forked from jonbarron/website
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
executable file
·1218 lines (1161 loc) · 67.2 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
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta name=viewport content=“width=800”>
<meta name="generator" content="HTML Tidy for Linux/x86 (vers 11 February 2007), see www.w3.org">
<style type="text/css">
/* Color scheme stolen from Sergey Karayev */
a {
color: #1772d0;
text-decoration: none;
}
a:focus,
a:hover {
color: #f09228;
text-decoration: none;
}
body,
td,
th,
tr,
p,
a {
font-family: 'Lato', Verdana, Helvetica, sans-serif;
font-size: 14px
}
strong {
font-family: 'Lato', Verdana, Helvetica, sans-serif;
font-size: 14px;
}
heading {
font-family: 'Lato', Verdana, Helvetica, sans-serif;
font-size: 22px;
}
papertitle {
font-family: 'Lato', Verdana, Helvetica, sans-serif;
font-size: 14px;
font-weight: 700
}
name {
font-family: 'Lato', Verdana, Helvetica, sans-serif;
font-size: 32px;
}
.one {
width: 160px;
height: 160px;
position: relative;
}
.two {
width: 160px;
height: 160px;
position: absolute;
transition: opacity .2s ease-in-out;
-moz-transition: opacity .2s ease-in-out;
-webkit-transition: opacity .2s ease-in-out;
}
.fade {
transition: opacity .2s ease-in-out;
-moz-transition: opacity .2s ease-in-out;
-webkit-transition: opacity .2s ease-in-out;
}
span.highlight {
background-color: #ffffd0;
}
</style>
<link rel="icon" type="image/png" href="seal_icon.png">
<title>Jon Barron</title>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<link href='http://fonts.googleapis.com/css?family=Lato:400,700,400italic,700italic' rel='stylesheet' type='text/css'>
</head>
<body>
<table width="800" border="0" align="center" cellspacing="0" cellpadding="0">
<tr>
<td>
<table width="100%" align="center" border="0" cellspacing="0" cellpadding="20">
<tr>
<td width="67%" valign="middle">
<p align="center">
<name>Jon Barron Jon Barron</name>
</p>
<p>I am a senior research scientist at <a href="https://research.google.com/">Google Research</a>, where I work on computer vision and computational photography. At Google I've worked on <a href="http://googleresearch.blogspot.com/2014/04/lens-blur-in-new-google-camera-app.html">Lens Blur</a>, <a href="http://googleresearch.blogspot.com/2014/10/hdr-low-light-and-high-dynamic-range.html">HDR+</a>, <a href="https://www.google.com/get/cardboard/jump/">Jump</a>, <a href="https://ai.googleblog.com/2017/10/portrait-mode-on-pixel-2-and-pixel-2-xl.html">Portrait Mode</a>, and <a href="https://www.youtube.com/watch?v=JSnB06um5r4">Glass</a>.
</p>
<p>
I did my PhD at <a href="http://www.eecs.berkeley.edu/">UC Berkeley</a>, where I was advised by <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a> and funded by the <a href="http://www.nsfgrfp.org/">NSF GRFP</a>. I've spent time at <a href="https://en.wikipedia.org/wiki/Google_X">Google[x]</a>, <a href="http://groups.csail.mit.edu/vision/welcome/">MIT CSAIL</a>, <a href="http://www.captricity.com/">Captricity</a>, <a href="https://www.nasa.gov/ames">NASA Ames</a>, <a href="http://www.google.com/">Google NYC</a>, the <a href="http://mrl.nyu.edu/">NYU MRL</a>, <a href="http://www.nibr.com/">Novartis</a>, and <a href="http://www.astrometry.net/">Astrometry.net</a>. I did my bachelors at the <a href="http://cs.toronto.edu">University of Toronto</a>.
</p>
<p align=center>
<a href="mailto:jonbarron@gmail.com">Email</a>  / 
<a href="JonBarron-CV.pdf">CV</a>  / 
<a href="JonBarron-bio.txt">Biography</a>  / 
<a href="https://scholar.google.com/citations?hl=en&user=jktWnL8AAAAJ">Google Scholar</a>  / 
<a href="http://www.linkedin.com/in/jonathanbarron/"> LinkedIn </a>
</p>
</td>
<td width="33%">
<img src="JonBarron_circle.jpg">
</td>
</tr>
</table>
<table width="100%" align="center" border="0" cellspacing="0" cellpadding="20">
<tr>
<td width="100%" valign="middle">
<heading>Research</heading>
<p>
I'm interested in computer vision, machine learning, statistics, optimization, image processing, virtual reality, and computational photography. Much of my research is about inferring the physical world (shape, depth, motion, paint, light, colors, etc) from images. I have also worked in astronomy and biology. Representative papers are <span class="highlight">highlighted</span>.
