-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
757 lines (695 loc) · 38.5 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
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
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
<!DOCTYPE HTML>
<html lang="en"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<title>Chen Wei</title>
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<link rel="stylesheet" type="text/css" href="stylesheet.css">
<link rel="icon" type="image/png" href="images_new/JHU_icon.jpg">
</head>
<body>
<table style="width:100%;max-width:800px;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:0px">
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr style="padding:0px">
<td style="padding:5%;width:63%;vertical-align:middle">
<p style="text-align:center">
<name>Chen Wei</name>
</p>
<p>
I am a final-year Computer Science Ph.D. student at Johns Hopkins University,
where I am fortunate to be advised by <a href="https://en.wikipedia.org/wiki/Bloomberg_Distinguished_Professorships">Bloomberg Distinguished Professor</a>
<a href="https://cs.jhu.edu/~ayuille/">Alan L. Yuille</a>. I received my B.S. with honors from <a href="https://english.pku.edu.cn/">Peking University</a>.
I spent great time at <a href="https://deepmind.google/">Google DeepMind</a>, <a href="https://ai.meta.com/research/">FAIR @ Meta</a>.
</p>
<p style="text-align:center">
<a href="mailto:[email protected]">weichen3012 [at] gmail [dot] com</a>  / 
<a href="https://scholar.google.com/citations?hl=en&user=LHQGpBUAAAAJ">Google Scholar</a>  / 
<a href="https://twitter.com/ChenWei23787135">Twitter</a>
</p>
</td>
<td style="padding:2.5%;width:30%;max-width:30%;">
<a href="images_new/profile_2.jpg"><img style="width:100%;max-width:100%;border-radius:50%;" alt="profile photo" src="images_new/profile_2.jpg" class="hoverZoomLink"></a>
</td>
</tr>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding-top:20px;padding-bottom:20px;width:100%;vertical-align:middle">
<heading>Research Interest</heading>
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
My research lies at the intersection of <b>machine learning</b> and <b>computer vision</b>, driven by the goal of developing flexible and general-purpose intelligent systems.
My research has primarily focused on two critical steps toward this goal, (1) to encode enormously complex raw visual signals to structured representations that are easier to understand and process for machine, and (2) to connect visual representations to higher-order cognition and commonsense reasoning about the world:
</td>
</tr>
<tr>
<td width="100%" valign="middle">
<ul>
<li><b>Generative Understanding</b>: <a href="https://weichen582.github.io/diffmae.html">constructing</a> and <a href="https://dediffusion.github.io/">learning from</a> generative models of the world.</li>
</ul>
<ul>
<li><b>Vision and Language</b>: <a href="https://dediffusion.github.io/">connecting</a> <a href="https://arxiv.org/abs/2211.11446">visual representations</a> to commonsense knowledge base like large language models to achieve higher-order cognition.</li>
</ul>
<ul>
<li><b>Self-Supervised Learning</b>: discovering natural structures in the raw data to obtain <a href="https://arxiv.org/abs/2010.02217">structured representations</a> that are scalable to wide domains in <a href="https://arxiv.org/abs/2111.07832">images</a> and <a href="https://arxiv.org/abs/2112.09133">videos</a>, and applicable to a <a href="https://arxiv.org/abs/2306.00989">variety </a> of <a href="https://arxiv.org/abs/2203.11709">tasks</a>.</li>
</ul>
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
<span style="color:red">I am on the job market for 2024!</span> Would love to chat more if you are interested. I am also happy to give talks on my research in related seminars.
