sup>,</span>
<span class="author-block">
<a href="https://github.com/xiaomomy" target="_blank">Qianyu Peng</a><sup>1*</sup>,</span>
<span class="author-block">
<a href="THIRD AUTHOR PERSONAL LINK" target="_blank">Yuting Hu</a><sup>1</sup>,</span>
<span class="author-block">
<a href="4 AUTHOR PERSONAL LINK" target="_blank">Zhantao Yang</a><sup>1</sup>,</span>
<span class="author-block">
<a href="4 AUTHOR PERSONAL LINK" target="_blank">Han Zhang</a><sup>1</sup>,</span>
<span class="author-block">
<a href="5 AUTHOR PERSONAL LINK" target="_blank">Zhao Pu</a><sup>1</sup>,</span>
<span class="author-block">
<a href="5 AUTHOR PERSONAL LINK" target="_blank">Ruili Feng</a><sup>2</sup>,</span>
<span class="author-block">
<a href="5 AUTHOR PERSONAL LINK" target="_blank">Fan Cheng</a><sup>1†</sup></span>
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block">
<sup>1</sup>Shanghai Jiao Tong University<br>
<sup>2</sup>University of Waterloo<br>
</span>
<div style="text-align: center; margin-top: 1em;">2025</div>
<span class="eql-cntrb"><small><br><sup>*</sup>Indicates Equal Contribution</small></span>
<span class="corr-author"><small><br><sup>†</sup>Corresponding Author</small></span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- Arxiv PDF link -->
<span class="link-block">
<a href="https://arxiv.org/abs/2508.03442" 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>
<!-- Supplementary PDF link -->
<!-- <span class="link-block">
<a href="static/pdfs/supplementary_material.pdf" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Supplementary</span>
</a>
</span> -->
<!-- Github link -->
<!-- <span class="link-block">
<a href="https://github.com/zhushangwen/LTC-Accel" 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> -->
<!-- ArXiv abstract Link -->
<!-- <span class="link-block">
<a href="https://arxiv.org/abs/2508.03442" 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> -->
<!-- <span class="link-block">
<a href="https://colab.research.google.com/drive/1h0SM3dp7oUzjBdahmwEcJn9RCq2gryD1#scrollTo=q7u54mHpyU4T"
target="_blank"
class="external-link button is-normal is-rounded"
style="background-color: black; color: white; border-color: #F9AB00;">
<span class="icon">
<i class="fas fa-code"></i> -->
<!-- </span>
<span>Colab</span>
<span class="icon" style="margin-left: 5px;">
<img src="https://colab.research.google.com/img/colab_favicon.ico"
alt="Colab"
style="height: 1em; vertical-align: middle;"> 官方Colab图标
</span> -->
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<div id="results-carousel">
<!-- 第一行6张图 - 缩小横向间距 -->
<div class="columns is-centered is-gapless" style="margin-bottom: 0.5em;"> <!-- 缩小行间距 -->
<div class="column is-one-sixth" style="padding: 0.1rem;"> <!-- 缩小列间距 -->
<center>
<img src="static/images/sd35/org10/0.png" alt="Image 1"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/sd35/org30/0.png" alt="Image 2"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/sd35/ours10/0.png" alt="Image 3"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/sd35/org10/34.png" alt="Image 4"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/sd35/org30/34.png" alt="Image 5"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/sd35/ours10/34.png" alt="Image 6"/>
</center>
</div>
</div>
<!-- 第二行6张图 - 缩小横向间距 -->
<div class="columns is-centered is-gapless" style="margin-bottom: 0.5em;"> <!-- 缩小行间距 -->
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/org10/21.png" alt="Image 7"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/org30/21.png" alt="Image 8"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/ours10/21.png" alt="Image 9"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/org10/30.png" alt="Image 10"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/org30/30.png" alt="Image 11"/>
</center>
</div>
<div class="column is-one-sixth" style="padding: 0.1rem;">
<center>
<img src="static/images/lumina/ours10/30.png" alt="Image 12"/>
</center>
</div>
</div>
<h2 class="subtitle is-6 has-text-centered" style="margin-top: 1em; color: gray; font-size: medium;">
Results on Stable Diffusion v3.5 (upper) and Lumina-Next (lower). Left: Baseline with 10 steps. Middle: Baseline with 30 steps. Right: RAAG with 10 steps.