</p>
</td>
</tr>
</table>
<table width="100%" align="center" border="0" cellspacing="0" cellpadding="20">
<tr onmouseout="portrait_stop()" onmouseover="portrait_start()">
<td width="25%">
<div class="one">
<div class="two" id='portrait_image'><img src='portrait_after.png'></div>
<img src='portrait_before.png'>
</div>
<script type="text/javascript">
function portrait_start() {
document.getElementById('portrait_image').style.opacity = "1";
}
function portrait_stop() {
document.getElementById('portrait_image').style.opacity = "0";
}
portrait_stop()
</script>
</td>
<td valign="top" width="75%">
<a href="https://drive.google.com/file/d/13i6DlS9UhGVKmwslLUFnKBwdxFRVQeQj/view?usp=sharing">
<papertitle>Synthetic Depth-of-Field with a Single-Camera Mobile Phone</papertitle>
</a>
<br>
<a href="http://people.csail.mit.edu/nwadhwa/">Neal Wadhwa</a>,
<a href="http://rahuldotgarg.appspot.com/">Rahul Garg</a>,
<a href="http://graphics.stanford.edu/~dejacobs/">David E. Jacobs</a>, Bryan E. Feldman, Nori Kanazawa, Robert Carroll,
<a href="http://www.cs.cmu.edu/~ymovshov/">Yair Movshovitz-Attias</a>,
<strong>Jonathan T. Barron</strong>, Yael Pritch,
<a href="http://graphics.stanford.edu/~levoy/">Marc Levoy</a>
<br>
<em>SIGGRAPH</em>, 2018
<br>
<a href="https://arxiv.org/abs/1806.04171">arxiv</a> /
<a href="https://ai.googleblog.com/2017/10/portrait-mode-on-pixel-2-and-pixel-2-xl.html">blog post</a> /
<a href="Wadhwa2018.bib">bibtex</a>
<p></p>
<p>Dual pixel cameras and semantic segmentation algorithms can be used for shallow depth of field effects.</p>
<p>This system is the basis for "Portrait Mode" on the Google Pixel 2 smartphones</p>
</td>
</tr>
<tr onmouseout="aperture_stop()" onmouseover="aperture_start()">
<td width="25%">
<div class="one">
<div class="two" id='aperture_image'><img src='aperture_after.jpg'></div>
<img src='aperture_before.jpg'>
</div>
<script type="text/javascript">
function aperture_start() {
document.getElementById('aperture_image').style.opacity = "1";
}
function aperture_stop() {
document.getElementById('aperture_image').style.opacity = "0";
}
aperture_stop()
</script>
</td>
<td valign="top" width="75%">
<a href="https://drive.google.com/open?id=1MpvxcW7OTJP321QL_q4ZLQ8D653bZZzy">
<papertitle>Aperture Supervision for Monocular Depth Estimation</papertitle>
</a>
<br>
<a href="https://people.eecs.berkeley.edu/~pratul/">Pratul P. Srinivasan</a>,
<a href="http://rahuldotgarg.appspot.com/">Rahul Garg</a>,
<a href="http://people.csail.mit.edu/nwadhwa/">Neal Wadhwa</a>,
<a href="http://graphics.stanford.edu/~renng/">Ren Ng</a>,
<strong>Jonathan T. Barron</strong>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2018
<br>
<a href="https://github.com/google/aperture_supervision">code</a> /
<a href="Srinivasan2018.bib">bibtex</a>
<p></p>
<p>Varying a camera's aperture provides a supervisory signal that can teach a neural network to do monocular depth estimation.</p>
</td>
</tr>
<tr onmouseout="deepburst_stop()" onmouseover="deepburst_start()">
<td width="25%">
<div class="one">
<div class="two" id='deepburst_image'><img src='deepburst_after.png'></div>
<img src='deepburst_before.png'>
</div>
<script type="text/javascript">
function deepburst_start() {
document.getElementById('deepburst_image').style.opacity = "1";
}
function deepburst_stop() {
document.getElementById('deepburst_image').style.opacity = "0";
}
deepburst_stop()
</script>
</td>
<td valign="top" width="75%">
<a href="https://drive.google.com/file/d/1GAH8ijyZ7GnoBnQFANEzdXinHrE4vvXn/view?usp=sharing">
<papertitle>Burst Denoising with Kernel Prediction Networks</papertitle>
</a>
<br>
<a href="https://people.eecs.berkeley.edu/~bmild/">Ben Mildenhall</a>,
<strong>Jonathan T. Barron</strong>,
<a href="http://people.csail.mit.edu/jiawen/">Jiawen Chen</a>,
<a href="http://www.dsharlet.com/">Dillon Sharlet</a>,
<a href="http://graphics.stanford.edu/~renng/">Ren Ng</a>, Robert Carroll
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2018   <font color=#FF8080><strong>(Spotlight)</strong></font>
<br>
<a href="https://drive.google.com/file/d/1aqk3Q-L2spjLZh2yRWKUWIDcZkGjQ7US/view?usp=sharing">supplement</a> /
<a href="https://github.com/google/burst-denoising">code</a> /
<a href="Mildenhall2018.bib">bibtex</a>
<p></p>
<p>We train a network to predict linear kernels that denoise noisy bursts from cellphone cameras.</p>
</td>
</tr>
<tr onmouseout="friendly_stop()" onmouseover="friendly_start()">
<td width="25%">
<div class="one">
<div class="two" id='friendly_image'><img src='friendly_after.png'></div>
<img src='friendly_before.png'>
</div>
<script type="text/javascript">
function friendly_start() {
document.getElementById('friendly_image').style.opacity = "1";
}
function friendly_stop() {
document.getElementById('friendly_image').style.opacity = "0";
}
friendly_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1w_0djhL0QgC_fbehnJ0c-J23_kW_420p/view?usp=sharing">
<papertitle>A Hardware-Friendly Bilateral Solver for Real-Time Virtual Reality Video</papertitle>
</a>
<br>
<a href="https://homes.cs.washington.edu/~amrita/">Amrita Mazumdar</a>, <a href="http://homes.cs.washington.edu/~armin/">Armin Alaghi</a>, <strong>Jonathan T. Barron</strong>, <a href="https://www.cs.unc.edu/~gallup/">David Gallup</a>, <a href="https://homes.cs.washington.edu/~luisceze/">Luis Ceze</a>, <a href="https://homes.cs.washington.edu/~oskin/">Mark Oskin</a>, <a href="http://homes.cs.washington.edu/~seitz/">Steven M. Seitz</a>