</td>
</tr>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding-top:20px;padding-bottom:20px;width:100%;vertical-align:middle">
<heading>News</heading>
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[12/01/2023] Gave a talk at Cornell Tech, hosted by <a href="https://www.cs.cornell.edu/~snavely/">Prof. Noah Snavely</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
<span style="color:red">[11/01/2023] Check out the <a href="https://dediffusion.github.io/">cool examples</a> from <a href="https://arxiv.org/abs/2311.00618">De-Diffusion Makes Text a Strong Cross-Modal Interface</a>!</span>
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[10/31/2023] Gave a talk at <a href="https://sites.google.com/view/visionseminar">MIT Visual Computing Seminar</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[10/27/2023] Gave a talk at <a href="https://cvl.psych.ucla.edu/">The Computational Vision and Learning Lab @ UCLA</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[08/29/2023] I have been selected as one of the <a href="https://eecsrisingstars2023.cc.gatech.edu/">EECS Rising Stars 2023</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[07/16/2023] <a href="https://weichen582.github.io/diffmae.html">DiffMAE</a> and <a href="https://arxiv.org/abs/2211.11446">SMAUG</a> accepted to ICCV 2023.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[04/23/2023] <a href="https://github.com/facebookresearch/hiera">Hiera</a> accepted to ICML 2023 as oral presentation.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[04/06/2023] Excited to release our new work on reformulating <a href="https://arxiv.org/abs/2304.03283">Diffusion Models as Masked Autoencoders</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[10/19/2022] Proud to be an <a href="https://eccv2022.ecva.net/program/outstanding-reviewers/">ECCV 2022 outstanding reviewer</a>.
</td>
</tr>
<tr>
<td style="padding-bottom:10px;width:100%;vertical-align:middle">
[09/07/2022] Passed my Graduate Board Oral exam and now a PhD candidate :)
</td>
</tr>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding-top:20px;padding-bottom:20px;width:100%;vertical-align:middle">
<heading>Publications</heading>
</td>
</tr>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;">
<tr onmouseout="dediffusion_stop()" onmouseover="dediffusion_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-top:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='dediffusion_image'><img style="width: 120%; height: auto;" src='images_new/dediffusion_text.png'></div>
<img style="width: 120%; height: auto;" src='images_new/corgi.jpeg'>
</div>
<script type="text/javascript">
function dediffusion_start() {
document.getElementById('dediffusion_image').style.opacity = "1";
}
function dediffusion_stop() {
document.getElementById('dediffusion_image').style.opacity = "0";
}
dediffusion_stop()
</script>
</td>
<td style="padding:40px;padding-top:70px;width:70%;vertical-align:top">
<a href="https://arxiv.org/abs/2311.00618">
<papertitle> De-Diffusion Makes Text a Strong Cross-Modal Interface </papertitle>
</a>
<br>
<b>Chen Wei</b>,
<a href="https://www.cs.jhu.edu/~cxliu/", class="author">Chenxi Liu</a>,
<a href="https://www.cs.jhu.edu/~syqiao/", class="author">Siyuan Qiao</a>,
<a href="https://zhishuai.xyz/", class="author">Zhishuai Zhang</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://jiahuiyu.com/", class="author", class="author">Jiahui Yu</a>
<br>
<em>arXiv </em>, 2023
<br>
<a href="https://arxiv.org/abs/2311.00618">arXiv</a> /
<a href="https://dediffusion.github.io/">project page</a>
<br>
</td>
</tr>
<tr onmouseout="i2a_stop()" onmouseover="i2a_start()">
<td style="padding:40px;padding-left:50px;padding-top:15px;padding-bottom:100px;vertical-align:middle">
<div class="one">
<div class="two" id='i2a_image'><img style="width: 110%; height: auto;" src='images_new/i2a_2.png'></div>
<img style="width: 110%; height: auto;" src='images_new/i2a_1.png'>
</div>
<script type="text/javascript">
function i2a_start() {
document.getElementById('i2a_image').style.opacity = "1";
