-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathpublications copy.html
1388 lines (1134 loc) · 90 KB
/
publications copy.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 XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<meta http-equiv="Content-Style-Type" content="text/css">
<meta name="generator" content="pandoc">
<title>ravikiran</title>
<script async="" src="./ravikiran_files/analytics.js"></script><script src="./ravikiran_files/jquery-1.js"></script>
<link rel="stylesheet" href="./ravikiran_files/bootstrap.css">
<link rel="stylesheet" href="./ravikiran_files/bootstrap-theme.css">
<script src="./ravikiran_files/bootstrap.js"></script>
<style>
/* http://stackoverflow.com/questions/18325779/bootstrap-3-collapse-show-state-with-chevron-icon */
.panel-heading .accordion-toggle:before {
font-family: 'Glyphicons Halflings';
content: "\e114";
float: left;
color: black;
padding-right: 6px;
}
.panel-heading .accordion-toggle.collapsed:before {
content: "\e080";
}
table, th, td {
border: 0px solid black;
border-collapse: collapse;
}
</style>
<style type="text/css">
</style>
</head>
<body>
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
})(window,document,'script','//www.google-analytics.com/analytics.js','ga');
ga('create', 'UA-11803764-2', 'uchicago.edu');
ga('require', 'displayfeatures');
ga('send', 'pageview');
</script>
<nav class="navbar navbar-default navbar-static-top" role="navigation">
<div class="container">
<ul class="nav navbar-nav">
<li><a href="index.html"><font style="font-size:20px; font-weight:500;" color="#003380">Ravi Kiran Sarvadevabhatla</font></a></li>
<li><a href="publications.html"><font size="4px" color="#E62E00"><b>Publications</b></font></a></li>
<li><a href="research.html"><font size="4px" color="#E62E00"><b>Research</b></font></a></li>
</ul>
</div>
</nav>
<div class="container">
<a name="publications"></a>
<p style="margin:-15px 0px 0px 0px;"></p>
<font size="4">Full list also on <a href="https://scholar.google.co.in/citations?user=oLJTcXIAAAAJ&hl=en">Google scholar</a></font>
<br>
<br>
<p style="margin:-2.5px 0px 0px 0px;"></p>
<br>
<h2>2020</h2>
<br>
<!-- TOPO MAPPING -->
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/iros20topological.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Topological Mapping for Manhattan-like Repetitive Environments</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">Sai Shubodh Puligilla, Satyajit Tourani, Tushar Vaidya, Udit Singh Parihar, Ravi Kiran Sarvadevabhatla, K. Madhava Krishna</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em><strong> International Conference on Robotics and Automation (ICRA) </strong></em>, 2020</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> This paper explores the role of topological understanding and benefits of such an understanding to the robot SLAM framework.</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://iiitaphyd-my.sharepoint.com/:b:/g/personal/shubodh_sai_research_iiit_ac_in/EdIIaEofWgJKlX1KiLo9n6oBqClcOk_oJgQwk7svtrwRIw?e=clSwFV" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#Toporepetitiveabs" href="#Toporepetitiveabs-list">Abstract</a>
<!--a class="buttonSS" data-toggle="collapse" data-parent="#Toporepetitive" href="#Toporepetitive-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="Deligan-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;"> @InProceedings{Gurumurthy_2017_CVPR, <br>
author = {Gurumurthy, Swaminathan and Kiran Sarvadevabhatla, Ravi and Venkatesh Babu, R.}, <br>
title = {DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data}, <br>
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, <br>
month = {July}, <br>
year = {2017} <br>
}
</p-->
<div id="Toporepetitive-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="Toporepetitiveabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
We showcase a topological mapping framework
for a challenging indoor warehouse setting. At the most abstract
level, the warehouse is represented as a Topological Graph
where the nodes of the graph represent a particular warehouse
topological construct (e.g. rackspace, corridor) and the edges
denote the existence of a path between two neighbouring
nodes or topologies. At the intermediate level, the map is
represented as a Manhattan Graph where the nodes and edges
are characterized by Manhattan properties and as a Pose Graph
at the lower-most level of detail. The topological constructs
are learned via a Deep Convolutional Network while the
relational properties between topological instances are learnt
via a Siamese-style Neural Network. In the paper, we show
that maintaining abstractions such as Topological Graph and
Manhattan Graph help in recovering an accurate Pose Graph
starting from a highly erroneous and unoptimized Pose Graph.