<br>
<em>High-Performance Graphics (HPG)</em>, 2017
<br>
<a href="https://sampa.cs.washington.edu/projects/vr-hw.html">project page</a>
<p></p>
<p>A reformulation of the bilateral solver can be implemented efficiently on GPUs and FPGAs.</p>
</td>
</tr>
<tr onmouseout="hdrnet_stop()" onmouseover="hdrnet_start()">
<td width="25%">
<div class="one">
<div class="two" id='hdrnet_image'><img src='hdrnet_after.jpg'></div>
<img src='hdrnet_before.jpg'>
</div>
<script type="text/javascript">
function hdrnet_start() {
document.getElementById('hdrnet_image').style.opacity = "1";
}
function hdrnet_stop() {
document.getElementById('hdrnet_image').style.opacity = "0";
}
hdrnet_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1jQY3CTMnLX7PeGUzYLso9H1eCsZyWbwg/view?usp=sharing">
<papertitle>Deep Bilateral Learning for Real-Time Image Enhancement</papertitle>
</a>
<br>
<a href="http://www.mgharbi.com">Michaël Gharbi</a>, <a href="http://people.csail.mit.edu/jiawen/">Jiawen Chen</a>, <strong>Jonathan T. Barron</strong>, <a href="https://people.csail.mit.edu/hasinoff/">Samuel W. Hasinoff</a>, <a href="http://people.csail.mit.edu/fredo/">Frédo Durand </a>
<br>
<em>SIGGRAPH</em>, 2017
<br>
<a href="https://groups.csail.mit.edu/graphics/hdrnet/">project page</a> /
<a href="https://www.youtube.com/watch?v=GAe0qKKQY_I">video</a> /
<a href="GharbiSIGGRAPH2017.bib">bibtex</a> /
<a href="http://news.mit.edu/2017/automatic-image-retouching-phone-0802">p</a><a href="https://www.wired.com/story/googles-new-algorithm-perfects-photos-before-you-even-take-them/">r</a><a href="https://petapixel.com/2017/08/02/new-ai-can-retouch-photos-snap/">e</a><a href="https://www.theverge.com/2017/8/2/16082272/google-mit-retouch-photos-machine-learning">s</a><a href="http://gizmodo.com/clever-camera-app-uses-deep-learning-to-perfectly-retou-1797474282">s</a>
<p></p>
<p>By training a deep network in bilateral space we can learn a model for high-resolution and real-time image enhancement.</p>
</td>
</tr>
<tr>
<td width="25%">
<img src='loss.png'>
</td>
<td valign="top" width="75%">
<p>
<a href="https://arxiv.org/abs/1701.03077">
<papertitle>A More General Robust Loss Function</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>
<br>
<em>arXiv Preprint</em>, 2017
<br>
<p></p>
<p>A single robust loss function is a superset of many other common robust loss functions.</p>
</td>
</tr>
<tr onmouseout="ffcc_stop()" onmouseover="ffcc_start()">
<td width="25%">
<div class="one">
<div class="two" id='ffcc_image'><img src='ffcc_after.jpg'></div>
<img src='ffcc_before.jpg'>
</div>
<script type="text/javascript">
function ffcc_start() {
document.getElementById('ffcc_image').style.opacity = "1";
}
function ffcc_stop() {
document.getElementById('ffcc_image').style.opacity = "0";
}
ffcc_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1VDWAS7HgiufTNPP7CQY00KmJP71QIZAy/view?usp=sharing">
<papertitle>Fast Fourier Color Constancy</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, Yun-Ta Tsai
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2017
<br>
<a href="https://drive.google.com/file/d/1b5zdR5UYPTkXa2UgiLhi-PP89bzINJSR/view?usp=sharing">supplement</a> /
<a href="https://youtu.be/rZCXSfl13rY">video</a> /
<a href="BarronTsaiCVPR2017.bib">bibtex</a> /
<a href="https://github.com/google/ffcc">code</a> /
<a href="https://drive.google.com/open?id=0B4nuwEMaEsnmWkJQMlFPSFNzbEk">output</a> /
<a href="https://blog.google/products/photos/six-tips-make-your-photos-pop/">blog post</a> /
<a href="https://9to5google.com/2017/03/03/google-photos-auto-white-balance/">p</a><a href="https://www.engadget.com/2017/03/03/google-photos-automatically-fixes-your-pictures-white-balance/">r</a><a href="https://lifehacker.com/google-photos-will-now-automatically-adjust-the-white-b-1793009155">e</a><a href="https://petapixel.com/2017/03/06/google-photos-will-now-automatically-white-balance-snapshots/">s</a><a href="http://www.theverge.com/2017/3/3/14800062/google-photos-auto-white-balance-android">s</a>
<p></p>
<p>Color space can be aliased, allowing white balance models to be learned and evaluated in the frequency domain. This improves accuracy by 13-20% and speed by 250-3000x.</p>
<p>This technology is used by <a href="https://photos.google.com/">Google Photos</a> and <a href="https://www.google.com/maps">Google Maps</a>.</p>
</td>
</tr>
<tr onmouseout="jump_stop()" onmouseover="jump_start()" bgcolor="#ffffd0">
<td width="25%">
<div class="one">
<div class="two" id='jump_image'><img src='jump_anim.gif'></div>
<img src='jump_still.png'>
</div>
<script type="text/javascript">
function jump_start() {
document.getElementById('jump_image').style.opacity = "1";
}
function jump_stop() {
document.getElementById('jump_image').style.opacity = "0";
}
jump_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1RBnTrtzqmuO8uj3GQaR5vBJZjIC3Jxjn/view?usp=sharing">
<papertitle>Jump: Virtual Reality Video</papertitle>
</a>
<br>
