-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathindex.html
497 lines (452 loc) · 22.6 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
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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description" content="BenchLMM: Benchmarking Cross-style Visual Capability of Large Multimodal Models">
<meta name="keywords" content="multimodal chatbot">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>BenchLMM</title>
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/css/bulma.min.css">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/4.5.2/css/bootstrap.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/css/all.min.css">
<link rel="stylesheet" href="./static/css/index.css">
<link rel="icon" href="imgs/tittle_fig.png">
<link href="https://fonts.googleapis.com/icon?family=Material+Icons" rel="stylesheet">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/5.15.1/js/all.min.js"></script>
<script type="module" src="https://gradio.s3-us-west-2.amazonaws.com/3.27.0/gradio.js"></script>
</head>
<style>
.section {
margin-bottom: -30px; /* Adjust this value as needed to reduce the space */
}
.expandable-card .card-text-container {
max-height: 200px;
overflow-y: hidden;
position: relative;
}
.expandable-card.expanded .card-text-container {
max-height: none;
}
.expand-btn {
position: relative;
display: none;
background-color: rgba(255, 255, 255, 0.8);
/* margin-top: -20px; */
/* justify-content: center; */
color: #510c75;
border-color: transparent;
}
.expand-btn:hover {
background-color: rgba(200, 200, 200, 0.8);
text-decoration: none;
border-color: transparent;
color: #510c75;
}
.expand-btn:focus {
outline: none;
text-decoration: none;
}
.expandable-card:not(.expanded) .card-text-container:after {
content: "";
position: absolute;
bottom: 0;
left: 0;
width: 100%;
height: 90px;
background: linear-gradient(rgba(255, 255, 255, 0.2), rgba(255, 255, 255, 1));
}
.expandable-card:not(.expanded) .expand-btn {
margin-top: -40px;
}
.card-body {
padding-bottom: 5px;
}
.vertical-flex-layout {
justify-content: center;
align-items: center;
height: 100%;
display: flex;
flex-direction: column;
gap: 5px;
}
.figure-img {
max-width: 100%;
height: auto;
}
.adjustable-font-size {
font-size: calc(0.5rem + 2vw);
}
.chat-history {
flex-grow: 1;
overflow-y: auto;
/* overflow-x: hidden; */
padding: 5px;
border-bottom: 1px solid #ccc;
margin-bottom: 10px;
}
#gradio pre {
background-color: transparent;
}
</style>
<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">
<img src="imgs/tittle_fig.png" alt="BenchLMM_face" width="100">
<h1 class="title is-1 publication-title">BenchLMM: <span class="is-size-2"><span class="is-size-1">B</span>enchmarking <span class="is-size-1">C</span>ross-style <span class="is-size-1">V</span>isual <span class="is-size-1">C</span>apability of <span class="is-size-1">L</span>arge <span class="is-size-1">M</span>ultimodal <span class="is-size-1">M</span>odels</span></h1>
<div class="is-size-5 publication-authors">
<!-- First Group of 3 Authors -->
<div class="author-group">
<span class="author-block">
<a href="https://github.com/RizhaoCai" style="color:#f68946;font-weight:normal;">Rizhao Cai<sup>*1</sup></a>,
</span>
<span class="author-block">
<a href="https://github.com/ZiruiSongBest" style="color:#f68946;font-weight:normal;">Zirui Song<sup>*2,3</sup></a>,
</span>
</div>
<!-- Second Group of 3 Authors -->
<div class="author-group">
<span class="author-block">
<a href="https://dayan-guan.github.io/" style="color:#008AD7;font-weight:normal;">Dayan Guan<sup>1</sup></a>,
</span>
<span class="author-block">
<a href="https://zhenhaochenofficial.github.io" style="color:#f68946;font-weight:normal;">Zhenhao Chen<sup>4</sup></a>,
</span>
<span class="author-block">
<a href="https://github.com/dylanli-hang" style="color:#f68946;font-weight:normal;">Yaohang Li<sup>2,3</sup></a>,
</span>
<span class="author-block">
<a href="" style="color:#f68946;font-weight:normal;">Xing Luo<sup>5</sup></a>,
</span>
<span class="author-block">
<a href="https://www.researchgate.net/scientific-contributions/Chenyu-Yi-2169232914" style="color:#f68946;font-weight:normal;">Chenyu Yi<sup>1</sup></a>,
</span>
</div>
<!-- Third Group of 4 Authors -->
<div class="author-group">
<span class="author-block">
<a href="https://personal.ntu.edu.sg/eackot/" style="color:#f68946;font-weight:normal;">Alex Kot<sup>1</sup></a>,
</span>
</div>
</div>
<div class="is-size-5 publication-authors">
Nanyang Technological University<sup>1</sup><br>
</div>
<div class="is-size-5 publication-authors">
University of Technology Sydney<sup>2</sup><br>
</div>
<div class="is-size-5 publication-authors">
Northeastern University<sup>3</sup><br>
</div>
<div class="is-size-5 publication-authors">
Mohamed bin Zayed University of Artificial Intelligence<sup>4</sup><br>
</div>
<div class="is-size-5 publication-authors">
Zhejiang University<sup>5</sup><br>
</div>
<div class="is-size-6 publication-authors">
<span class="author-block"><sup>*</sup>Equally contributing first authors</span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<span class="link-block">
<a href="https://arxiv.org/abs/2312.02896" 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://github.com/AIFEG/BenchLMM" 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>
<span class="link-block">
<a href="https://huggingface.co/datasets/AIFEG/BenchLMM/" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="far fa-images"></i>
</span>
<span>Demo</span>
</a>
</span>
<span class="link-block">