We show how this is achieved by embedding topological and
Manhattan relations as well as Manhattan Graph aided loop
closure relations as constraints in the backend Pose Graph
optimization framework. The recovery of near ground-truth
Pose Graph on real-world indoor warehouse scenes vindicate
the efficacy of the proposed framework.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/new_burst.png"></div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2019</h2>
<br>
<!-- FUSION RE-ID -->
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/fusion-reid.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Operator-in-the-Loop Deep Sequential Multi-camera Feature Fusion for Person Re-identification</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">Navaneet Murthy, Shashank Shekhar, <strong>Ravi Kiran Sarvadevabhatla</strong>, R. Venkatesh Babu, Anirban Chakraborty</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em><strong> IEEE Transactions on Information Forensics and Security (IEEE T-IFS) </strong></em>, 2019</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> An intelligent sequential fusion technique for multi-camera person reidentification (re-id). The approach is designed to not only improve re-id accuracy but to also learn increasingly better feature representations as observations from additional cameras are fused. </p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1807.07295.pdf" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#Fusionreidabs" href="#Fusionreidabs-list">Abstract</a>
<!--a class="buttonSS" data-toggle="collapse" data-parent="#Fusionreid" href="#Fusionreid-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="Deligan-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;"> @InProceedings{Gurumurthy_2017_CVPR, <br>
author = {Gurumurthy, Swaminathan and Kiran Sarvadevabhatla, Ravi and Venkatesh Babu, R.}, <br>
title = {DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data}, <br>
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, <br>
month = {July}, <br>
year = {2017} <br>
}
</p-->
<div id="Fusionreid-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="Fusionreidabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Given a target image as query, person re-identification systems retrieve a ranked list of candidate matches on a per-camera
basis. In deployed systems, a human operator scans these lists and labels sighted targets by touch or mouse-based selection.
However, classical re-id approaches generate per-camera lists independently. Therefore, target identifications by operator in a subset
of cameras cannot be utilized to improve ranking of the target in remaining set of network cameras. To address this shortcoming, we
propose a novel sequential multi-camera re-id approach. The proposed approach can accommodate human operator inputs and
provides early gains via a monotonic improvement in target ranking. At the heart of our approach is a fusion function which operates on
deep feature representations of query and candidate matches. We formulate an optimization procedure custom-designed to
incrementally improve query representation. Since existing evaluation methods cannot be directly adopted to our setting, we also
propose two novel evaluation protocols. The results on two large-scale re-id datasets (Market-1501, DukeMTMC-reID) demonstrate
that our multi-camera method significantly outperforms baselines and other popular feature fusion schemes. Additionally, we conduct a
comparative subject-based study of human operator performance. The superior operator performance enabled by our approach makes
a compelling case for its integration into deployable video-surveillance systems.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/new_burst.png"></div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<!-- INDISCAPES -->
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/indiscapes19.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Indiscapes: Instance segmentation networks for layout parsing of historical indic manuscripts <strong><font color=red> [ORAL]</font> </strong></h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">Abhishek Prusty, Aitha Sowmya, Abhishek Trivedi, Ravi Kiran Sarvadevabhatla</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In IAPR International Conference on Document Analysis and Recognition <em><strong> (ICDAR) </strong></em>, 2019</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;">
We introduce Indiscapes - the largest publicly available layout annotated dataset of historical Indic manuscript images.
</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1912.07025.pdf" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#Indiscapesabs" href="#Indiscapesabs-list">Abstract</a>
<a href="http://ihdia.iiit.ac.in/" target="_blank" class="buttonPP">Project page</a>
<!--a class="buttonSS" data-toggle="collapse" data-parent="#Indiscapes" href="#Indiscapes-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="Deligan-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;"> @InProceedings{Gurumurthy_2017_CVPR, <br>
author = {Gurumurthy, Swaminathan and Kiran Sarvadevabhatla, Ravi and Venkatesh Babu, R.}, <br>
title = {DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data}, <br>
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, <br>
month = {July}, <br>
year = {2017} <br>
}
</p-->
<div id="Indiscapes-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="Indiscapesabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Historical palm-leaf manuscript and early paper documents from Indian subcontinent form an important part of the world’s literary and cultural heritage. Despite their importance, large-scale annotated Indic manuscript image datasets do
not exist. To address this deficiency, we introduce Indiscapes, the first ever dataset with multi-regional layout annotations
for historical Indic manuscripts. To address the challenge of
large diversity in scripts and presence of dense, irregular layout elements (e.g. text lines, pictures, multiple documents per
image), we adapt a Fully Convolutional Deep Neural Network
architecture for fully automatic, instance-level spatial layout
parsing of manuscript images. We demonstrate the effectiveness
of proposed architecture on images from the Indiscapes dataset.
For annotation flexibility and keeping the non-technical nature
of domain experts in mind, we also contribute a custom, webbased GUI annotation tool and a dashboard-style analytics portal.