<a href="http://mi.eng.cam.ac.uk/~ra312/">Robert Anderson</a>, <a href="https://www.cs.unc.edu/~gallup/">David Gallup</a>, <strong>Jonathan T. Barron</strong>, <a href="https://mediatech.aalto.fi/~janne/index.php">Janne Kontkanen</a>, <a href="https://www.cs.cornell.edu/~snavely/">Noah Snavely</a>, <a href="http://carlos-hernandez.org/">Carlos Hernández</a>, <a href="https://homes.cs.washington.edu/~sagarwal/">Sameer Agarwal</a>, <a href="https://homes.cs.washington.edu/~seitz/">Steven M Seitz</a>
<br>
<em>SIGGRAPH Asia</em>, 2016
<br>
<a href="https://drive.google.com/file/d/11D4eCDXqqFTtZT0WS2COJE0hsAN3QEww/view?usp=sharing">supplement</a> /
<a href="https://www.youtube.com/watch?v=O0qUYynupTI">video</a> /
<a href="Anderson2016.bib">bibtex</a> /
<a href="https://blog.google/products/google-vr/jump-using-omnidirectional-stereo-vr-video/">blog post</a>
<p></p>
<p>Using computer vision and a ring of cameras, we can make video for virtual reality headsets that is both stereo and 360°.</p>
<p>This technology is used by <a href="https://vr.google.com/jump/">Jump</a>. </p>
<p></p>
</a>
</p>
</td>
</tr>
<tr onmouseout="hdrp_stop()" onmouseover="hdrp_start()">
<td width="25%">
<div class="one">
<div class="two" id='hdrp_image'><img src='hdrp_after.jpg'></div>
<img src='hdrp_before.jpg'>
</div>
<script type="text/javascript">
function hdrp_start() {
document.getElementById('hdrp_image').style.opacity = "1";
}
function hdrp_stop() {
document.getElementById('hdrp_image').style.opacity = "0";
}
hdrp_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/open?id=1SSSmVHWbMQ7sZMOredSVWVJXbXobkyzA">
<papertitle>Burst Photography for High Dynamic Range and Low-Light Imaging on Mobile Cameras</papertitle>
</a>
<br>
<a href="http://people.csail.mit.edu/hasinoff/">Samuel W. Hasinoff</a>, <a href="http://www.dsharlet.com/">Dillon Sharlet</a>, <a href="http://www.geisswerks.com/">Ryan Geiss</a>, <a href="http://people.csail.mit.edu/abadams/">Andrew Adams</a>, <strong>Jonathan T. Barron</strong>, Florian Kainz, <a href="http://people.csail.mit.edu/jiawen/">Jiawen Chen</a>, <a href="http://graphics.stanford.edu/~levoy/">Marc Levoy</a>
<br>
<em>SIGGRAPH Asia</em>, 2016
<br>
<a href="http://hdrplusdata.org/">project page</a> /
<a href="https://drive.google.com/open?id=15EUuSDi1BtHUgQCaiooVrD44qYKIC3vx">supplement</a> /
<a href="Hasinoff2016.bib">bibtex</a>
<p></p>
<p>Mobile phones can take beautiful photographs in low-light or high dynamic range environments by aligning and merging a burst of images.</p>
<p>This technology is used by the <a href="https://research.googleblog.com/2014/10/hdr-low-light-and-high-dynamic-range.html">Nexus HDR+</a> feature. </p>
<p></p>
</a>
</p>
</td>
</tr>
<tr onmouseout="bs_stop()" onmouseover="bs_start()" bgcolor="#ffffd0">
<td width="25%">
<div class="one">
<div class="two" id='bs_image'><img src='BS_after.jpg'></div>
<img src='BS_before.jpg'>
</div>
<script type="text/javascript">
function bs_start() {
document.getElementById('bs_image').style.opacity = "1";
}
function bs_stop() {
document.getElementById('bs_image').style.opacity = "0";
}
bs_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1zFzCaFwkGK1EGmJ_KEqb-ZsRJhfUKN2S/view?usp=sharing">
<papertitle>The Fast Bilateral Solver</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="https://cs.stanford.edu/~poole/">Ben Poole</a>
<br>
<em>European Conference on Computer Vision (ECCV)</em>, 2016   <font color="red"><strong>(Best Paper Honorable Mention)</strong></font>
<br>
<a href="http://arxiv.org/abs/1511.03296">arXiv</a> /
<a href="https://drive.google.com/file/d/0B4nuwEMaEsnmdEREcjhlSXM2NGs/view?usp=sharing">supplement</a> /
<a href="BarronPooleECCV2016.bib">bibtex</a> /
<a href="http://videolectures.net/eccv2016_barron_bilateral_solver/">video (they messed up my slides, use →)</a> /
<a href="https://drive.google.com/file/d/19x1AeN0PFus6Pjrd8nR-vCmJ6bNEefsC/view?usp=sharing">keynote</a>
(or <a href="https://drive.google.com/file/d/1p9nduiymK9jUh7WfwlsMjBfW8RoNe_61/view?usp=sharing">PDF</a>) /
<a href="https://github.com/poolio/bilateral_solver">code</a> /
<a href="https://drive.google.com/file/d/0B4nuwEMaEsnmaDI3bm5VeDRxams/view?usp=sharing">depth super-res results</a> /
<a href="BarronPooleECCV2016_reviews.txt">reviews</a>
<p></p>
<p>Our solver smooths things better than other filters and faster than other optimization algorithms, and you can backprop through it.
<p></p>
</a>
</p>
</td>
</tr>
<tr onmouseout="diverdi_stop()" onmouseover="diverdi_start()">
<td width="25%">
<div class="one">
<div class="two" id='diverdi_image'><img src='diverdi_after.jpg'></div>
<img src='diverdi_before.jpg'>
</div>
<script type="text/javascript">
function diverdi_start() {
document.getElementById('diverdi_image').style.opacity = "1";
}
function diverdi_stop() {
document.getElementById('diverdi_image').style.opacity = "0";
}
diverdi_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1mmT-LuK_eBZsl3qp4-fAshEPdgfbgvNE/view?usp=sharing">
<papertitle>Geometric Calibration for Mobile, Stereo, Autofocus Cameras</papertitle>
</a>
<br>
<a href="http://www.stephendiverdi.com/">Stephen DiVerdi</a>,
<strong>Jonathan T. Barron</strong>
<br>
<em>Winter Conference on Applications of Computer Vision (WACV)</em>, 2016
<br>
<a href="Diverdi2016.bib">bibtex</a>
<p></p>
<p>Standard techniques for stereo calibration don't work for cheap mobile cameras.
