-
Notifications
You must be signed in to change notification settings - Fork 0
/
2024.html
1465 lines (1309 loc) · 76.1 KB
/
2024.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
---
title: "EPIC-KITCHENS Dataset"
layout: default
---
<!-- About -->
<section class="bg-light" id="about">
<div class="container">
<div class="row">
<div class="col-lg-12">
<video autoplay muted loop width="100%">
<source src="{{ site.baseurl }}/static/videos/03x03_videoWall.webm" type="video/webm">
<source src="{{ site.baseurl }}/static/videos/03x03_videoWall.mp4" type="video/mp4">
<source src="{{ site.baseurl }}/static/videos/03x03_videoWall.mp4" type="video/mp4">
Sorry, we cannot display the EPIC-KITCHENS-100 video wall as
your browser doesn't support HTML5 video.
</video>
</div>
</div>
<div class="row">
<div class="col-lg-12">
<h2 class="section-heading text-uppercase">News</h2>
<h5 class="text-muted" style="text-align:left;">
<ul>
<li>Feb 2024: Leaderboards Open for the 2024 Challenge</li>
<li>Jan 2024: EPIC-KITCHENS team has joined the <a href="https://egovis.github.io">EgoVis Board, and will be co-organising a Joint workshop @CVPR in Seattle this June</a></li>
<li>27 June 2023: <a href="#results">Results are publicly announced</a> and leaderboards are open</li>
<li>1 June 2023: Leaderboards are closed</li>
<li>21 May 2023: 10 days to go until leaderboards close. We have now increased number of submission/day to 2.</li>
<li>23 Jan 2023: Leaderboards are now open. Check <a href="#challenges">challenges</a> for details.</li>
<li>23 Jan 2023: <a href="https://epic-kitchens.github.io/epic-sounds">EPIC-SOUNDS dataset now public</a>. ArXiv paper describing the dataset is coming out soon</li>
<li>16 Aug 2022: Announcing <a href="https://epic-kitchens.github.io/VISOR">VISOR (Segmentations and Object Relations) annotations</a></li>
<li>31 July 2022: The <a href="./Reports/EPIC-KITCHENS-Challenges-2022-Report.pdf">2022 Challenges Report is now available</a></li>
<li>3 July 2022: <a href="https://epic-kitchens.github.io/2022.html#results">Results of the 2022 challenges</a> are now public</li>
<li>Watch the dataset's <a href="https://www.youtube.com/watch?v=8IzkrWAfAGg" target="_blank">trailer</a> and <a href="https://www.youtube.com/watch?v=MUlyXDDzbZU&t=4s">video demonstration</a> on YouTube</li>
</ul>
</h5>
</div>
</div>
<div class="row">
<!-- news column -->
<div class="col-md-4">
<h4 class="service-heading">What is EPIC-KITCHENS-100?</h4>
<p class="text-muted">The <b>large-scale dataset in first-person (egocentric) vision</b>; multi-faceted, audio-visual, <b>non-scripted</b> recordings in native environments - i.e. the wearers' homes, capturing all daily activities in the kitchen over multiple days. Annotations are collected using a novel 'Pause-and-Talk' narration interface.</p>
</div>
<!-- characteristics column -->
<div class="col-md-4">
<h4 class="service-heading">Characteristics</h4>
<ul class="text-muted">
<li class="text-muted">45 kitchens - 4 cities</li>
<li class="text-muted">Head-mounted camera</li>
<li class="text-muted"><b>100</b> hours of recording - Full HD</li>
<li class="text-muted">20M frames</li>
<li class="text-muted">Multi-language narrations</li>
<li class="text-muted">90K action segments</li>
<li class="text-muted">20K unique narrations</li>
<li class="text-muted">97 verb classes, 300 noun classes</li>
<li class="text-muted">5 challenges</li>
</ul>
</div>
<!-- udated column -->
<div class="col-md-4">
<h4 class="service-heading">Previous versions...</h4>
<ul class="text-muted">
<li class="text-muted">2022 Challenges: <a href="https://epic-kitchens.github.io/2022.html#results">Results</a>, <a href="./Reports/EPIC-KITCHENS-Challenges-2022-Report.pdf">Tech Report</a> [<a href="./Reports/2022-bibtex.txt">Bibtex</a>]</li>
<li class="text-muted">2021 Challenges: <a href="https://epic-kitchens.github.io/2021.html#results">Results</a>, <a href="./Reports/EPIC-KITCHENS-Challenges-2021-Report.pdf">Tech Report</a> [<a href="./Reports/2021-bibtex.txt">Bibtex</a>]</li>
<li class="text-muted">Refer to <a href="2020-55.html">EPIC-KITCHENS-55</a> for previous version</li>
<li class="text-muted">2020 Challenges: <a href="https://epic-kitchens.github.io/2020-55.html#results">Results</a>, <a href="./Reports/EPIC-KITCHENS-Challenges-2020-Report.pdf">Tech Report</a> [<a href="./Reports/2020-bibtex.txt">Bibtex</a>]</li>