}
function i2a_stop() {
document.getElementById('i2a_image').style.opacity = "0";
}
i2a_stop()
</script>
</td>
<td style="padding:40px;padding-top:15px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2311.15551">
<papertitle> Instruct2Attack: Language-Guided Semantic Adversarial Attacks </papertitle>
</a>
<br>
<a href="https://joellliu.github.io/", class="author">Jiang Liu</a>,
<b>Chen Wei</b>,
<a href="https://aiem.jhu.edu/people/yuxiang-guo/", class="author">Yuxiang Guo</a>,
<a href="https://heng14.github.io/", class="author">Heng Yu</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://www.cs.umd.edu/~sfeizi/", class="author", class="author">Soheil Feizi</a>,
<a href="https://samuel930930.github.io/", class="author", class="author">Chun Pong Lau</a>,
<a href="https://engineering.jhu.edu/ece/faculty/rama-chellappa/", class="author", class="author">Rama Chellappa</a>
<br>
<em>arXiv </em>, 2023
<br>
<a href="https://arxiv.org/abs/2311.15551">arXiv</a>
<br>
</td>
</tr>
<tr onmouseout="diffmae_stop()" onmouseover="diffmae_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-top:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='diffmae_image'><img style="width: 120%; height: auto;" src='diffmae/interpolation/input.png'></div>
<img style="width: 120%; height: auto;" src='diffmae/interpolation/2.png'>
</div>
<script type="text/javascript">
function diffmae_start() {
document.getElementById('diffmae_image').style.opacity = "1";
}
function diffmae_stop() {
document.getElementById('diffmae_image').style.opacity = "0";
}
diffmae_stop()
</script>
</td>
<td style="padding:40px;padding-top:70px;width:70%;vertical-align:top">
<a href="https://arxiv.org/abs/2304.03283">
<papertitle> Diffusion Models as Masked Autoencoders </papertitle>
</a>
<br>
<b>Chen Wei</b>,
<a href="https://karttikeya.github.io/", class="author">Karttikeya Mangalam</a>,
<a href="http://www.cs.cmu.edu/~poyaoh/", class="author">Po-Yao Huang</a>,
<a href="https://lyttonhao.github.io/", class="author">Yanghao Li</a>,
<a href="https://haoqifan.github.io/", class="author">Haoqi Fan</a>,
<a href="https://howardhsu.github.io/", class="author">Hu Xu</a>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<a href="https://cihangxie.github.io/", class="author">Cihang Xie</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://feichtenhofer.github.io/", class="author", class="author">Christoph Feichtenhofer</a>
<br>
<em>ICCV </em>, 2023
<br>
<a href="https://arxiv.org/abs/2304.03283">arXiv</a> /
<a href="https://weichen582.github.io/diffmae.html">project page</a> /
<a href="https://www.marktechpost.com/2023/04/11/a-new-ai-research-integrates-masking-into-diffusion-models-to-develop-diffusion-masked-autoencoders-diffmae-a-self-supervised-framework-designed-for-recognizing-and-generating-images-and-videos/">Marktechpost</a>
<br>
</td>
</tr>
<tr>
<td style="padding:40px;padding-left:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='smaug_image'><img style="width: 150%; height: auto;" src='images_new/smaug.png'></div>
</div>
</td>
<td style="padding:40px;padding-top:30px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2211.11446">
<papertitle> SMAUG: Sparse Masked Autoencoder for Efficient Video-Language Pre-Training </papertitle>
</a>
<br>
<a href="https://scholar.google.com/citations?user=0WFC2w0AAAAJ", class="author">Yuanze Lin</a>,
<strong>Chen Wei</strong>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://cihangxie.github.io/", class="author">Cihang Xie</a>
<br>
<em>ICCV </em>, 2023
<br>
<a href="https://arxiv.org/abs/2211.11446">arXiv</a>
<br>
</td>
</tr>
<tr onmouseout="hiera_stop()" onmouseover="hiera_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-left:40px;padding-top:0px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='hiera_image'><img style="width: 110%; height: auto;" src='images_new/hiera_before.png'></div>
<img style="width: 110%; height: auto;" src='images_new/hiera_after.png'>
</div>
<script type="text/javascript">
function hiera_start() {
document.getElementById('hiera_image').style.opacity = "1";