Overall, our contributions set the stage for enabling downstream
applications such as OCR and word-spotting in historical Indic
manuscripts at scale.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/Oral-session-icon.png"></div>
<div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/new_burst.png"></div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2018</h2>
<br>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="ravikiran_files/pictionary-robot-camready-1.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Pictionary-style word-guessing on hand-drawn object sketches: dataset, analysis and deep network models</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Shiv Surya, Trisha Mittal, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em><strong>IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)</strong></em>, 2018</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> Journal version of the AAAI-18 paper</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://drive.google.com/file/d/1uvzUyyFalV7Fyp_XEX2hkg7CKilK2G90/view?usp=sharing" target="_blank" class="buttonTT">Paper (Pre-print)</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchGuessJourabs" href="#SketchGuessJourabs-list">Abstract</a>
<a href="https://github.com/val-iisc/sketchguess" target="_blank" class="buttonPP">Project page</a>
<!--a class="buttonSS" data-toggle="collapse" data-parent="#SketchGuess" href="#SketchGuess-list">Bibtex</a--><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchGuessJour-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="SketchGuessJourabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
The ability of intelligent agents to play games in human-like fashion is popularly considered a benchmark of progress in Artificial Intelligence. In our work, we introduce the first computational model aimed at Pictionary, the popular word-guessing social game. We first introduce Sketch-QA, a guessing task. Styled after Pictionary, Sketch-QA uses incrementally accumulated sketch stroke sequences as visual data. Sketch-QA involves asking a fixed question (``What object is being drawn?") and gathering open-ended guess-words from human guessers. We analyze the resulting dataset and present many interesting findings therein. To mimic Pictionary-style guessing, we propose a deep neural model which generates guess-words in response to temporally evolving human-drawn object sketches. Our model even makes human-like mistakes while guessing, thus amplifying the human mimicry factor. We evaluate our model on the large-scale guess-word dataset generated via Sketch-QA task and compare with various baselines. We also conduct a Visual Turing Test to obtain human impressions of the guess-words generated by humans and our model. Experimental results demonstrate the promise of our approach for Pictionary and similarly themed games.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="ravikiran_files/pictionary-robot-camready-1.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Game of Sketches: Deep
Recurrent Models of Pictionary-style Word Guessing</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Shiv Surya, Trisha Mittal, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em><strong>AAAI</strong></em>, 2018</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> The first-ever deep neural network for mimicking Pictionary-style guessing with object sketches as input. </p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1801.09356" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchGuessabs" href="#SketchGuessabs-list">Abstract</a>
<a href="https://github.com/val-iisc/sketchguess" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SketchGuess" href="#SketchGuess-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchGuess-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="SketchGuessabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
The ability of intelligent agents to play games in human-like fashion is popularly considered a benchmark of progress in Artificial Intelligence. Similarly, performance on multi-disciplinary tasks such as Visual Question Answering (VQA) is considered a marker for gauging progress in Computer Vision. In our work, we bring games and VQA together. Specifically, we introduce the first computational model aimed at Pictionary, the popular word-guessing social game. We first introduce Sketch-QA, an elementary version of Visual Question Answering task. Styled after Pictionary, Sketch-QA uses incrementally accumulated sketch stroke sequences as visual data. Notably, Sketch-QA involves asking a fixed question ("What object is being drawn?") and gathering open-ended guess-words from human guessers. We analyze the resulting dataset and present many interesting findings therein. To mimic Pictionary-style guessing, we subsequently propose a deep neural model which generates guess-words in response to temporally evolving human-drawn sketches. Our model even makes human-like mistakes while guessing, thus amplifying the human mimicry factor. We evaluate our model on the large-scale guess-word dataset generated via Sketch-QA task and compare with various baselines. We also conduct a Visual Turing Test to obtain human impressions of the guess-words generated by humans and our model. Experimental results demonstrate the promise of our approach for Pictionary and similarly themed games.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2017</h2>
<br>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/SketchParse.svg"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">SketchParse : Towards Rich Descriptions for Poorly Drawn Sketches using Multi-Task Hierarchical Deep Networks</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Isht Dwivedi, Abhijat Biswas, Sahil Manocha, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM Multimedia (<strong>ACMMM</strong>)</em>, 2017</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> We explore the problem of parsing sketched objects, i.e. given a freehand line sketch of an object, determine its salient attributes (e.g. category, semantic parts, pose). To this end, we propose SketchParse, the first deep-network architecture for fully automatic parsing of freehand object sketches. </p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1709.01295" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchParseabs" href="#SketchParseabs-list">Abstract</a>
<a href="http://val.cds.iisc.ac.in/sketchparse/" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SketchParse" href="#SketchParse-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchParse-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