<a href="https://huggingface.co/datasets/AIFEG/BenchLMM/blob/main/BenchGPT.zip" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-database"></i>
</span>
<span>Dataset</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<h4 class="subtitle has-text-justified">
Large Multimodal Models (LMMs) such as GPT-4V and LLaVA have shown remarkable capabilities in visual reasoning with common image styles. However, their robustness against diverse style shifts, crucial for practical applications, remains largely unexplored. In this paper, we propose a new benchmark, BenchLMM, to assess the robustness of LMMs against three different styles: artistic image style, imaging sensor style, and application style, where each style has five sub-styles. Utilizing BenchLMM, we comprehensively evaluate state-of-the-art LMMs and reveal: 1) LMMs generally suffer performance degradation when working with other styles; 2) An LMM performs better than another model in common style does not guarantee its superior performance in other styles; 3) LMMs' reasoning capability can be enhanced by prompting LMMs to predict the style first, based on which we propose a versatile and training-free method for improving LMMs; 4) An intelligent LMM is expected to interpret the causes of its errors when facing stylistic variations. We hope that our benchmark and analysis can shed new light on developing more intelligent and versatile LMMs. </h4>
</div>
</div>
</section>
<!-- <div style="text-align:center;">
<iframe width="1024" height="720" src="https://www.youtube.com/embed/0dZ4dlNIGTY" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
</div> -->
<!-- <section class="section" style="background-color:#efeff081">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">🔥Highlights</h2>
<div class="content has-text-justified">
<ol type="1">
<li><b>GLaMM Introduction</b>. We present the <span style="color: #997300;">Grounding Large Multimodal Model (GLaMM)</span>, the first-of-its-kind model capable of generating natural language responses that are seamlessly integrated with object segmentation masks. Unlike existing models, GLaMM accommodates both textual and optional visual prompts, facilitating enhanced multimodal user interaction. </li>
<br>
<li><b>Novel Task & Evaluation</b>. Recognizing the lack of standardized benchmarks for visually grounded conversations, we propose a new task of <span style="color: #997300;">Grounded Conversation Generation</span> (GCG). Alongside, we introduce a comprehensive evaluation protocol to measure the efficacy of models in this novel setting, filling a significant gap in the literature.</li>
<br>
<li><b>GranD Dataset Creation</b>. To facilitate model training and evaluation, we create <span style="color: #997300;">GranD - Grounding-anything Dataset</span>, a large-scale densely annotated dataset. Developed using an automatic annotation pipeline and verification criteria, it encompasses 7.5M unique concepts grounded in 810M regions. Additionally, we produce <span style="color: #997300;">GranD-f</span>, a high-quality dataset explicitly designed for the GCG task, by re-purposing existing open-source datasets.</li>
</ol>
</div>
</div>
</div>
</div>
</section> -->
<!--Model Arch-->
<section class="section"style="background-color:#efeff081">
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">Motivation</h2>
</div>
</div>
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
We propose BenchLMM to investigate the cross-style capability of Large Multimodal Models (LMMs). This includes their proficiency in tasks like analyzing images with diverse styles, processing images acquired from non-RGB cameras, and interpreting images sourced from specific application knowledge. In this section, we will elaborate on how we build the benchmark that encompasses these diverse styles.
</p>
</div>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/demo.png">
<figcaption>
This picture shows benchmark examples. For each specific field, one example image and its corresponding Q&A have been shown. Note that, for a simple presentation, the questions in Domestic Robot and Open Game have been simplified from multiple-choice format. Please see the Appendix for more examples and detailed questions.
</figcaption>
</figure>
</div>
</div>
</div>
</section>
<!-- Model Arch -->
<!--Dataset-->
<section id="Benchmark" class="section">
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">Evaluate Public LMMs</h2>
</div>
</div>
<!--Dataset Pipeline-->
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
In most existing works, LMMs are predominantly evaluated using images in the 'Photo' style, leading to a gap in understanding their performance across diverse artistic styles. We extend the evaluation scope by examining LMMs' performance with various artistic styles beyond the common 'Photo' style. Results, as detailed in Table, reveal a notable decline in LMMs' effectiveness when processing these artistic styles. This trend suggests a potential overfitting of LMMs to the 'Photo' style, highlighting their limited adaptability to varied artistic styles, a capability that humans typically possess. Interestingly, GPT-4V, despite being a robust commercial model, exhibits similar limitations in handling diverse styles.
</p>
</div>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/Table1.png">
<figcaption>
The figure shows the evaluations of public LMMs on cross-style BenchLMM.
Note that Average<sup>*</sup> represents the
% out-of-distribution (OOD)
average accuracy computed over five artistic-style benchmarks.