}
function hiera_stop() {
document.getElementById('hiera_image').style.opacity = "0";
}
hiera_stop()
</script>
</td>
<td style="padding:40px;padding-top:50px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2203.11709">
<papertitle> Hiera: A Hierarchical Vision Transformer without the Bells-and-Whistles </papertitle>
</a>
<br>
<a href="https://scholar.google.com/citations?user=4LWx24UAAAAJ", class="author">Chaitanya Ryali</a>*,
<a href="https://scholar.google.com/citations?user=aMpbemkAAAAJ", class="author">Yuan-Ting Hu</a>*,
<a href="https://dbolya.github.io/", class="author">Daniel Bolya</a>*,
<strong>Chen Wei</strong>,
<a href="https://haoqifan.github.io/", class="author">Haoqi Fan</a>,
<a href="http://www.cs.cmu.edu/~poyaoh/", class="author">Po-Yao Huang</a>,
<a href="https://scholar.google.com/citations?hl=en&user=Qwm6ZOYAAAAJ", class="author">Vaibhav Aggarwal</a>,
<a href="https://scholar.google.com/citations?hl=en&user=42v1i_YAAAAJ", class="author">Arkabandhu Chowdhury</a>,
<a href="https://omidpoursaeed.github.io/", class="author">Omid Poursaeed</a>,
<a href="https://faculty.cc.gatech.edu/~judy/", class="author">Judy Hoffman</a>,
<a href="https://people.eecs.berkeley.edu/~malik/", class="author">Jitendra Malik</a>,
<a href="https://lyttonhao.github.io/", class="author">Yanghao Li</a>*,
<a href="https://feichtenhofer.github.io/", class="author">Christoph Feichtenhofer</a>*
<br>
<em>ICML </em>, 2023 <strong>Oral</strong>
<br>
<a href="https://arxiv.org/abs/2306.00989/">arXiv</a> /
<a href="https://github.com/facebookresearch/hiera">code</a>
<br>
</td>
</tr>
<tr>
<td style="padding:40px;padding-left:20px;padding-bottom:0px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='dmae_image'><img style="width: 130%; height: auto;" src='images_new/dmae.png'></div>
</div>
</td>
<td style="padding:40px;padding-top:10px;padding-bottom:0px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2208.12256">
<papertitle> Masked Autoencoders Enable Efficient Knowledge Distillers </papertitle>
</a>
<br>
<a class="author">Yutong Bai</a>,
<a href="https://zeyuwang.netlify.app/", class="author">Zeyu Wang</a>,
<a href="https://lambert-x.github.io/", class="author">Junfei Xiao</a>,
<strong>Chen Wei</strong>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://yuyinzhou.github.io/", class="author">Yuyin Zhou</a>,
<a href="https://cihangxie.github.io/", class="author">Cihang Xie</a>
<br>
<em>CVPR </em>, 2023
<br>
<a href="https://arxiv.org/abs/2208.12256">arXiv</a> /
<a href="https://github.com/UCSC-VLAA/DMAE">code</a>
<br>
</td>
</tr>
<tr onmouseout="cp2_stop()" onmouseover="cp2_start()">
<td style="padding:40px;padding-left:30px;padding-top:20px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='cp2_image'><img style="width: 120%; height: auto;" src='images_new/cp2_after.png''></div>
<img style="width: 120%; height: auto;" src='images_new/cp2_before.png'>
</div>
<script type="text/javascript">
function cp2_start() {
document.getElementById('cp2_image').style.opacity = "1";
}
function cp2_stop() {
document.getElementById('cp2_image').style.opacity = "0";
}
cp2_stop()
</script>
</td>
<td style="padding:40px;padding-top:40px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2203.11709">
<papertitle> CP2: Copy-Paste Contrastive Pretraining for Semantic Segmentation </papertitle>
</a>
<br>
<a href="https://wangf3014.github.io/home/", class="author">Feng Wang</a>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<strong>Chen Wei</strong>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://shenwei1231.github.io/", class="author">Wei Shen</a>
<br>
<em>ECCV </em>, 2022
<br>
<a href="https://arxiv.org/abs/2203.11709">arXiv</a> /
<a href="https://github.com/wangf3014/CP2">code</a>
<br>
</td>
</tr>
<tr>
<td style="padding:40px;padding-left:40px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='li_image'><img style="width: 120%; height: auto;" src='images_new/eccv2023li.png'></div>
</div>
</td>
<td style="padding:40px;padding-top:50px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2205.01721">