</p>
</div>
<div id="SketchParseabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
The ability to semantically interpret hand-drawn line sketches, although very challenging, can pave way for novel applications in multimedia. We propose
SketchParse, the first deep-network architecture for fully automatic parsing of freehand object sketches. SketchParse is configured as a two-level fully convolutional network. The first level contains shared layers common to all
object categories. The second level contains a number of expert sub-networks. Each expert specializes in parsing sketches from object categories which contain structurally similar parts. Effectively, the two-level configuration
enables our architecture to scale up efficiently as additional categories are added. We introduce a router layer which (i) relays sketch features from shared layers to the correct expert (ii) eliminates the need to manually specify
object category during inference. To bypass laborious part-level annotation, we sketchify photos from semantic object-part image datasets and use them for training. Our architecture also incorporates object pose prediction as a novel
auxiliary task which boosts overall performance while providing supplementary information regarding the sketch. We demonstrate SketchParse's abilities (i) on two challenging large-scale sketch datasets (ii) in parsing unseen,
semantically related object categories (iii) in improving fine-grained sketch-based image retrieval. As a novel application, we also outline how SketchParse's output can be used to generate caption-style descriptions for
hand-drawn sketches.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="http://val.cds.iisc.ac.in/deligan/GAN-DeLiGAN.svg"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">Gurumurthy Swaminathan<strong>Ravi Kiran Sarvadevabhatla</strong>, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"> </p>
<p style="font-size:13.4px">In <em>Computer Vision and Pattern Recognition (<strong>CVPR</strong>)</em>, 2017</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> We propose DeLiGAN -- a novel image generative model for diverse and limited training data scenarios. Across a number of image modalities including hand-drawn sketches, we show that DeLiGAN
generates diverse samples. To quantitatively characterize intra-class diversity of generated samples, we also introduce a modified version of "inception-score", a measure found to correlate well with human assessment of generated samples. </p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1706.02071" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#Deliganabs" href="#Deliganabs-list">Abstract</a>
<a href="http://val.cds.iisc.ac.in/deligan/" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#Deligan" href="#Deligan-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="Deligan-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;"> @InProceedings{Gurumurthy_2017_CVPR, <br>
author = {Gurumurthy, Swaminathan and Kiran Sarvadevabhatla, Ravi and Venkatesh Babu, R.}, <br>
title = {DeLiGAN : Generative Adversarial Networks for Diverse and Limited Data}, <br>
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)}, <br>
month = {July}, <br>
year = {2017} <br>
}
</p>
</div>
<div id="Deliganabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
A class of recent approaches for generating images, called Generative Adversarial Networks (GAN), have been used to generate impressively realistic images of objects, bedrooms, handwritten digits and a variety of other image modalities. However, typical GAN-based approaches require large amounts of training data to capture the diversity across the image modality. In this paper, we propose DeLiGAN -- a novel GAN-based architecture for diverse and limited training data scenarios. In our approach, we reparameterize the latent generative space as a mixture model and learn the mixture model's parameters along with those of GAN. This seemingly simple modification to the GAN framework is surprisingly effective and results in models which enable diversity in generated samples although trained with limited data. In our work, we show that DeLiGAN can generate images of handwritten digits, objects and hand-drawn sketches, all using limited amounts of data. To quantitatively characterize intra-class diversity of generated samples, we also introduce a modified version of "inception-score", a measure which has been found to correlate well with human assessment of generated samples.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/SketchGaze.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Object category understanding via eye fixations on freehand sketches</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Sudharshan Suresh, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>IEEE Transactions on Image Processing (<strong>TIP</strong>)</em>, 2017</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> We analyze the results of a free-viewing gaze fixation study conducted on 3904 freehand sketches distributed across 160 object categories. Our analysis shows that fixation sequences exhibit marked consistency within a sketch, across sketches of a category and even across suitably grouped sets of categories. In our paper, we show that this multi-level consistency in the fixation data can be exploited to (a) predict a test sketch's category given only its fixation sequence and (b) build a computational model which predicts part-labels underlying fixations on objects. </p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1703.06554" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchGazeabs" href="#SketchGazeabs-list">Abstract</a>
<a href="http://val.serc.iisc.ernet.in/sketchgaze/" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SketchGaze" href="#SketchGaze-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchGaze-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">@article{sarvadevabhatla2017object,<br>
title={Object category understanding via eye fixations on freehand sketches},<br>
author={Sarvadevabhatla, Ravi Kiran and Suresh, Sudharshan and Babu, R Venkatesh}, <br>
journal={IEEE Transactions on Image Processing}, <br>
volume={26}, <br>
number={5}, <br>
pages={2508--2518}, <br>
year={2017}, <br>
publisher={IEEE} <br>
}</p>
</div>
<div id="SketchGazeabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