</td>
</tr>
<tr onmouseout="dt_stop()" onmouseover="dt_start()">
<td width="25%">
<div class="one">
<div class="two" id='dt_image'><img src='DT_edge.jpg'></div>
<img src='DT_image.jpg'>
</div>
<script type="text/javascript">
function dt_start() {
document.getElementById('dt_image').style.opacity = "1";
}
function dt_stop() {
document.getElementById('dt_image').style.opacity = "0";
}
dt_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/178Xj2PZ1w6hZJpucU-TiZOoCemJmvsVQ/view?usp=sharing">
<papertitle>Semantic Image Segmentation with Task-Specific Edge Detection Using CNNs and a Discriminatively Trained Domain Transform</papertitle>
</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2016
<br>
<a href="http://liangchiehchen.com/">Liang-Chieh Chen</a>, <strong>Jonathan T. Barron</strong>, <a href="http://ttic.uchicago.edu/~gpapan/">George Papandreou</a>, <a href="http://www.cs.ubc.ca/~murphyk/">Kevin Murphy</a>, <a href="http://www.stat.ucla.edu/~yuille/">Alan L. Yuille</a>
<br>
<a href="Chen2016.bib">bibtex</a></em> /
<a href="http://liangchiehchen.com/projects/DeepLab.html">project page</a> /
<a href="https://bitbucket.org/aquariusjay/deeplab-public-ver2">code</a>
<p></p>
<p>By integrating an edge-aware filter into a convolutional neural network we can learn an edge-detector while improving semantic segmentation.</p>
</td>
</tr>
<tr onmouseout="ccc_stop()" onmouseover="ccc_start()" bgcolor="#ffffd0">
<td width="25%">
<div class="one">
<div class="two" id='ccc_image'><img src='ccc_after.jpg'></div>
<img src='ccc_before.jpg'>
</div>
<script type="text/javascript">
function ccc_start() {
document.getElementById('ccc_image').style.opacity = "1";
}
function ccc_stop() {
document.getElementById('ccc_image').style.opacity = "0";
}
ccc_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1id74VNDL8ACrrWf6vYgN2M4kS8gd4n7w/view?usp=sharing">
<papertitle>Convolutional Color Constancy</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>
<br>
<em>International Conference on Computer Vision (ICCV)</em>, 2015
<br>
<a href="https://drive.google.com/file/d/1vO3sVOMihmpNqsuASeR46Y_iME0lOANR/view?usp=sharing">supplement</a> / <a href="BarronICCV2015.bib">bibtex</a> / <a href="https://youtu.be/saHwKY9rfx0">video</a> </a> (or <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmalBNUzlENUJSVDg/view?usp=sharing">mp4</a>)
<p></p>
<p>By framing white balance as a chroma localization task we can discriminatively learn a color constancy model that beats the state-of-the-art by 40%.</p>
</td>
</tr>
<tr>
<td width="25%">
<img src='Shelhamer2015.jpg'>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1stygV71uBruD7Ck9CaAQr7nREvr3DtUL/view?usp=sharing">
<papertitle>Scene Intrinsics and Depth from a Single Image</papertitle>
</a>
<br>
<a href="http://imaginarynumber.net/">Evan Shelhamer</a>, <strong>Jonathan T. Barron</strong>, <a href="http://www.eecs.berkeley.edu/%7Etrevor/">Trevor Darrell</a>
<br>
<em>International Conference on Computer Vision (ICCV) Workshop</em>, 2015
<br>
<a href="Shelhamer2015.bib">bibtex</a>
<p></p>
<p>The monocular depth estimates produced by fully convolutional networks can be used to inform intrinsic image estimation.</p>
</td>
</tr>
<tr bgcolor="#ffffd0" onmouseout="defocus_stop()" onmouseover="defocus_start()">
<td width="25%">
<div id='lens_blurry' class='hidden'><img src="BarronCVPR2015_anim.gif"></div>
<div id='lens_sharp'>
<a href="BarronCVPR2015_anim.gif"><img src="BarronCVPR2015_still.jpg"></a>
</div>
<script type="text/javascript">
function defocus_start() {
document.getElementById('lens_blurry').style.display = 'inline';
document.getElementById('lens_sharp').style.display = 'none';
}
function defocus_stop() {
document.getElementById('lens_blurry').style.display = 'none';
document.getElementById('lens_sharp').style.display = 'inline';
}
defocus_stop()
</script>
</td>
<td valign="top" width="75%">
<p>
<a href="https://drive.google.com/file/d/1R4RdaBZIs-uJobhIFs9yKf3jIsaHQNH0/view?usp=sharing">
<papertitle>Fast Bilateral-Space Stereo for Synthetic Defocus</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://people.csail.mit.edu/abadams/">Andrew Adams</a>, <a href="http://people.csail.mit.edu/yichangshih/">YiChang Shih, <a href="http://carlos-hernandez.org/">Carlos Hernández</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2015   <font color="red"><strong>(Oral Presentation)</strong></font>
<br>
<a href="https://drive.google.com/file/d/125qgMdqeT1vojMIijIKcOF099LjUgUOL/view?usp=sharing">abstract</a> / <a href="https://drive.google.com/file/d/1HGGvVOGxmPjvgdK5q3UD1Qb5Nttg6kq9/view?usp=sharing">supplement</a> / <a href="BarronCVPR2015.bib">bibtex</a> /
<a href="http://techtalks.tv/talks/fast-bilateral-space-stereo-for-synthetic-defocus/61624/">talk</a> /
<a href="https://drive.google.com/file/d/0B4nuwEMaEsnmSzZZdUJSMllSUkE/view?usp=sharing">keynote</a> (or <a href="https://drive.google.com/open?id=0B4nuwEMaEsnmZ1ZXUzBCWDJYeFU">PDF</a>)
<p></p>
<p>By embedding a stereo optimization problem in "bilateral-space" we can very quickly solve for an edge-aware depth map, letting us render beautiful depth-of-field effects.</p>
<p>This technology is used by the <a href="http://googleresearch.blogspot.com/2014/04/lens-blur-in-new-google-camera-app.html">Google Camera "Lens Blur"</a> feature. </p>
<p></p>
</a>
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="PABMM2015.jpg" alt="PontTuset" width="160" style="border-style: none">
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1EUvfslce9iqCbJ_IAE4J1ser6hpse_uh/view?usp=sharing" id="MCG_journal">
<papertitle>Multiscale Combinatorial Grouping for Image Segmentation and Object Proposal Generation</papertitle>
</a>
<br>
<a href="http://imatge.upc.edu/web/people/jordi-pont-tuset">Jordi Pont-Tuset</a>, <a href="http://www.cs.berkeley.edu/~arbelaez/">Pablo Arbeláez</a>, <strong>Jonathan T. Barron</strong>, <a href="http://imatge.upc.edu/web/ferran">Ferran Marqués</a>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) </em>, 2017
<br>
<a href="http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/mcg/">project page</a> /
<a href="PontTusetTPAMI2017.bib">bibtex</a> /
<a href="https://drive.google.com/file/d/1AiB78Fy7QVA3KqgcooyzMAC5L8HhNzjz/view?usp=sharing">fast eigenvector code</a>
</p>
<p>
We produce state-of-the-art contours, regions and object candidates, and we compute normalized-cuts eigenvectors 20× faster.