<li class="text-muted">2019 Challenges: <a href="https://epic-kitchens.github.io/2019.html#results">Results</a>, <a href="Reports/EPIC-Kitchens-Challenges-2019-Report.pdf">Tech Report</a> [<a href="./Reports/2019-bibtex.txt">Bibtex</a>]</li>
</ul>
</div>
</div>
<!-- video banner row -->
</div>
</section>
<!-- Stats -->
<section id="stats">
<div class="container">
<div class="row">
<div class="col-lg-12 text-center">
<h2 class="section-heading text-uppercase">EPIC-KITCHENS-100 Stats and Figures</h2>
<h3 class="section-subheading text-muted">Some graphical representations of our dataset and annotations</h3>
</div>
</div>
<div class="row justify-content-md-center text-center">
<div class="col-md centered" style="padding:1rem;">
<img src="{{ site.baseurl }}/static/img/stats-figures/verb_categories.png" style="width: 100%" class="img-responsive"/>
</div>
</div>
<div class="row justify-content-md-center text-center">
<div class="col-md centered" style="padding:1rem;">
<img src="{{ site.baseurl }}/static/img/stats-figures/noun_categories.png" style="width: 100%" class="img-responsive"/>
</div>
</div>
<div class="row justify-content-md-center text-center">
<div class="col-md-6 centered" style="padding:1rem;">
<img src="{{ site.baseurl }}/static/img/stats-figures/pipeline.png" style="width: 100%" class="img-responsive"/>
<h4>Annotation Pipeline</h4>
</div>
<div class="col-md-6 centered" style="padding:1rem; vertical-align:bottom">
<!-- TO ADD GRAPH: replace div below, ex: above <img> tag -->
<img src="{{ site.baseurl }}/static/img/stats-figures/masks.png" style="width: 100%" class="img-responsive"/>
<h4>Automatic Annotations</h4>
</div>
</div>
<!-- <div class="row justify-content-md-center text-center">
<div class="col-md-4 centered" style="padding:1rem;">
<div id="graph4" style="width: 100%" class="img-responsive"></div>
<h4>Resolution</h4>
</div>
<div class="col-md-4 centered" style="padding:1rem;">
<div id="graph5" style="width: 100%" class="img-responsive"></div>
<h4>Number of Frames</h4>
</div>
<div class="col-md-4 centered" style="padding:1rem;">
<div id="graph6" style="width: 100%" class="img-responsive"></div>
<h4>Total number of hours</h4>
</div>
</div>
<div class="row justify-content-md-center text-center">
<div class="col-md-4 centered" style="padding:1rem;">
<div id="graph7" style="width: 100%" class="img-responsive"></div>
<h4>Number of annotators<br/>used per video</h4>
</div>
<div class="col-md-4 centered" style="padding:1rem;">
<div id="graph8" style="width: 100%" class="img-responsive"></div>
<h4>Splits</h4>
</div>
</div> -->
<!-- <div class="col-md-6">
<div class="card" style="border: solid 2px; background-color: #373435ff; margin-bottom:5px;">
<h1 style=" color: white; text-decoration: underline; text-decoration-color: #ed323eff;"> Baseline Models </h1>
<div id="graph9"></div>
</div>
<div class="card" style="border: solid 2px; background-color: #ed323eff;">
<h1 style=" color: white; text-decoration: underline; text-decoration-color: #373435ff;"> State of the Art Results </h1>
<div id="graph10"></div>
</div>
</div> -->
</div>
</section>
<section class="bg-light" id="downloads">
<div class="container">
<div class="row">
<div class="col-md-12 text-center">
<h2 class="section-heading text-uppercase">Download</h2>
<h3 class="section-subheading text-muted">Dataset publicly available for research purposes</h3>
</div>
</div>
<div class="row">
<div class="col-md-12">
<h4 class="section-subheading" id="downloadFiles">Data and Download Script</h4>
<hr/>
<p><b>Erratum [Important]:</b> We have recently detected an error in our pre-extracted RGB and Optical flow frames for two videos in our dataset. This does not affect the videos themselves or any of the annotations in this github. However, if you've been using our pre-extracted frames, you can fix the error at your end by <a href="https://github.com/epic-kitchens/epic-kitchens-100-annotations/blob/master/README.md#erratum">following the instructions in this link.</a></p><hr/>
<p>
Extended Sequences (+RGB Frames, Flow Frames, Gyroscope + accelerometer data): Available at <a href="http://dx.doi.org/10.5523/bris.2g1n6qdydwa9u22shpxqzp0t8m">Data.Bris servers (740GB zipped)</a> or <a href="https://academictorrents.com/details/c92b4a3cd3834e9af9666ac82379ff15ca289a83">via Academic Torrents</a>