The best and the second best results are highlighted in and respectively.
All the numbers are presented in % and the full score is 100%.
</figcaption>
</figure>
</div>
</div>
</div>
<br>
<!-- Dataset Samples
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
Below we present more evaluation results.
</p>
</div>
</div>
</div> -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/Tabel2.png">
<figcaption>
Evaluations of public LMMs on cross-sensor BenchLMM.
</figcaption>
</figure>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/Table3.png">
<figcaption>
Evaluations of public LMMs on cross-task BenchLMM.
</figcaption>
</figure>
</div>
</div>
</div>
<br>
<br>
<!--Subsection-->
<section id="Examples" class="section">
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths">
<h2 class="title is-3">More examples of our benchmark</h2>
</div>
</div>
<!--Dataset Pipeline-->
<!-- <div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
In most existing works, LMMs are predominantly evaluated using images in the 'Photo' style, leading to a gap in understanding their performance across diverse artistic styles. We extend the evaluation scope by examining LMMs' performance with various artistic styles beyond the common 'Photo' style. Results, as detailed in Table, reveal a notable decline in LMMs' effectiveness when processing these artistic styles. This trend suggests a potential overfitting of LMMs to the 'Photo' style, highlighting their limited adaptability to varied artistic styles, a capability that humans typically possess. Interestingly, GPT-4V, despite being a robust commercial model, exhibits similar limitations in handling diverse styles.
</p>
</div>
</div>
</div> -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/P1-new_1.png">
<!-- <figcaption>
</figcaption> -->
</figure>
</div>
</div>
</div>
<br>
<!-- Dataset Samples
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<div class="content has-text-justified">
<p>
Below we present more evaluation results.
</p>
</div>
</div>
</div> -->
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/P2-new_1.png">
<!-- <figcaption>
Evaluations of public LMMs on cross-sensor BenchLMM.
</figcaption> -->
</figure>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-six-fifths" style="display: flex; align-items: flex-start; justify-content: center;">
<figure style="text-align: center;">
<img id="teaser" width="100%" src="imgs/P3-new_1.png">
<!-- <figcaption>
Evaluations of public LMMs on cross-task BenchLMM.
</figcaption> -->
</figure>
</div>
</div>
</div>
<br>
<br>
<!--Subsection-->
<!--Downstream-->
<!--Conv -->
<style>
#BibTeX {
margin-bottom: -80px; /* Adjust the negative margin as needed */
}
#Acknowledgement {
margin-top: -80px; /* Adjust the negative margin as needed */
}
</style>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>
@misc{cai2023benchlmm,
title={BenchLMM: Benchmarking Cross-style Visual Capability of Large Multimodal Models},
author={Rizhao Cai and Zirui Song and Dayan Guan and Zhenhao Chen and Xing Luo and Chenyu Yi and Alex Kot},
year={2023},
eprint={2312.02896},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
</code></pre>
</div>
</section>
<section class="section" id="Acknowledgement">
<div class="container is-max-desktop content">
<h2 class="title">Acknowledgement</h2>
<p>
This research is supported in part by the Rapid-Rich Object Search (ROSE) Lab of Nanyang Technological University and the NTU-PKU Joint Research Institute (a collaboration between NTU and Peking University that is sponsored by a donation from the Ng Teng Fong Charitable Foundation). We are deeply grateful to Yaohang Li from the University of Technology Sydney for his invaluable assistance in conducting the experiments, and to Jingpu Yang, Helin Wang, Zihui Cui, Yushan Jiang, Fengxian Ji, and Yuxiao Hang from NLULab@NEUQ (Northeastern University at Qinhuangdao, China) for their meticulous efforts in annotating the dataset. We also would like to thank Prof. Miao Fang (PI of NLULab@NEUQ) for his supervision and insightful suggestion during discussion on this project.
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>
</section>
</body>
</html>
<div style="text-align: center;">
<a href="https://github.com/AIFEG" target="_blank">
<img src="imgs/logos/AIFEG.png" width="100" height="100" alt="AIFEG Logo">
</a>
<a href="https://www.ntu.edu.sg/rose" target="_blank">
<img src="imgs/logos/ROSE.png" width="200" height="100" alt="ROSE Logo">
</a>
<a href="https://www.ntu.edu.sg" target="_blank">
<img src="imgs/logos/NTU.png" width="130" height="100" alt="AIFEG Logo">
</a>
<a href="https://www.uts.edu.au/" target="_blank">
<img src="imgs/logos/UTS.png" width="100" height="100" alt="UTS Logo">
</a>
<a href="https://www.NEU.edu.cn/" target="_blank">
<img src="imgs/logos/NEU.png" width="100" height="100" alt="NEU Logo">
</a>
<a href="https://mbzuai.ac.ae" target="_blank">
<img src="imgs/logos/MBZUAI_logo.png" width="360" height="85" alt="MBZUAI Logo">
</a>
<a href="https://www.ZJU.edu.cn/" target="_blank">
<img src="imgs/logos/ZJU_jmages.png" width="100" height="100" alt="NEU Logo">
</a>
</div>