</p>
<p>
This paper subsumes our CVPR 2014 paper.
</p>
</td>
</tr>
<tr bgcolor="#ffffd0" onmouseout="sirfs_stop()" onmouseover="sirfs_start()">
<td width="25%">
<div class="one">
<div class="two" id='sirfs_image'>
<a href="Estee.png"><img src='Estee_160.png' style="border-style: none"></a>
</div>
<a href="Estee.png"><img src='Estee_160_prodB2.png' style="border-style: none"></a>
</div>
<script type="text/javascript">
function sirfs_start() {
document.getElementById('sirfs_image').style.opacity = "1";
}
function sirfs_stop() {
document.getElementById('sirfs_image').style.opacity = "0";
}
sirfs_stop()
</script>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1RvyCiDMg--jyO8lLBvopp0o271LvREoa/view?usp=sharing" id="SIRFS">
<papertitle>Shape, Illumination, and Reflectance from Shading</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)</em>, 2015
<br>
<a href="https://drive.google.com/file/d/1D3k6u4Ek2dWm2Yf7kl1Vu_g1HFSxztqF/view?usp=sharing">supplement</a> / <a href="BarronMalikTPAMI2015.bib">bibtex</a> / <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmVWpfa19mbUxIYW8/view?usp=sharing">keynote</a> (or <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmazJvLXJUb0NuM1U/view?usp=sharing">powerpoint</a>, <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmTDBUWE96VHJndjg/view?usp=sharing">PDF</a>) / <a href="http://www.youtube.com/watch?v=NnePYprvFvA">video</a> / <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmem4tdm93ZDVWRUE/view?usp=sharing">code & data</a> / <a href="why_did_this_paper_come_out_in_2015.txt">rant</a> / <a href="https://drive.google.com/file/d/11X5Zfjy7Q7oP_V2rtqy2f5-x9YgQUAFd/view?usp=sharing">kudos</a>
</p>
<p>
We present <strong>SIRFS</strong>, which can estimate shape, chromatic illumination, reflectance, and shading from a single image of an masked object.
</p>
<p>
This paper subsumes our CVPR 2011, CVPR 2012, and ECCV 2012 papers.
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="ArbalaezCVPR2014.jpg" alt="ArbalaezCVPR2014" width="160" height="120" style="border-style: none">
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1M0wijHY134F9ETBgO8mjeuKUSblTRLG0/view?usp=sharing">
<papertitle>Multiscale Combinatorial Grouping</papertitle>
</a>
<br>
<a href="http://www.cs.berkeley.edu/~arbelaez/">Pablo Arbeláez</a>, <a href="http://imatge.upc.edu/web/people/jordi-pont-tuset">Jordi Pont-Tuset</a>, <strong>Jonathan T. Barron</strong>, <a href="http://imatge.upc.edu/web/ferran">Ferran Marqués</a>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2014
<br>
<a href="http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/mcg/">project page</a> /
<a href="ArbelaezCVPR2014.bib">bibtex</a>
</p>
<p>This paper is subsumed by <a href="#MCG_journal">our journal paper</a>.</p>
<br>
</p>
</td>
</tr>
<tr onmouseout="flyspin_stop()" onmouseover="flyspin_start()">
<td width="25%">
<div id='flyspin' class='hidden'><img src="BarronICCV2013_160.gif"></div>
<div id='flystill'>
<a href="BarronICCV2013.gif"><img src="BarronICCV2013_160.jpg"></a>
</div>
<script type="text/javascript">
function flyspin_start() {
document.getElementById('flyspin').style.display = 'inline';
document.getElementById('flystill').style.display = 'none';
}
function flyspin_stop() {
document.getElementById('flyspin').style.display = 'none';
document.getElementById('flystill').style.display = 'inline';
}
flyspin_stop()
</script>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1shvItvx_8Sb8QNXhrOXkuRmx2618iwNJ/view?usp=sharing">
<papertitle>Volumetric Semantic Segmentation using Pyramid Context Features</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~arbelaez/">Pablo Arbeláez</a>, <a href="http://big.lbl.gov/">Soile V. E. Keränen</a>, <a href="http://www.lbl.gov/gsd/biggin.html">Mark D. Biggin</a>,
<br> <a href="http://dwknowles.lbl.gov/">David W. Knowles</a>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>International Conference on Computer Vision (ICCV)</em>, 2013
<br>
<a href="https://drive.google.com/file/d/1htiLpMAcYLtuBthmAb4XHnOYxUbkfnqR/view?usp=sharing">supplement</a> /
<a href="https://drive.google.com/file/d/1qoYeFNa443myn2SfcdhmCsYBqE9xQrPD/view?usp=sharing">poster</a> /
<a href="BarronICCV2013.bib">bibtex</a> / <a href="http://www.youtube.com/watch?v=Y56-FcfnlVA&hd=1">video 1</a> (or <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmZ1ZLaHdQYzAxNlU/view?usp=sharing">mp4</a>) / <a href="http://www.youtube.com/watch?v=mvRoYuP6-l4&hd=1">video 2</a> (or <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmZ1ZLaHdQYzAxNlU/view?usp=sharing">mp4</a>) / <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmSF9YdWJjQmh4QW8/view?usp=sharing">code & data</a>
</p>
<p>
We present a technique for efficient per-voxel linear classification, which enables accurate and fast semantic segmentation of volumetric Drosophila imagery.