The study of eye gaze fixations on photographic images is an active research area. In contrast, the image subcategory of freehand sketches has not received as much attention for such studies. In this paper, we analyze the results of a free-viewing gaze fixation study conducted on 3904 freehand sketches distributed across 160 object categories. Our analysis shows that fixation sequences exhibit marked consistency within a sketch, across sketches of a category and even across suitably grouped sets of categories. This multi-level consistency is remarkable given the variability in depiction and extreme image content sparsity that characterizes hand-drawn object sketches. In our paper, we show that the multi-level consistency in the fixation data can be exploited to (a) predict a test sketch's category given only its fixation sequence and (b) build a computational model which predicts part-labels underlying fixations on objects. We hope that our findings motivate the community to deem sketch-like representations worthy of gaze-based studies vis-a-vis photographic images.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<br>
<h2>2016</h2>
<br>
<p style="margin:-12.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/epivis-epitome-hac.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Analyzing object categories via novel category-ranking measures defined on visual feature emebddings</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Raviteja Meesala, Manjunath Hegde, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>Indian Conference on Graphics, Vision and Image Processing (<strong>ICVGIP</strong>)</em>, Guwahati,India, 2016</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;">Visualizing category-level features using color-coding is impractical when number of categories is large. This paper presents an approach which utilizes the geometrical attributes of per-category feature collections to order the categories. Our approach enables a novel viewpoint for exploring large-scale object category collections.</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="http://dl.acm.org/ft_gateway.cfm?id=3010037&ftid=1822926&dwn=1&CFID=708922244&CFTOKEN=15898682" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#FeatureViz_icvgipabs" href="#FeatureViz_icvgipabs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#FeatureViz_icvgip" href="#FeatureViz_icvgip-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="FeatureViz_icvgip-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">@inproceedings{Sarvadevabhatla:2016:AOC:3009977.3010037, <br>
author = {Sarvadevabhatla, Ravi Kiran and Meesala, Raviteja and Hegde, Manjunath and R., Venkatesh Babu}, <br>
title = {Analyzing Object Categories via Novel Category Ranking Measures Defined on Visual Feature Embeddings}, <br>
booktitle = {Proceedings of the Tenth Indian Conference on Computer Vision, Graphics and Image Processing}, <br>
series = {ICVGIP '16}, <br>
year = {2016}, <br>
isbn = {978-1-4503-4753-2}, <br>
location = {Guwahati, Assam, India}, <br>
pages = {79:1--79:6}, <br>
articleno = {79}, <br>
numpages = {6}, <br>
url = {http://doi.acm.org/10.1145/3009977.3010037}, <br>
doi = {10.1145/3009977.3010037}, <br>
acmid = {3010037}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA}, <br>
} <br>
</p>
</div>
<div id="FeatureViz_icvgipabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Visualizing 2-D/3-D embeddings of image features can help gain an intuitive understanding of the image category landscape. However, popular methods for visualizing such embeddings (e.g. color-coding by category) are impractical when the number of categories is large. To address this and other shortcomings, we propose novel quantitative measures defined on image feature embeddings. Each measure produces a ranked ordering of the categories and provides an intuitive vantage point from which to view the entire set of categories. As an experimental testbed, we use deep features obtained from category-epitomes, a recently introduced minimalist visual representation, across 160 object categories. We embed the features in a visualization-friendly yet similarity-preserving 2-D manifold and analyze the inter/intra-category distributions of these embeddings using the proposed measures. Our analysis demonstrates that the category ordering methods enable new insights for the domain of large-category object representations. Moreover, our ordering measure approach is general in nature and can be applied to any feature-based representation of categories.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<!--div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/new_burst.png"></div-->
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:14px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/overview-accv2016-semantic.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">'Part'-ly first among equals: Part-based benchmarking for state-of-the-art object recognition architectures</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Shanthakumar Venkatraman, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>Asian Conference on Computer Vision (<strong>ACCV</strong>)</em>, Taipei,Taiwan ROC, 2016</p>
<!--p style="font-size:13.4px">Arxiv preprint arXiv:1611.03382</p-->
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;">Top-1/Top-5 error based benchmarking results for large-scale object recognition datasets do not reveal which aspects of recognition problem (robustness to occlusion, loss of global detail) the classifiers are good at. Moreover, the overall approach provides a falsely optimistic picture due to dataset bias. In this paper, we propose a novel semantic-part based dataset and benchmarking approach which overcomes the shortcomings mentioned above.</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<a href="https://arxiv.org/pdf/1611.07703v2.pdf" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#PBBMabs" href="#PBBMabs-list">Abstract</a>
<a href="http://val.serc.iisc.ernet.in/pbbm/" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#PBBM" href="#PBBM-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="PBBM-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@article{DBLP:journals/corr/Sarvadevabhatla16b,<br>
author = {Ravi Kiran Sarvadevabhatla and Shanthakumar Venkatraman and R. Venkatesh Babu},<br>
title = {'Part'ly first among equals: Semantic part-based benchmarking for state-of-the-art object recognition systems},<br>
journal = {CoRR}, <br>
volume = {abs/1611.07703}, <br>
year = {2016}, <br>
url = {http://arxiv.org/abs/1611.07703},<br>
timestamp = {Thu, 01 Dec 2016 19:32:08 +0100}, <br>