<papertitle> In Defense of Image Pre-Training for Spatiotemporal Recognition</papertitle>
</a>
<br>
<a href="https://xhl-video.github.io/xianhangli/", class="author">Xianhang Li</a>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<strong>Chen Wei</strong>,
<a href="https://meijieru.com/", class="author">Jieru Mei</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://yuyinzhou.github.io/", class="author">Yuyin Zhou</a>,
<a href="https://cihangxie.github.io/", class="author">Cihang Xie</a>
<br>
<em>ECCV </em>, 2022
<br>
<a href="https://arxiv.org/abs/2205.01721">arXiv</a> /
<a href="https://github.com/UCSC-VLAA/Image-Pretraining-for-Video">code</a>
<br>
</td>
</tr>
<tr onmouseout="maskfeat_stop()" onmouseover="maskfeat_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-top:0px;padding-left:30px;width:30%;vertical-align:middle">
<br></br>
<div class="one">
<div class="two" id='maskfeat_image'><img style="width: 120%; height: auto;" src='images_new/maskfeat_after.png'></div>
<img style="width: 120%; height: auto;" src='images_new/maskfeat_before.png'>
</div>
<script type="text/javascript">
function maskfeat_start() {
document.getElementById('maskfeat_image').style.opacity = "1";
}
function maskfeat_stop() {
document.getElementById('maskfeat_image').style.opacity = "0";
}
maskfeat_stop()
</script>
</td>
<td style="padding:40px;padding-top:60px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/pdf/2112.09133.pdf">
<papertitle> Masked Feature Prediction for Self-Supervised Visual Pre-Training </papertitle>
</a>
<br>
<strong>Chen Wei*</strong>,
<a href="https://haoqifan.github.io/", class="author">Haoqi Fan</a>,
<a href="https://vcl.ucsd.edu/~sxie/", class="author">Saining Xie</a>,
<a href="https://chaoyuan.org/", class="author">Chao-Yuan Wu</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="https://feichtenhofer.github.io/", class="author">Christoph Feichtenhofer*</a>
<br>
<em>CVPR </em>, 2022
<br>
<a href="https://arxiv.org/pdf/2112.09133.pdf">arXiv</a> /
<a href="https://github.com/facebookresearch/SlowFast">code at pySlowFast</a>
<br>
<br>
<a class="highlight", href="https://www.paperdigest.org/2023/04/most-influential-cvpr-papers-2023-04/">Most Influential CVPR 2023 Papers</a>
</td>
</tr>
<tr onmouseout="ibot_stop()" onmouseover="ibot_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-left:30px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='ibot_image'><img style="width: 120%; height: auto;" src='images_new/ibot_after.png'></div>
<img style="width: 120%; height: auto;" src='images_new/ibot_before.png'>
</div>
<script type="text/javascript">
function ibot_start() {
document.getElementById('ibot_image').style.opacity = "1";
}
function ibot_stop() {
document.getElementById('ibot_image').style.opacity = "0";
}
ibot_stop()
</script>
</td>
<td style="padding:40px;padding-top:80px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/pdf/2111.07832.pdf">
<papertitle> iBOT: Image BERT Pre-Training with Online Tokenizer </papertitle>
</a>
<br>
<a href="https://shallowtoil.github.io/", class="author">Jinghao Zhou</a>,
<strong>Chen Wei</strong>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<a href="https://shenwei1231.github.io/", class="author">Wei Shen</a>,
<a href="https://cihangxie.github.io/", class="author">Cihang Xie</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="http://www.taokong.org/", class="author">Tao Kong</a>
<br>
<em>ICLR </em>, 2022
<br>
<a href="https://arxiv.org/pdf/2111.07832.pdf">arXiv</a> /
<a href="https://github.com/bytedance/ibot">code</a> /
<a href="https://medium.com/syncedreview/meet-ibot-a-masked-image-modelling-framework-that-enables-bert-like-pretraining-for-vision-da01002115e7">press</a>
<br>
<br>
<br>
<a class="highlight", href="https://dinov2.metademolab.com/">Improved and Scaled up to foundation model DINOv2</a>
</td>
</tr>
<tr onmouseout="crest_stop()" onmouseover="crest_start()" bgcolor="#ffffd0">
<td style="padding:40px;padding-left:30px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='crest_image'><img style="width: 120%; height: auto;" src='images_new/crest_after.png'></div>