</p>
<p>
Original Sequences (+RGB and Flow Frames): Available at <a href="http://dx.doi.org/10.5523/bris.3h91syskeag572hl6tvuovwv4d">Data.Bris servers (1.1TB zipped)</a> or <a href="https://academictorrents.com/details/d08f4591d1865bbe3436d1eb25ed55aae8b8f043">via Academic Torrents</a>
</p>
<p> Automatic annotations (masks, hands and objects): Available for download at <a href="https://data.bris.ac.uk/data/dataset/3l8eci2oqgst92n14w2yqi5ytu">Data.Bris server (10 GB)</a>. We also have two Repos that will allow you to visualise and utilise these automatic annotations for <a href="https://github.com/epic-kitchens/epic-kitchens-100-object-masks">hand-objects</a> as well as <a href="https://github.com/epic-kitchens/epic-kitchens-100-object-masks">masks</a>.</p>
<p> We also offer a <a href="https://github.com/epic-kitchens/epic-kitchens-download-scripts">python script to download</a> various parts of the dataset</p>
<h4 class="section-subheading">Annotations and Pipeline</h4>
<p>All action segment annotations (Train/Val/Test) for all challenges are available at <a href="https://github.com/epic-kitchens/epic-kitchens-100-annotations">EPIC-KITCHENS-100-annotations repo</a></p>
<p>All masks/segmentation annotations are available from <a href="https://epic-kitchens.github.io/VISOR/#downloads">VISOR webpage</a></p>
<p>Audio-only annotations are available from <a href="https://epic-kitchens.github.io/epic-sounds/#downloads">EPIC-SOUNDS webpage</a></p>
<p>Code to visualise and utilise automatic annotations is available for both <a href="https://github.com/epic-kitchens/epic-kitchens-100-object-masks">object masks</a> and <a href="https://github.com/epic-kitchens/epic-kitchens-100-hand-object-bboxes">hand-object detections</a>.</p>
<p>The EPIC Narrator, used to collect narrations for EPIC-KITCHENS-100 is open-sourced at <a href="https://github.com/epic-kitchens/epic-kitchens-100-narrator">EPIC-Narrator repo</a></p>
<section>
<h4 class="section-subheading">Publication(s)</h4>
<p>Cite our IJCV paper (Open Access 2021 - Published 2022):
<a href="https://link.springer.com/content/pdf/10.1007/s11263-021-01531-2.pdf">PDF</a> or <a href="https://arxiv.org/abs/2006.13256">Arxiv</a>:</p>
<pre class="bibtex"><code>@ARTICLE{Damen2022RESCALING,
title={Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100},
author={Damen, Dima and Doughty, Hazel and Farinella, Giovanni Maria and Furnari, Antonino
and Ma, Jian and Kazakos, Evangelos and Moltisanti, Davide and Munro, Jonathan
and Perrett, Toby and Price, Will and Wray, Michael},
journal = {International Journal of Computer Vision (IJCV)},
year = {2022},
volume = {130},
pages = {33–55},
Url = {https://doi.org/10.1007/s11263-021-01531-2}
} </code></pre>
<p>Additionally, cite the original paper
(<a href="https://arxiv.org/abs/1804.02748">available now on Arxiv</a> and
<a href="http://openaccess.thecvf.com/content_ECCV_2018/html/Dima_Damen_Scaling_Egocentric_Vision_ECCV_2018_paper.html">the CVF</a>):</p>
<pre class="bibtex"><code>
@INPROCEEDINGS{Damen2018EPICKITCHENS,
title={Scaling Egocentric Vision: The EPIC-KITCHENS Dataset},
author={Damen, Dima and Doughty, Hazel and Farinella, Giovanni Maria and Fidler, Sanja and
Furnari, Antonino and Kazakos, Evangelos and Moltisanti, Davide and Munro, Jonathan
and Perrett, Toby and Price, Will and Wray, Michael},
booktitle={European Conference on Computer Vision (ECCV)},
year={2018}
} </code></pre>
<p>An extended journal paper is avaliable at:
(available now on <a href="https://ieeexplore.ieee.org/document/9084270">IEEE</a> and a <a href="https://arxiv.org/abs/2005.00343">preprint on Arxiv</a> and ):</p>
<pre class="bibtex"><code>
@ARTICLE{Damen2021PAMI,
title={The EPIC-KITCHENS Dataset: Collection, Challenges and Baselines},
author={Damen, Dima and Doughty, Hazel and Farinella, Giovanni Maria and Fidler, Sanja and
Furnari, Antonino and Kazakos, Evangelos and Moltisanti, Davide and Munro, Jonathan
and Perrett, Toby and Price, Will and Wray, Michael},
journal={IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)},
year={2021},
volume={43},
number={11},
pages={4125-4141},
doi={10.1109/TPAMI.2020.2991965}
} </code></pre>
</section>
</div>
</div>
<div class="row">
<div class="col-md-12">