</h2>
</div>
</div>
</div>
</section>
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<!-- 第一行：两个 GIF -->
<div class="columns is-centered">
<div class="column is-half">
<div class="item">
<center>
<img src="static/images/org10.gif" alt="GIF 1"/>
</center>
</div>
</div>
<div class="column is-half">
<div class="item">
<center>
<img src="static/images/org.gif" alt="GIF 2"/>
</center>
</div>
</div>
</div>
<!-- 第二行：两个 GIF -->
<div class="columns is-centered">
<div class="column is-half">
<div class="item">
<center>
<img src="static/images/next10.gif" alt="GIF 3"/>
</center>
</div>
</div>
<div class="column is-half">
<div class="item">
<center>
<img src="static/images/rg.gif" alt="GIF 4"/>
</center>
</div>
</div>
</div>
<h2 class="subtitle is-6 has-text-centered" style="margin-top: 1em; color: gray; font-size: medium;">
Results comparison: WAN2.1-14B (left) vs WAN2.1-1.4B (right). Top row: original videos; Bottom row: RAAG-processed.
</h2>
</div>
</div>
</section>
<!-- Teaser video-->
<!-- <section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/images/example.png"
type="video/mp4">
</video>
<h2 class="subtitle has-text-centered">
LTC-Accel is a training-free acceleration method by identifying Local Local Transition Coherence and approximating
certain steps to reduce redundant computations.
It is both efficient and widely compatible with various models and schedulers.
</h2>
</div>
</div>
</section> -->
<!-- End teaser video -->
<!-- Paper abstract -->
<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-3">Abstract</h2>
<div class="content has-text-justified">
<p>
Flow-based generative models have recently achieved remarkable progress in
image and video synthesis, with classifier-free guidance (CFG) becoming the
standard tool for high-fidelity, controllable generation. However, despite
their practical success, little is known about how guidance interacts with
different stages of the sampling process-especially in the fast, low-step
regimes typical of modern flow-based pipelines. In this work, we uncover
and analyze a fundamental instability: the earliest reverse steps are
acutely sensitive to the guidance scale, owing to a pronounced spike in the
relative strength (RATIO) of conditional to unconditional predictions.
Through rigorous theoretical analysis and empirical validation, we show that
this RATIO spike is intrinsic to the data distribution, independent of the model
architecture, and causes exponential error amplification when paired with strong
guidance. To address this, we propose a simple, theoretically grounded,
RATIO-aware adaptive guidance schedule that automatically dampens the guidance scale
at early steps based on the evolving RATIO, using a closed-form exponential decay.
Our method is lightweight, requires no additional inference overhead, and is
compatible with standard flow frameworks. Experiments across state-of-the-art image
(SD3.5, Lumina) and video (WAN2.1) models demonstrate that our approach enables up
to 3x faster sampling while maintaining or improving generation quality,
robustness, and semantic alignment. Extensive ablation studies further confirm the
generality and stability of our schedule across models, datasets, and hyperparameters. Our findings highlight the critical role of stepwise guidance adaptation in unlocking the full potential of fast flow-based generative models.
</p>
<b>Keywords: Rectified Flow, Classifier-Free Guidance, Early-Step Instability</b>
</div>
</div>
</div>
</div>
</section>
<!-- End paper abstract -->
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<div class="columns is-centered has-text-centered">
<h2 class="title is-3">Method Overview<br></h2>
</div>
<div class="columns" style="margin-top: 1.5em;">
<div class="column" style="padding-left: 15rem; color: black; font-size: medium;">
<h4>Our method is based on the key concept RATIO, defined as the relative strength of conditional to uncondtional predictions:</h4>
</div>
</div>
<div class="item" style="margin-top: 1.5em">
<center>
<img src="static/images/def_ratio.png" alt="MY ALT TEXT" width="900"/>
</center>
</div>
<h3 class="subtitle is-6 has-text-centered" style="margin-top: 1.5em; color: gray; font-size: medium;"">
Definition of RATIO.
</h3>
<div class="columns" style="margin-top: 1.5em;">
<div class="column" style="padding-left: 15rem; color: black; font-size: medium;">
<h4>RAAG provides a dynamic guidance schedule adaptive to RATIO at each step as follows:</h4>
</div>
</div>
<div class="item" style="margin-top: 1.5em">
<center>
<img src="static/images/method.png" alt="MY ALT TEXT" width="400"/>
</center>
</div>
<h3 class="subtitle is-6 has-text-centered" style="margin-top: 1.5em; color: gray; font-size: medium;"">
A concise illustration of RAAG's procedure.