biburl = {http://dblp.uni-trier.de/rec/bib/journals/corr/Sarvadevabhatla16b}, <br>
bibsource = {dblp computer science bibliography, http://dblp.org} <br> }
</p>
</div>
<div id="PBBMabs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
An examination of object recognition challenge leaderboards (ILSVRC, PASCAL-VOC) reveals that the top-performing classifiers typically exhibit small differences amongst themselves in terms of error rate/mAP. To better differentiate the top performers, additional criteria are required. Moreover, the (test) images, on which the performance scores are based, predominantly contain fully visible objects. Therefore, `harder' test images, mimicking the challenging conditions (e.g. occlusion) in which humans routinely recognize objects, need to be utilized for benchmarking. To address the concerns mentioned above, we make two contributions. <i>First</i>, we systematically vary the level of local object-part content, global detail and spatial context in images from PASCAL VOC 2010 to create a new benchmarking dataset dubbed PPSS-12. <i>Second</i>, we propose an object-part based benchmarking procedure which quantifies classifiers' robustness to a range of visibility and contextual settings. The benchmarking procedure relies on a semantic similarity measure that naturally addresses potential semantic granularity differences between the category labels in training and test datasets, thus eliminating manual mapping. We use our procedure on the PPSS-12 dataset to benchmark top-performing classifiers trained on the ILSVRC-2012 dataset. Our results show that the proposed benchmarking procedure enables additional differentiation among state-of-the-art object classifiers in terms of their ability to handle missing content and insufficient object detail. Given this capability for additional differentiation, our approach can potentially supplement existing benchmarking procedures used in object recognition challenge leaderboards.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<!--div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/new_burst.png"></div-->
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:9px 1.5% 0 0; width:14.1%;" href=""><img width="100%" height="100%" src="./ravikiran_files/Alexnet-GRU-Overview.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Enabling my robot to play Pictionary: Recurrent Neural Networks for sketch recognition</h4>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px"><strong>Ravi Kiran Sarvadevabhatla</strong>, Jogendra Nath Kundu, R. Venkatesh Babu</p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM Multimedia Conference (<strong>ACMMM</strong>)</em>, Amsterdam, The Netherlands 2016</p>
<!--p style="font-size:13.4px">Arxiv preprint arXiv:1608.07711</p-->
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;">We propose a Recurrent Neural Network architecture which exploits the long-term sequential and structural regularities in sketch stroke data for large-scale recognition of hand-drawn object sketches.</p>
<a href="https://arxiv.org/pdf/1608.03369v1.pdf" target="_blank" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchRNNMM16abs" href="#SketchRNNMM16abs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SketchRNNMM16" href="#SketchRNNMM16-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchRNNMM16-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@inproceedings{Sarvadevabhatla:2016:EMR:2964284.2967220,<br>
author = {Sarvadevabhatla, Ravi Kiran and Kundu, Jogendra and R, Venkatesh Babu},<br>
title = {Enabling My Robot To Play Pictionary: Recurrent Neural Networks For Sketch Recognition},<br>
booktitle = {Proceedings of the 2016 ACM Conference on Multimedia},<br>
year = {2016},<br>
location = {Amsterdam, The Netherlands}, <br>
pages = {247--251}, <br>
url = {http://doi.acm.org/10.1145/2964284.2967220}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA},<br>
}
</p>
</div>
<div id="SketchRNNMM16abs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Freehand sketching is an inherently sequential process. Yet, most approaches for hand-drawn sketch recognition either ignore this sequential aspect or exploit it in an ad-hoc manner. In our work, we propose a recurrent neural network architecture for sketch object recognition which exploits the long-term sequential and structural regularities in stroke data in a scalable manner. Specifically, we introduce a Gated Recurrent Unit based framework which leverages deep sketch features and weighted per-timestep loss to achieve state-of-the-art results on a large database of freehand object sketches across a large number of object categories. The inherently online nature of our framework is especially suited for on-the-fly recognition of objects as they are being drawn. Thus, our framework can enable interesting applications such as camera-equipped robots playing the popular party game Pictionary with human players and generating sparsified yet recognizable sketches of objects.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:15px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/switchnet_full.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">SwiDeN: Convolutional Neural Networks for Depiction Invariant Object Recognition</h4>
<p style="margin:-9.0px 0px 0px 0px;"><strong>Ravi Kiran Sarvadevabhatla</strong>, Shiv Surya, Srinivas SS Kruthiventi, R. Venkatesh Babu</p>
<p style="font-size:13.4px"></p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM Multimedia Conference (<strong>ACMMM</strong>)</em>, Amsterdam, The Netherlands 2016</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> In this paper, we present SwiDeN: our Convolutional Neural Network (CNN) architecture which recognizes objects regardless of how they are visually depicted (line drawing, realistic shaded drawing, photograph etc.) </p>
<a href="https://arxiv.org/pdf/1607.08764v1.pdf" class="buttonTT">Paper</a>
<a href="https://github.com/val-iisc/swiden" target="_blank" class="buttonPP">Code</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SwidenMM16abs" href="#SwidenMM16abs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SwidenMM16" href="#SwidenMM16-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SwidenMM16-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@inproceedings{Sarvadevabhatla:2016:SCN:2964284.2967208,<br>
author = {Sarvadevabhatla, Ravi Kiran and Surya, Shiv and Kruthiventi, Srinivas S S and R., Venkatesh Babu}, <br>
title = {SwiDeN: Convolutional Neural Networks For Depiction Invariant Object Recognition}, <br>