<h4 class="section-subheading">Disclaimer </h4>
<p>EPIC-KITCHENS-55 and EPIC-KITCHENS-100 were collected as a tool for research in computer vision. The dataset may have unintended biases (including those of a societal, gender or racial nature).</p>
</div>
</div>
<div class="row">
<div class="col-md-12">
<h4 class="section-subheading">Copyright <img alt="Creative Commons License" style="border-width:1px;float:left;margin-right:15px;margin-bottom:0px;" src="https://i.creativecommons.org/l/by-nc/3.0/88x31.png"/></h4>
<p>
All datasets and benchmarks on this page are copyright by us and published under the <a rel="license" href="https://creativecommons.org/licenses/by-nc/4.0/">Creative Commons Attribution-NonCommercial 4.0 International</a> License. This means that you must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. You may not use the material for commercial purposes.
</p>
<p>For commercial licenses of EPIC-KITCHENS and any of its annotations, email us at <a href="mailto:uob-epic-kitchens@bristol.ac.uk">uob-epic-kitchens@bristol.ac.uk</a></p>
</div>
</div>
</div>
</section>
<!-- Challenges -->
<section id="challenges">
<div class="container">
<div class="row">
<div class="col-md-12 text-center">
<h2 class="section-heading text-uppercase">EPIC-KITCHENS-100 2024 Challenges</h2>
<h3 class="section-subheading text-muted">Challenge Details with links to ★NEW★ Codalab Leaderboards</h3>
</div>
</div>
<div class="row">
<div class="col-md-12">
<b>leaderboards are now open for the <b>challenge phase from Feb 2024</b>.<br/>
<b>In 2024, we have 9 open challenges. These are</b>
<ul>
<li><a href="#vos">Semi-Supervised Video Object Segmentation Challenge</a></li>
<li><a href="#hos">Hand-Object Segmentation Challenge</a></li>
<li><a href="#tracking">TREK-150 Object Tracking Challenge</a></li>
<li><a href="#sounds">EPIC-SOUNDS Audio-Based Interaction Recognition</a></li>
<li><a href="#sounds2">EPIC-SOUNDS Audio-Based Interaction <b>Detection</b> **NEW in 2024** </a></li>
<li><a href="#challenge-action-recognition">Action Recognition</a></li>
<li><a href="#challenge-action-detection">Action Detection</a></li>
<li><a href="#challenge-action-anticipation">Action Anticipation</a></li>
<li><a href="#challenge-domain-adaptation">UDA for Action Recognition</a></li>
<li><a href="#challenge-action-retrieval">Multi-Instance Retrieval</a></li>
</ul>
<h4 class="subheading">EPIC-Kitchens 2024 Challenges</h4>
<div class="row">
<div class="col-md-3">
Feb 10th 2024,
</div>
<div class="col-md-9">
All leaderboards are open
</div>
</div>
<div class="row">
<div class="col-md-3">
June 1st 2024,
</div>
<div class="col-md-9">
Server Submission Deadline at 00:00:00 UTC
</div>
</div>
<div class="row">
<div class="col-md-3">
June 6th 2024,
</div>
<div class="col-md-9">
Deadline for Submission of Technical Reports on CMT <a href="https://cmt3.research.microsoft.com/EgoVis2024/">HERE</a>
</div>
</div>
<div class="row">
<div class="col-md-3">
June 17 2024,
</div>
<div class="col-md-9">
Results announced at 1st EgoVis workshop in Seattle <a href="https://egovis.github.io">EPIC@EgoVis@CVPR2024 workshop</a>
</div>
</div>
</div>
</div>
<br/>
<div class="row">
<!-- <div class="col-md-12">
<h4 class="subheading">Open Testing Phase Guidelines</h4>
<p>The five leaderboards are available for testing. This is not a formal challenge. Please revisit the website next January for 2023 challenge guideilnes. The CodaLab server pages detail submission format and evaluation metrics. </p>
<p>To <b>submit to any of the five competitions</b>, you need to register an account for that challenge using a valid institute (university/company) email address and <a href="https://forms.office.com/r/u3Hjpvb5Ks">fill this form with your team's details</a>. A single registration per research team is allowed. We perform a manual check for each submission, and expect to accept registrations within 2 working days.</p>
<p>For all challenges the maximum submissions per day is limited to 1, and the overall maximum number of submissions per team is limited to 50 overall, submitted once a day. This includes any failed submissions due to formats - please do not contact us to ask for increasing this limit.</p>