<br>
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="3DSP_160.jpg" alt="3DSP" width="160" height="120" style="border-style: none">
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/0B4nuwEMaEsnmbG1tOGIta3N1Wjg/view?usp=sharing" id="3DSP">
<papertitle>3D Self-Portraits</papertitle>
</a>
<br>
<a href="http://www.hao-li.com/">Hao Li</a>, <a href="http://www.evouga.com/">Etienne Vouga</a>, Anton Gudym, <a href="http://www.cs.princeton.edu/~linjiel/">Linjie Luo</a>, <strong>Jonathan T. Barron</strong>, Gleb Gusev
<br>
<em>SIGGRAPH Asia</em>, 2013
<br>
<a href="http://www.youtube.com/watch?v=DmUkbZ0QMCA">video</a> / <a href="http://shapify.me/">shapify.me</a> / <a href="3DSP_siggraphAsia2013.bib">bibtex</a>
</p>
<p>
Our system allows users to create textured 3D models of themselves in arbitrary poses using only a single 3D sensor.
<br>
</p>
</td>
</tr>
<tr onmouseout="rgbd_stop()" onmouseover="rgbd_start()">
<td width="25%">
<div id='rgbd_anim' class='hidden'><img src="SceneSIRFS.gif"></div>
<div id='rgbd_still'><img src="SceneSIRFS-still.jpg"></div>
<script type="text/javascript">
function rgbd_start() {
document.getElementById('rgbd_anim').style.display = 'inline';
document.getElementById('rgbd_still').style.display = 'none';
}
function rgbd_stop() {
document.getElementById('rgbd_anim').style.display = 'none';
document.getElementById('rgbd_still').style.display = 'inline';
}
rgbd_stop()
</script>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1snypSLhzC0jXCchJRsWpcDZ7Es5hDmXo/view?usp=sharing">
<papertitle>Intrinsic Scene Properties from a Single RGB-D Image</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2013   <font color="red"><strong>(Oral Presentation)</strong></font>
<br>
<a href="https://drive.google.com/file/d/1cLUw72WpgdZ_3TQAjJABdgywqjBfn_Mq/view?usp=sharing">supplement</a> / <a href="BarronMalikCVPR2013.bib">bibtex</a> / <a href="http://techtalks.tv/talks/intrinsic-scene-properties-from-a-single-rgb-d-image/58614/">talk</a> / <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmWW1CZGJPbi12R0k/view?usp=sharing">keynote</a> (or <a href="https://drive.google.com/file/d/19q3EFf6GIb4UFcCN2DVU2jVKpxRj5kxf/view?usp=sharing">powerpoint</a>, <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmMzQ4ZVp1SWdnVkk/view?usp=sharing">PDF</a>) / <a href="https://drive.google.com/file/d/0B4nuwEMaEsnmOXFOTm5oUHpRamc/view?usp=sharing">code & data</a>
</p>
<p>By embedding mixtures of shapes & lights into a soft segmentation of an image, and by leveraging the output of the Kinect, we can extend SIRFS to scenes.
<br>
<br> TPAMI Journal version: <a href="https://drive.google.com/file/d/1iQiUxZvjPPnb8rFCwXYesTgFSRk7mkAq/view?usp=sharing">version</a> / <a href="BarronMalikTPAMI2015B.bib">bibtex</a>
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="Boundary.jpg" alt="Boundary_png" style="border-style: none"></a>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1H4YPovfrvcce3HGMEhidwU2l2fTcNR5y/view?usp=sharing">
<papertitle>Boundary Cues for 3D Object Shape Recovery</papertitle>
</a>
<br>
<a href="http://www.kevinkarsch.com/">Kevin Karsch</a>,
<a href="http://web.engr.illinois.edu/~liao17/">Zicheng Liao</a>,
<a href="http://web.engr.illinois.edu/~jjrock2/">Jason Rock</a>,
<strong>Jonathan T. Barron</strong>,
<a href="http://www.cs.illinois.edu/homes/dhoiem/">Derek Hoiem</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2013
<br>
<a href="https://drive.google.com/file/d/0B4nuwEMaEsnmLUQ5SVJTcUZIYXc/view?usp=sharing">supplement</a> / <a href="KarschCVPR2013.bib">bibtex</a>
</p>
<p>Boundary cues (like occlusions and folds) can be used for shape reconstruction, which improves object recognition for humans and computers.