booktitle = {Proceedings of the 2016 ACM on Multimedia Conference}, <br>
series = {MM '16}, <br>
year = {2016}, <br>
isbn = {978-1-4503-3603-1}, <br>
location = {Amsterdam, The Netherlands}, <br>
pages = {187--191}, <br>
numpages = {5}, <br>
url = {http://doi.acm.org/10.1145/2964284.2967208}, <br>
doi = {10.1145/2964284.2967208}, <br>
acmid = {2967208}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA}, <br>
keywords = {convolutional neural networks, deep learning, depiction-invariance, object category recognition}, <br>
}
</p>
</div>
<div id="SwidenMM16abs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Current state of the art object recognition architectures achieve impressive performance but are typically specialized for a single depictive style (e.g. photos only, sketches only). In this paper, we present SwiDeN: our Convolutional Neural Network (CNN) architecture which recognizes objects regardless of how they are visually depicted (line drawing, realistic shaded drawing, photograph etc.). In SwiDeN, we utilize a novel `deep' depictive style-based switching mechanism which appropriately addresses the depiction-specific and depiction-invariant aspects of the problem. We compare SwiDeN with alternative architectures and prior work on a 50-category Photo-Art dataset containing objects depicted in multiple styles. Experimental results show that SwiDeN outperforms other approaches for the depiction-invariant object recognition problem.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:15px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/epitomes-to-wordle.jpg"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Analyzing Structural Characteristics of Object Category Representations From Their Semantic-part Distributions</h4>
<p style="margin:-9.0px 0px 0px 0px;"><strong>Ravi Kiran Sarvadevabhatla</strong>, R. Venkatesh Babu</p>
<p style="font-size:13.4px"></p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM Multimedia Conference (<strong>ACMMM</strong>)</em>, Amsterdam, The Netherlands 2016</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> To analyze intra/inter-category variations of object apperance, we present an approach which represents the relative frequency of object part presence as a category-level word cloud. In this paper, we explore the word cloud style visualizations to characterize <i>category-epitomes</i>, a novel visual representation for objects designed by us in a previous work. </p>
<a href="http://dl.acm.org/ft_gateway.cfm?id=2967190&ftid=1796462&dwn=1&CFID=874283070&CFTOKEN=86936959" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#SketchPartsMM16abs" href="#SketchPartsMM16abs-list">Abstract</a>
<a href="http://val.serc.iisc.ernet.in/eotd/epipartviz/" target="_blank" class="buttonPP">Project page</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#SketchPartsMM16" href="#SketchPartsMM16-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="SketchPartsMM16-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@inproceedings{Sarvadevabhatla:2016:ASC:2964284.2967190,<br>
author = {Sarvadevabhatla, Ravi Kiran and R, Venkatesh Babu}, <br>
title = {Analyzing Structural Characteristics of Object Category Representations From Their Semantic-part Distributions}, <br>
booktitle = {Proceedings of the 2016 ACM on Multimedia Conference}, <br>
series = {MM '16}, <br>
year = {2016}, <br>
isbn = {978-1-4503-3603-1}, <br>
location = {Amsterdam, The Netherlands}, <br>
pages = {97--101}, <br>
numpages = {5}, <br>
url = {http://doi.acm.org/10.1145/2964284.2967190}, <br>
doi = {10.1145/2964284.2967190}, <br>
acmid = {2967190}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA}, <br>
keywords = {freehand sketch, object category representation, semantic part, visualization}, <br>
}
</p>
</div>
<div id="SketchPartsMM16abs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
Studies from neuroscience show that part-mapping computations are employed by human visual system in the process of object recognition. In this paper, we present an approach for analyzing semantic-part characteristics of object category representations. For our experiments, we use category-epitome, a recently proposed sketch-based spatial representation for objects. To enable part-importance analysis, we first obtain semantic-part annotations of hand-drawn sketches originally used to construct the epitomes. We then examine the extent to which the semantic-parts are present in the epitomes of a category and visualize the relative importance of parts as a word cloud. Finally, we show how such word cloud visualizations provide an intuitive understanding of category-level structural trends that exist in the category-epitome object representations. Our method is general in applicability and can also be used to analyze part-based visual object representations for other depiction methods such as photographic images.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2015</h2>
<br>
<p style="margin:-12.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:15px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/primal-sketches-examples.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Eye of the Dragon : Exploring discriminatively minimalist sketch-based abstractions for object categories <strong>[ORAL]</strong></h4>
<p style="margin:-9.0px 0px 0px 0px;"><strong>Ravi Kiran Sarvadevabhatla</strong>, R. Venkatesh Babu</p>
<p style="font-size:13.4px"></p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM Multimedia Conference (<strong>ACMMM</strong>)</em>, Brisbane, Australia 2015</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> In this paper, we introduce a novel visual representation derived from freehand sketches of objects. This representation, called <i>category-epitome</i>, is designed to be a sparsified yet recognizable version of the original sketch. We examine various interesting properties of <i>category-epitome</i>s. </p>
<a href="http://dl.acm.org/ft_gateway.cfm?id=2806230&ftid=1633068&dwn=1&CFID=874283070&CFTOKEN=86936959" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#CatEpitomes15abs" href="#CatEpitomes15abs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#CatEpitomes15" href="#CatEpitomes15-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="CatEpitomes15-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@inproceedings{Sarvadevabhatla:2015:EDE:2733373.2806230,<br>
author = {Sarvadevabhatla, Ravi Kiran and R, Venkatesh Babu}, <br>
title = {Eye of the Dragon: Exploring Discriminatively Minimalist Sketch-based Abstractions for Object Categories}, <br>