<p>To <b>submit</b> your results, follow the JSON submission format, upload your results and give time for the evaluation to complete (in the order of several minutes). <b>Note our new rules on declaring the supervision level, given our proposed scale, for each submission.</b> After the evaluation is complete, the results automatically appear on the public leaderboards but you are allowed to withdraw these at any point in time.</p>
</div>
</div>-->
<div class="row">
<div class="col-md-12">
<h4 class="subheading">Challenges Guidelines</h4>
<p>The <b>nine</b> challenges below and their test sets and evaluation servers are available via CodaLab. The leaderboards will decide the winners for each individual challenge. For each challenge, the CodaLab server page details submission format and evaluation metrics. </p>
<p>This year, we offer <b>four</b> new challenges in: Semi-Supervised Video Object Segmentation using the <a href="http://epic-kitchens.github.io/VISOR/">VISOR</a> annotations, Hand-object-segmentations using the <a href="http://epic-kitchens.github.io/VISOR/">VISOR</a> annotations, single-object tracking and audio-based action recognition using the <a href="https://epic-kitchens.github.io/epic-sounds/">epic-sounds</a> dataset.</p>
<p>To <b>enter any of the nine competitions</b>, you need to register an account for that challenge using a valid institute (university/company) email address. To enable your account, <a href="https://forms.office.com/r/u3Hjpvb5Ks">fill this form with your team's details</a>. A single registration per research team is allowed. We perform a manual check for each submission, and expect to accept registrations within 2 working days.</p>
<p>For all challenges, the maximum submissions per day is limited to 1, and the overall maximum number of submissions per team is limited to 50 overall, submitted once a day. This includes any failed submissions due to formats - please do not contact us to ask for increasing this limit.</p>
<p>To <b>submit</b> your results, follow the JSON submission format, upload your results and give time for the evaluation to complete (in the order of several minutes). <b>Note our new rules on declaring the supervision level, given our proposed scale, for each submission.</b> After the evaluation is complete, the results automatically appear on the public leaderboards but you are allowed to withdraw these at any point in time.</p>
<p>For the Semi-Supervised VOS challenge, an additional step is required by submitting the model/code using a Docker. Please refer to this challenge's section for more details.</p>
<p>To <b>participate</b> in the challenge, you need to have your results on the public leaderboard, along with an informative team name (that represents your institute or the collection of institutes participating in the work), as well as brief information on your method. You are also required to submit a report on your method (in the form of 2-6 pages) to the EPIC@CVPR workshop by June 6th, detailing your entry's technical details.</p>
<p>To <b>submit your technical report</b>, use the <a href="https://cvpr.thecvf.com/Conferences/2024/AuthorGuidelines">CVPR 2024 camera ready author kit (no blind submission)</a>, and submit a report of 2-6 pages inclusive of any references to the EgoVis@CVPR2024 CMT3 website <a href="https://cmt3.research.microsoft.com/EgoVis2024/">Link Here</a>. Please select the track "EPIC-Kitchens 2024 Challenges - Technical Papers" when submitting your pdf. These technical reports will be combined into an overall report of the EPIC-Kitchens challenges.</p>
<p>Make the most of the starter packs available with the challenges, and should you have any questions, please use our info email <a href="mailto:uob-epic-kitchens@bristol.ac.uk">uob-epic-kitchens@bristol.ac.uk</a></p>
</div>
</div>
<div class="row">
<div class="col-md-12">
<h4 class="subheading">Frequently Asked Questions</h4>
<ul>
<li>
<b>Q.</b> Who is allowed to participate?<br/>
<b>A.</b> Any researcher, whether in academia or industry, is invited to participate in the EPIC-KITCHENS-100 2022 Challenges and open testing phase. We only request a valid official email address, associated with an institution, for registraton. This ensures we limit the number of submissions per team. Do not use your personal email for registration. You registration request will be declined without further explanation.