<img style="width: 120%; height: auto;" src='images_new/crest_before.png'>
</div>
<script type="text/javascript">
function crest_start() {
document.getElementById('crest_image').style.opacity = "1";
}
function crest_stop() {
document.getElementById('crest_image').style.opacity = "0";
}
crest_stop()
</script>
</td>
<td style="padding:40px;padding-top:60px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/pdf/2102.09559.pdf">
<papertitle> CReST: A Class-Rebalancing Self-Training Framework for Imbalanced Semi-Supervised Learning </papertitle>
</a>
<br>
<strong>Chen Wei</strong>,
<a href="https://sites.google.com/site/kihyuksml/", class="author">Kihyuk Sohn</a>,
<a href="https://scholar.google.com/citations?user=Vu3vH2sAAAAJ&hl=en", class="author">Clayton Mellina</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>,
<a href="http://users.umiacs.umd.edu/~fyang/", class="author">Fan Yang</a>
<br>
<em>CVPR </em>, 2021
<br>
<a href="https://arxiv.org/pdf/2102.09559.pdf">arXiv</a> /
<a href="https://github.com/google-research/crest">code</a> /
<a href="files/CReST_poster.pdf">poster</a> /
<a href="https://www.youtube.com/watch?v=cbXRfYBIt0w">video</a>
<br>
</td>
</tr>
<tr onmouseout="co2_stop()" onmouseover="co2_start()">
<td style="padding:40px;padding-left:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='co2_image'><img style="width: 130%; height: auto;" src='images_new/co2_after.png'></div>
<img style="width: 130%; height: auto;" src='images_new/co2_before.png'>
</div>
<script type="text/javascript">
function co2_start() {
document.getElementById('co2_image').style.opacity = "1";
}
function co2_stop() {
document.getElementById('co2_image').style.opacity = "0";
}
co2_stop()
</script>
</td>
<td style="padding:40px;padding-top:65px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/2010.02217">
<papertitle> CO2: Consistent Contrast for Unsupervised Visual Representation Learning</papertitle>
</a>
<br>
<strong>Chen Wei</strong>,
<a href="https://csrhddlam.github.io/", class="author">Huiyu Wang</a>,
<a href="https://shenwei1231.github.io/", class="author">Wei Shen</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>
<br>
<em>ICLR </em>, 2021
<br>
<a href="https://arxiv.org/abs/2010.02217">arXiv</a> /
<a href="https://openreview.net/forum?id=U4XLJhqwNF1">Open Review</a> /
<a href="https://iclr.cc/virtual/2021/poster/3354">video</a>
<br>
</td>
</tr>
<tr onmouseout="puzzle_stop()" onmouseover="puzzle_start()">
<td style="padding:40px;padding-left:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='puzzle_image'><img style="width: 130%; height: auto;" src='images_new/puzzle_after.png'></div>
<img style="width: 130%; height: auto;" src='images_new/puzzle_before.png'>
</div>
<script type="text/javascript">
function puzzle_start() {
document.getElementById('puzzle_image').style.opacity = "1";
}
function puzzle_stop() {
document.getElementById('puzzle_image').style.opacity = "0";
}
puzzle_stop()
</script>
</td>
<td style="padding:40px;;padding-top:70px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/1812.00329">
<papertitle> Iterative Reorganization with Weak Spatial Constraints: </br> Solving Arbitrary Jigsaw Puzzles for Unsupervised Representation Learning</papertitle>
</a>
<br>
<strong>Chen Wei</strong>,
<a href="http://lingxixie.com/", class="author">Lingxi Xie</a>,
<a href="https://tonghelen.github.io/", class="author">Xutong Ren</a>,
<a href="http://yingdaxia.github.io/", class="author">Yingda Xia</a>,
<a href="https://scholar.google.com/citations?user=V-AVamQAAAAJ&hl", class="author">Chi Su</a>,
<a href="http://www.icst.pku.edu.cn/struct/people/liujiaying.html", class="author">Jiaying Liu</a>,
<a href="http://www.cs.utsa.edu/~qitian/", class="author">Qi Tian</a>,
<a href="http://cs.jhu.edu/~ayuille/", class="author">Alan Yuille</a>
<br>
<em>CVPR</em>, 2019
<br>
<a href="https://arxiv.org/abs/1812.00329">arXiv</a> /
<a href="https://github.com/weichen582/Unsupervised-Visual-Recognition-by-Solving-Arbitrary-Puzzles">code</a>
<br>
</td>