<br>
</p>
</td>
</tr>
<tr onmouseout="eccv12_stop()" onmouseover="eccv12_start()">
<td width="25%">
<div id='eccv12_anim' class='hidden'>
<a href="https://drive.google.com/file/d/1brxb58CfRPe7KEER4Q_fYS9B_J-hiS0t/view?usp=sharing"><img src="ECCV2012_small.gif"></a>
</div>
<div id='eccv12_still'><img src="ECCV2012_still.jpg"></div>
<script type="text/javascript">
function eccv12_start() {
document.getElementById('eccv12_anim').style.display = 'inline';
document.getElementById('eccv12_still').style.display = 'none';
}
function eccv12_stop() {
document.getElementById('eccv12_anim').style.display = 'none';
document.getElementById('eccv12_still').style.display = 'inline';
}
eccv12_stop()
</script>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1NczR4pJ-s0YBjCe0rCevMt8IM5JPuUrc/view?usp=sharing">
<papertitle>Color Constancy, Intrinsic Images, and Shape Estimation</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>European Conference on Computer Vision (ECCV)</em>, 2012
<br>
<a href="https://drive.google.com/file/d/1zuxhWZ3i6THvuRRBeE7dM_BJfDxO72Fq/view?usp=sharing">supplement</a> / <a href="BarronMalikECCV2012.bib">bibtex</a> / <a href="https://drive.google.com/file/d/12x8mhqpFsA6p0u6ZQW-ieRKF8hlQBKKe/view?usp=sharing">poster</a> / <a href="http://www.youtube.com/watch?v=NnePYprvFvA">video</a>
</p>
<p>This paper is subsumed by <a href="#SIRFS">SIRFS</a>.</p>
</td>
</tr>
<tr onmouseout="cvpr2012_stop()" onmouseover="cvpr2012_start()">
<td width="25%">
<div class="one">
<div class="two" id='cvpr2012_image'>
<img src='BarronCVPR2012_after.jpg' style="border-style: none"></div>
<img src='BarronCVPR2012_before.jpg' style="border-style: none">
</div>
<script type="text/javascript">
function cvpr2012_start() {
document.getElementById('cvpr2012_image').style.opacity = "1";
}
function cvpr2012_stop() {
document.getElementById('cvpr2012_image').style.opacity = "0";
}
cvpr2012_stop()
</script>
</td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/17RfINbE2dr2EjXp9MtGO0MHJLQmQVhvT/view?usp=sharing">
<papertitle>Shape, Albedo, and Illumination from a Single Image of an Unknown Object</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2012
<br>
<a href="https://drive.google.com/file/d/1Im_bUI42AP9VPoNtsjLajvtLRiwv39k3/view?usp=sharing">supplement</a> / <a href="BarronMalikCVPR2012.bib">bibtex</a> / <a href="https://drive.google.com/file/d/1IAlSF4k3_CEL9dfbaMiNTFPBoEkLhsRl/view?usp=sharing">poster</a>
</p>
<p>This paper is subsumed by <a href="#SIRFS">SIRFS</a>.</p>
</td>
</tr>
<tr>
<td width="25%"><img src="B3DO.jpg" alt="b3do" width="160" style="border-style: none"></td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1_S8EQyngbHQrB415o0XkQ4V9SMzdEgWT/view?usp=sharing">
<papertitle>A Category-Level 3-D Object Dataset: Putting the Kinect to Work</papertitle>
</a>
<br>
<a href="http://www.eecs.berkeley.edu/%7Eallie/">Allison Janoch</a>, <a href="http://sergeykarayev.com/">Sergey Karayev</a>, <a href="http://www.eecs.berkeley.edu/%7Ejiayq/">Yangqing Jia</a>, <strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/%7Emfritz/">Mario Fritz</a>, <a href="http://www.icsi.berkeley.edu/%7Esaenko/">Kate Saenko</a>, <a href="http://www.eecs.berkeley.edu/%7Etrevor/">Trevor Darrell</a>
<br>
<em>International Conference on Computer Vision (ICCV) 3DRR Workshop</em>, 2011
<br>
<a href="B3DO_ICCV_2011.bib">bibtex</a> / <a href="https://drive.google.com/file/d/1qf4-U5RhSw12O7gzQwW66SMQhs2FWYDW/view?usp=sharing">"smoothing" code</a>
</p>
<p>We present a large RGB-D dataset of indoor scenes and investigate ways to improve object detection using depth information.
<br>
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="safs.jpg" alt="safs_small" width="160" height="160" style="border-style: none"></td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1EZTOO5xezLYcyIFgAzs4KuZFLbTcwTDH/view?usp=sharing">
<papertitle>High-Frequency Shape and Albedo from Shading using Natural Image Statistics</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>Computer Vision and Pattern Recognition (CVPR)</em>, 2011
<br>
<a href="BarronMalikCVPR2011.bib">bibtex</a>
</p>
<p>This paper is subsumed by <a href="#SIRFS">SIRFS</a>.</p>
</td>
</tr>
<tr>
<td width="25%"><img src="fast_texture.jpg" alt="fast-texture" width="160" height="160"></td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1rc05NatkQVmUDlGCAYcHSrvAzTpU9knT/view?usp=sharing">
<papertitle>Discovering Efficiency in Coarse-To-Fine Texture Classification</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://www.cs.berkeley.edu/~malik/">Jitendra Malik</a>
<br>
<em>Technical Report</em>, 2010
<br>
<a href="BarronTR2010.bib">bibtex</a>
</p>
<p>We introduce a model and feature representation for joint texture classification and segmentation that learns how to classify accurately and when to classify efficiently. This allows for sub-linear coarse-to-fine classification.
<br>
</p>
</td>
</tr>
<tr>
<td width="25%"><img src="bd_promo.jpg" alt="blind-date" width="160" height="160"></td>
<td width="75%" valign="top">
<p>
<a href="https://drive.google.com/file/d/1PQjzKgFcrAesMIDJr-WDlCwuGUxZJZwO/view?usp=sharing">
<papertitle>Blind Date: Using Proper Motions to Determine the Ages of Historical Images</papertitle>
</a>
<br>
<strong>Jonathan T. Barron</strong>, <a href="http://cosmo.nyu.edu/hogg/">David W. Hogg</a>, <a href="http://www.astro.princeton.edu/~dstn/">Dustin Lang</a>, <a href="http://cs.nyu.edu/~roweis/">Sam Roweis</a>
<br>
<em>The Astronomical Journal</em>, 136, 2008
</p>
<p>Using the relative motions of stars we can accurately estimate the date of origin of historical astronomical images.</p>
</td>
</tr>
<tr>