booktitle = {Proceedings of the 23rd ACM International Conference on Multimedia}, <br>
series = {MM '15}, <br>
year = {2015},<br>
isbn = {978-1-4503-3459-4}, <br>
location = {Brisbane, Australia}, <br>
pages = {271--280}, <br>
numpages = {10}, <br>
url = {http://doi.acm.org/10.1145/2733373.2806230}, <br>
doi = {10.1145/2733373.2806230}, <br>
acmid = {2806230}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA}, <br>
keywords = {deep learning, freehand sketch, object category recognition}, <br>
}
</p>
</div>
<div id="CatEpitomes15abs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
As a form of visual representation, freehand line sketches are typically studied as an end product of the sketching process. However, from a recognition point of view, one can also study various orderings and properties of the primitive strokes that compose the sketch. Studying sketches in this manner has enabled us to create novel sparse yet discriminative sketch-based representations for object categories which we term category-epitomes. Concurrently, the epitome construction provides a natural measure for quantifying the sparseness underlying the original sketch, which we term epitome-score. We analyze category-epitomes and epitome-scores for hand-drawn sketches from a sketch dataset of 160 object categories commonly encountered in daily life. Our analysis provides a novel viewpoint for examining the complexity of representation for visual object categories.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
<div class="pull-right text-center" style="padding:4px 0.5% 0 0; width:5.0%;" href=""><img width="100%" src="./ravikiran_files/Oral-session-icon.png"></div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2012</h2>
<br>
<p style="margin:-12.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:15px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/captainmayi.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%"> Captain May I -- Proxemics studies examining factors that influence distance between humanoid robots, children and adults during Human-Robot Interaction </h4>
<p style="margin:-9.0px 0px 0px 0px;">Sandra Okita, Victor Ng-Thow-Hing, <strong>Ravi Kiran Sarvadevabhatla</strong></p>
<p style="font-size:13.4px"></p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>ACM/IEEE International Conference on Human-Robot Interaction (<strong>HRI</strong>)</em>, Boston, USA 2012</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> This paper examines how interaction distance between humans and robots vary due to factors such as age, initiator, gesture style, movement announcement. </p>
<a href="http://dl.acm.org/ft_gateway.cfm?id=2157756&ftid=1161758&dwn=1&CFID=875199498&CFTOKEN=10876262" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#Captain12abs" href="#Captain12abs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#Captain12" href="#Captain12-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="Captain12-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">
@inproceedings{Okita:2012:CMI:2157689.2157756, <br>
author = {Okita, Sandra Y. and Ng-Thow-Hing, Victor and Sarvadevabhatla, Ravi Kiran}, <br>
title = {Captain May I?: Proxemics Study Examining Factors That Influence Distance Between Humanoid Robots, Children, and Adults, During Human-robot Interaction}, <br>
booktitle = {Proceedings of the Seventh Annual ACM/IEEE International Conference on Human-Robot Interaction}, <br>
series = {HRI '12}, <br>
year = {2012}, <br>
isbn = {978-1-4503-1063-5}, <br>
location = {Boston, Massachusetts, USA}, <br>
pages = {203--204}, <br>
numpages = {2}, <br>
url = {http://doi.acm.org/10.1145/2157689.2157756}, <br>
doi = {10.1145/2157689.2157756}, <br>
acmid = {2157756}, <br>
publisher = {ACM}, <br>
address = {New York, NY, USA}, <br>
keywords = {human robot interaction, proxemics study, young children}, <br>
}
</p>
</div>
<div id="Captain12abs-list" class="panel-collapse collapse out" style="background-color:#ADEBFF; padding:9px 2.5% 3px 2.5%; border-radius:10px;">
<p style="margin:0px 0px 0px 0px;"></p>
<p style="font-size:14.1px;">
This proxemics study examines whether the physical distance between robots and humans differ based on the following factors: 1) age: children vs. adults, 2) who initiates the approach: humans approaching the robot vs. robot approaching humans, 3) prompting: verbal invitation vs. non-verbal gesture (e.g., beckoning), and 4) informing: announcement vs. permission vs. nothing. Results showed that both verbal and non-verbal prompting had significant influence on physical distance. Physiological data is also used to detect the appropriate timing of approach for a more natural and comfortable interaction.
</p>
<p style="margin:0px 0px 0px 0px;"></p>
</div>
</div>
</li>
</ul>
<p style="margin:-13.5px 0px 0px 0px;"></p>
<br>
<h2>2011</h2>
<br>
<p style="margin:-12.5px 0px 0px 0px;"></p>
<ul class="list-group">
<li class="list-group-item" style="padding:0 0 0.2% 1.1%;border:1px solid">
<div class="media">
<div class="pull-left text-center" style="padding:15px 1.5% 0 0; width:14.1%;" href=""><img width="100%" src="./ravikiran_files/facialexprec.png"></div>
<div class="pull-left text-left" style="padding:0% 0% 0% 0.1%; width:76.4%;">
<p style="margin:-3px 0px 0px 0px;"></p>
<h4 style="font-size:14.1px; line-height:120%">Adaptive Facial Expression Recognition Using Inter-modal Top-down context </h4>
<p style="margin:-9.0px 0px 0px 0px;"><strong>Ravi Kiran Sarvadevabhatla</strong>, Victor Ng-Thow-Hing, Mitchel Benovoy, Sam Musallam</p>
<p style="font-size:13.4px"></p>
<p style="margin:-10.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px">In <em>International Conference on Multimodal Interaction(<strong>ICMI</strong>) </em>, Alicante, Spain 2011</p>
<p style="margin:-9.0px 0px 0px 0px;"></p>
<p style="font-size:13.4px; color:#E62E00;"> This paper describes an approach for facial expression recognition which takes the effect of other concurrently active modalities (e.g. talking while emoting the expression) into account. </p>
<a href="http://dl.acm.org/ft_gateway.cfm?id=2070488&ftid=1058231&dwn=1&CFID=875199498&CFTOKEN=10876262" class="buttonTT">Paper</a>
<a class="buttonAA" data-toggle="collapse" data-parent="#ExpRec11abs" href="#ExpRec11abs-list">Abstract</a>
<a class="buttonSS" data-toggle="collapse" data-parent="#ExpRec11" href="#ExpRec11-list">Bibtex</a><br>
<p style="margin:1px 0px 0px 0px;"></p>
<div id="ExpRec11-list" class="panel-collapse collapse out" style="background-color:#FFE0C2; padding:0% 3% 1% 3%; border-radius:10px;">
<p style="font-size:15px;">