</tr>
<tr onmouseout="retinex_stop()" onmouseover="retinex_start()">
<td style="padding:40px;padding-left:15px;width:30%;vertical-align:middle">
<div class="one">
<div class="two" id='retinex_image'><img style="width: 130%; height: auto;" src='images_new/retinex_after.png'></div>
<img style="width: 130%; height: auto;" src='images_new/retinex_before.png'>
</div>
<script type="text/javascript">
function retinex_start() {
document.getElementById('retinex_image').style.opacity = "1";
}
function retinex_stop() {
document.getElementById('retinex_image').style.opacity = "0";
}
retinex_stop()
</script>
</td>
<td style="padding:40px;padding-top:105px;width:70%;vertical-align:middle">
<a href="https://arxiv.org/abs/1808.04560">
<papertitle> Deep Retinex Decomposition for Low-Light Enhancement </papertitle>
</a>
<br>
<strong>Chen Wei*</strong>,
<a href="https://daooshee.github.io/website/", class="author">Wenjing Wang</a>*,
<a href="https://flyywh.github.io/index.html", class="author">Wenhan Yang</a>,
<a href="http://www.icst.pku.edu.cn/struct/people/liujiaying.html", class="author">Jiaying Liu</a>
<br>
<em>BMVC</em>, 2018 <strong>Oral</strong>
<br>
<a href="https://arxiv.org/abs/1808.04560">arXiv</a> /
<a href="https://github.com/weichen582/RetinexNet">code</a> /
<a href="https://daooshee.github.io/BMVC2018website/">project page & dataset</a>
<br>
<br>
<a class="highlight", href="https://scholar.google.com/citations?hl=en&vq=eng_computervisionpatternrecognition&view_op=list_hcore&venue=T_DfB2ikUbwJ.2023">Most Cited BMVC Papers Over the Last Five Years</a>
</td>
</tr>
<tr onmouseout="glad_stop()" onmouseover="glad_start()">
<td style="padding:40px;padding-left:20px;padding-top:20px;width:30%;vertical-align:middle">
<br></br>
<div class="one">
<div class="two" id='glad_image'><img style="width: 130%; height: auto;" src='images_new/glad_after.png'></div>
<img style="width: 130%; height: auto;" src='images_new/glad_before.png'>
</div>
<script type="text/javascript">
function glad_start() {
document.getElementById('glad_image').style.opacity = "1";
}
function glad_stop() {
document.getElementById('glad_image').style.opacity = "0";
}
glad_stop()
</script>
</td>
<td style="padding:40px;width:70%;vertical-align:middle">
<a href="https://ieeexplore.ieee.org/document/8373911">
<papertitle> GLADNet: Low-Light Enhancement Network with Global Awareness</papertitle>
</a>
<br>
<a href="https://daooshee.github.io/website/", class="author">Wenjing Wang</a>*,
<strong>Chen Wei*</strong>,
<a href="https://flyywh.github.io/index.html", class="author">Wenhan Yang</a>,
<a href="http://www.icst.pku.edu.cn/struct/people/liujiaying.html", class="author">Jiaying Liu</a>
<br>
<em>FG Workshop</em>, 2018
<br>
<a href="http://39.96.165.147/Pub%20Files/2018/wwj_fg2018.pdf">PDF</a> /
<a href="https://github.com/weichen582/GLADNet">code</a> /
<a href="https://daooshee.github.io/fgworkshop18Gladnet/">project page</a>
<br>
</td>
</tr>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding-top:20px;padding-bottom:20px;width:100%;vertical-align:middle">
<heading>Experience</heading>
</td>
</tr>
<td width="100%" valign="middle">
<ul>
<li>Student Researcher, Google DeepMind, Summer 2023</li>
</ul>
<ul>
<li>Research Intern, Facebook AI Research (Meta AI), Summer 2022</li>
</ul>
<ul>
<li>Research Intern, Facebook AI Research, Summer 2021</li>
</ul>
<ul>
<li>Research Intern, Google Cloud AI, Summer 2020</li>
</ul>
<ul>
<li>Research Intern, Johns Hopkins University, Summer 2018</li>
</ul>
<ul>
<li>Engineering Practium Intern, Google, Summer 2017</li>
</ul>
<ul>
<li>Research Intern, <a href="http://www.icst.pku.edu.cn/struct/">STRUCT</a> Lab, Peking University, 2017 - 2019</li>
</ul>
</td>
</tr>
</table>
<table style="width:100%;border:0px;border-spacing:0px;border-collapse:separate;margin-right:auto;margin-left:auto;"><tbody>
<tr>
<td style="padding:0px">
<br>
<p style="text-align:right;font-size:small;">
Last update: Dec. 2023      <a href="https://jonbarron.info/">Template</a>
</p>
</td>
</tr>
</tbody></table>
</td>
</tr>
</table>
</body>
</html>