</li>
<li>
<b>Q.</b> Can I participate in more than one challenge? Do I need to register separately for each challenge?<br/>
<b>A.</b> Yes, and yes. You can participate in all challenges but you need to register separately for each. Winners for each challenge will be announced and there will be no 'overall' winner across challenges.
</li>
<li>
<b>Q.</b> Can I get additional submission limits to debug my file format?<br/>
<b>A.</b> No. Please check your format in advance. We do not offer additional allowance for submission failures.
</li>
<li>
<b>Q.</b> Can I participate in the challenge but not submit a report describing my method?<br/>
<b>A.</b> No. Entries to the challenge will only be considered if a technical report is submitted on time. This should not affect later publications of your method if you restrict your report to 4 pages including references.
</li>
<li>
<b>Q.</b> What happens after the server closes?<br/>
<b>A.</b> We will close the test server after the deadline and until the workshop, but open it again later for normal submissions of future papers.
</li>
<li>
<b>Q.</b> Are there any prizes given?<br/>
<b>A.</b> There are no monetary prizes. Certificates will be awarded.
</li>
<li>
<b>Q.</b> What is the submission limit?<br/>
<b>A.</b> Once a day, with a maximum of 50 submissions.
</li>
<li>
<b>Q.</b> Can you give me access to pre-trained models of baselines?<br/>
<b>A.</b> Yes, check the details of this per challenge.
</li>
<li>
<b>Q.</b> Can I use the large-scale <a href="https://ego4d-data.org/">Ego4D dataset</a> in the EPIC-KITCHENS challenges?<br/>
<b>A.</b> Yes, you can. You need to explicitly declare that in the Supervision Level Scales (SLS) of your submission as follows.
<ul>
<li>When using Ego4D for pretraining, set the Pretraining (PT) flag to level 3 when pretraining in self-supervised manner, and 4 when pretraining with supervision.</li>
<li>When using Ego4D labels as training data, set the Training Data (TD) flag to 4 if using Ego4D solely andd 5 if using Ego4D with other private datasets</li>
</ul>
</li>
</ul>
</div>
</div>
<div class="row">
<div class="col-md-12">
<section class="challenge" id="vos">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Semi-Supervised Video Object Segmentation</h5>
<p class="text-muted">
This challenge uses the newly published <a href="http://epic-kitchens.github.io/VISOR/">VISOR annotations</a>.<br/>
<b>Task. </b>
Given a sub-sequence of frames with M object masks in the first frame, the goal is to segment these through the remaining frames. Other objects not present in the first frame of the sub-sequence are excluded from this benchmark. Note that any of the M objects can be occluded or out of view, and can reappear during the subsequene.
</p>
<p><b>Validation Submission Phase. </b> Between Jan and end of March, interested parties can submit their results on the validation set to the leaderboard. Successful top submissions will be invited to evaluate on the hidden Test Set in the next phase</p>
<p><b>Test Submission (Final Test).</b> Interested participants who have already registered their interest through submitting their models to the Validation Server are now invited to evaluate their models on the test set. As we do not plan to release the first frame annotations for the test set, to avoid overfitting, you are asked to submit the trained models for inference through a Docker that will be evaluated at our end.
We have now announced <a href="https://github.com/epic-kitchens/C6-SemiVOS#instructions-to-evaluate-on-the-test-set"><b>Instructions to evaluate on the test set</b></a>.