</h3>
</div>
</div>
</div>
</div>
</section>
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<div class="columns is-centered has-text-centered">
<h2 class="title is-3">Results<br></h2>
</div>
<div class="columns is-centered has-text-centered" style="margin-top: 1.5em; margin-bottom: 1.5em; color: black; font-size: medium;">
<h3> The original images are depicted in the first (10-step) and second (30-step) row, and the images by RAAG (10-step) are shown in the bottom.</h3>
</div>
<div id="results-carousel">
<div class="item">
<!-- Your image here -->
<center>
<img src="static/images/sd35_display.png" style="width: 80%;" alt="MY ALT TEXT"/>
</center>
<h4 class="subtitle is-6 has-text-centered" style="margin-top: 0.5em; margin-bottom: 1.5em;">
Stable Diffusion v3.5.
</h4>
</div>
<div id="results-carousel">
<div class="item">
<!-- Your image here -->
<center>
<img src="static/images/lumina_display.png" style="width: 80%;" alt="MY ALT TEXT"/>
</center>
<h4 class="subtitle is-6 has-text-centered" style="margin-top: 0.5em; margin-bottom: 1.5em;">
Lumina-Next.
</h4>
</div>
</div>
</div>
</div>
</section>
<!-- Image carousel -->
<!-- <section class="hero is-small">
<div class="hero-body">
<div class="columns is-centered has-text-centered">
<h2 class="title is-3">Results<br></h2>
</div>
<div class="container">
<div id="results-carousel" class="carousel results-carousel">
<div class="item">
<img src="static/images/img_res.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
These images are generated by EDM based on Stable Diffusion v3.5.
Specifically, the first column presents images generated from 8-step
LTC-Accel accelerated from the 12-step original sampling process,
while the second and the third column are generated from 8-step and
12-step original sampling process.
</h2>
</div>
<div class="item">
<img src="static/images/video_res.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
These images are generated by DeepCache based on Stable Diffusion v2.
Specifically, the first column presents images generated from the
50-step original sampling process,
while the second and the third column are generated from 50-step
DeepCache and 38-step DeepCache accelerated by LTC-Accel.
</h2>
</div>
<div class="item">
<img src="static/images/example3.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
3-step generation on Animated-Diff-Lightning
accelerated by LTC-Accel.
</h2>
</div>
<div class="item">
<img src="static/images/example4.png" alt="MY ALT TEXT"/>
<h2 class="subtitle has-text-centered">
These images are generated by Align Your Steps based on Stable Diffusion v1.5.
Specifically, the first column presents images generated from the
8-step sampling process accelerated by Align Your Steps and LTC-Accel,
while the second and the third column are generated from 10-step
Align Your Steps and original sampling process.
</h2>
</div>
</div>
</div>
</div>
</section> -->
<!-- End image carousel -->
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>
@article{zhu2025raag,
title={RAAG: Ratio Aware Adaptive Guidance},
author={Zhu, Shangwen and Peng, Qianyu and Hu, Yuting and Yang, Zhantao and Zhang, Han and Pu, Zhao and Feng, Ruili and Cheng, Fan},
journal={arXiv preprint arXiv:2508.03442},
year={2025}
}
</code></pre>
</div>
</section>
<!-- Youtube video -->
<!--
<section class="hero is-small is-light">
<div class="hero-body">
<div class="container">
<h2 class="title is-3">Video Presentation</h2>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<div class="publication-video">
<iframe src="https://www.youtube.com/embed/JkaxUblCGz0" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>
</div>
</div>
</div>
</div>
</div>
</section>
-->
<!-- End youtube video -->
<!-- Video carousel -->
<!--
<section class="hero is-small">
<div class="hero-body">
<div class="container">
<h2 class="title is-3">Another Carousel (11-Step LTC-Accel/Original Video)</h2>
<div id="results-carousel" class="carousel results-carousel">
<div class="gif-container">
<div class="item item-video1">
<img src="static/videos/animation16.gif" alt="Animation 16">
</div>
<div class="item item-video2">
<img src="static/videos/animation11.gif" alt="Animation 11">
</div>
</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" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>BibTex Code Here</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.
You are free to borrow the source code of this website, we just ask that you link back to this page in the footer. <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>