You can now contact us at any point to try your model on the test server (<b>Open for submissions until 1st of June</b>). Winners of the challenge are only decided based on the performance on the hidden test set</p>
<b>Evaluation metrics. </b> Average of J-Decay and F-Recall as with other video object segmentation benchmarks.<br/>
<b>Lead:</b> Ahmad DarKhalil (University of Bristol) and Richard Higgins (Univ of Michigan)
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C6-SemiVOS">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C6-SemiVOS">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9767#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/9767#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9767#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/9767#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/VOS.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="hos">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Hand Object Segmentation</h5>
<p class="text-muted">
<b>Task. </b>
Segment hands and the corresponding objects being interacted with in images.<br/>
<b>Training input. </b> A set of images with hand masks, object masks and hand-object relations. <br/>
<b>Testing input. </b> A set of images in the test set.<br/>
<b>Splits. </b> VISOR Train and validation for training, evaluated on the VISOR test split. <br/>
<b>Evaluation metrics. </b> Mean Average Precision (mAP) @ IOU 0.1 to 0.5.<br/>
<b>Lead:</b> Dandan Shan (University of Michigan), Richard Higgins (University of Michigan) and David Fouhey (University of Michigan)
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C7-VISOR-HoS">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C7-VISOR-HoS">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9969#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/9969#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9969#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/9969#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/HOS.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="tracking">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">TREK-150 Object Tracking</h5>
<p class="text-muted">
<b>Task. </b>
The challenge requires to track an object instance through a first-person video sequence. The challenge will be carried out on the <a href="https://machinelearning.uniud.it/datasets/trek150/">TREK-150 dataset</a>, a subset of the EPIC-KITCHENS dataset labeled for single object tracking. More information on the dataset and downloads can be found <a href="https://machinelearning.uniud.it/datasets/trek150/">here</a>.
<br/>
<b>Lead: Matteo Dunnhofer (University of Udine), Antonino Furnari (University of Catania), Giovanni Maria Farinella (University of Catania), Christian Micheloni (University of Udine)</b>
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C8-Object-Tracking">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C8-Object-Tracking">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9597#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/9597#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9597#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/9597#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/tracker.jpg" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="sounds">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">EPIC-SOUNDS: Audio-Based Interaction Recognition</h5>
<p class="text-muted">
<b>Task. </b> Assign an audio-class label to a trimmed segment, indicating the class of interaction taking place in the video. This uses the <a href="http://epic-kitchens.github.io/epic-sounds/">EPIC-SOUNDS</a> dataset annotations. <br/>
<b>Training input. </b> A set of trimmed audio segments, each annotated with one of 44 class labels. <br/>
<b>Testing input. </b> A set of trimmed unlabelled audio segments. <br/>
<b>Splits. </b> Train/Val/Test splits are available
<a href="https://github.com/epic-kitchens/epic-sounds-annotations" target="_blank">here</a>. <br/>
<b>Evaluation metrics. </b> Top-1/5 accuracy for audio class, on the target test set, as well as mAC mAP and mAUC for class-balanced metrics.<br/>
<b>Leads:</b> Jaesung Huh (University of Oxford), Jacob Chalk (University of Bristol), Evangelos Kazkos (ex- Univ of Bristol), Dima Damen (University of Bristol) and Andrew Zisserman (University of Oxford)
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C9-epic-sounds">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C9-epic-sounds">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9729#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/9729#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/9729#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/9729#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Difference from visual and audio annotations is demonstrated in this teaser figure</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/epic-sounds.jpg" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="sounds2">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">EPIC-SOUNDS: Audio-Based Interaction <b>Detection</b></h5>
<p class="text-muted">
<b>Task. </b> Given an untrimmed video, find all temporal segments that indicate an interaction event and assign these a label. This challenge uses the <a href="http://epic-kitchens.github.io/epic-sounds/">EPIC-SOUNDS</a> dataset annotations. <br/>
<b>Training input. </b> A set of untrimmed videos, each video annotated with a number of, potentially overlapping, start-end times and a label out of 44 class labels. <br/>
<b>Testing input. </b> A set of untrimmed videos. <br/>
<b>Splits. </b> Train (shared with recognition) /Val (shared with recognition) /Test (distinct from recognition) splits are available
<a href="https://github.com/epic-kitchens/epic-sounds-annotations" target="_blank">here</a>. <br/>
<b>Evaluation metrics. </b> average mAP, on the target test set<br/>
<b>Leads:</b> Jacob Chalk (University of Bristol), Jaesung Huh (University of Oxford), Evangelos Kazkos (ex- Univ of Bristol), Dima Damen (University of Bristol) and Andrew Zisserman (University of Oxford)
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C10-epic-sounds-detection">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C10-epic-sounds-detection">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/17921#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/17921#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/17921#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/17921#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Difference from visual and audio annotations is demonstrated in this teaser figure</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/epic-sounds-detection.jpg" style="width: 100%" class="img-responsive"/>
</figure>
</section>
</div>
</div>
<div class="row">
<div class="col-md-12">
<h4 class="subheading">Original Challenges/Leaderboard Details</h4>
<ul>
<li><a href="#challenge-action-recognition">Action Recognition</a></li>
<li><a href="#challenge-action-detection">Action Detection</a></li>
<li><a href="#challenge-action-anticipation">Action Anticipation</a></li>
<li><a href="#challenge-domain-adaptation">Domain Adaptation for Action Recognition</a></li>
<li><a href="#challenge-action-retrieval">Multi-Instance Retrieval</a></li>
</ul>
<p class="text-muted">
<b>Splits. </b> The dataset is split in train/validation/test sets, with a ratio of roughly 75/10/15. <br/>
The action recognition, detection and anticipation challenges use all the splits. <br/>
The unsupservised domain adaptation and action retrieval challenges use different splits as detailed below. <br/>
You can download all the necessary annotations <a href="https://github.com/epic-kitchens/epic-kitchens-100-annotations" target="_blank">here</a>. <br/>
You can find more details about the splits in <a href="https://arxiv.org/pdf/2006.13256.pdf" target="_blank">our paper</a>.
</p>
<p class="text-muted">
<b>Evaluation. </b> All challenges are evaluated considering all segments in the Test split.
The action recognition and anticipation challenges are additionally evaluated considering unseen participants and tail classes. These are automatically evaluated in the scripts and you do not need to do anything specific to report these.<br/>
<b>Unseen participants. </b> The validation and test sets contain participants that are not present in the train set.
There are 2 unseen participants in the validation set, and another 3 participants in the test set.
The corresponding action segments are 1,065 and 4,110 respectively. <br/>
<b>Tail classes. </b> These are the set of smallest classes whose instances account for 20% of the total number of instances in
training. A tail action class contains either a tail verb class or a tail noun class.
<br/><br/>
</p>
<section class="challenge" id="challenge-action-recognition">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Action Recognition</h5>
<p class="text-muted">
<b>Task. </b>
Assign a (verb, noun) label to a trimmed segment.<br/>
<b>Training input (strong supervision). </b> A set of trimmed action segments, each annotated with a (verb, noun) label. <br/>
<b>Training input (weak supervision). </b> A set of <i>untrimmed videos</i>, each annotated with a list of (timestamp, verb, noun) labels.
Note that for each action you are given a <i>single, roughly aligned</i> timestamp, i.e. one timestamp located
around the action. Timestamps may be located over background frames or frames belonging to another action. <br/>
<b>Testing input. </b> A set of trimmed unlabelled action segments. <br/>
<b>Splits. </b> Train and validation for training, evaluated on the test split. <br/>
<b>Evaluation metrics. </b> Top-1/5 accuracy for verb, noun and action (verb+noun), calculated for all segments as well as
unseen participants and tail classes.
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C1-Action-Recognition">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C1-Action-Recognition">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/776#learn_the_details">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/776#learn_the_details">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/776#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/776#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/ar.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="challenge-action-detection">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Action Detection</h5>
<p class="text-muted">
<b>Task. </b>
Detect the start and the end of each action in an <i>untrimmed</i> video. Assign a (verb, noun) label to each
detected segment. <br/>
<b>Training input. </b> A set of trimmed action segments, each annotated with a (verb, noun) label. <br/>
<b>Testing input. </b> A set of <i>untrimmed</i> videos. <u>Important:</u> You are not allowed to use the knowledge of trimmed segments in the test set when reporting for this challenge.<br/>
<b>Splits. </b> Train and validation for training, evaluated on the test split. <br/>
<b>Evaluation metrics. </b> Mean Average Precision (mAP) @ IOU 0.1 to 0.5.
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C2-Action-Detection">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C2-Action-Detection">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/707#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/707#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/707#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/707#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/ad.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="challenge-action-anticipation">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Action Anticipation</h5>
<p class="text-muted">
<b>Task. </b>
Predict the (verb, noun) label of a future action observing a segment preceding its occurrence. <br/>
<b>Training input. </b> A set of trimmed action segments, each annotated with a (verb, noun) label. <br/>
<b>Testing input. </b> During testing you are allowed to observe a segment that <i>ends</i> at least one second before
the start of the action you are testing on.<br/>
<b>Splits. </b> Train and validation for training, evaluated on the test split. <br/>
<b>Evaluation metrics. </b> Top-5 recall averaged for all classes, as defined <a href="https://openaccess.thecvf.com/content_ECCVW_2018/papers/11133/Furnari_Leveraging_Uncertainty_to_Rethink_Loss_Functions_and_Evaluation_Measures_for_ECCVW_2018_paper.pdf" target="_blank">here</a>,
calculated for all segments as well as unseen participants and tail classes.
<br/>
<!--
<b>Caveat. </b> You are not allowed to observe any frame after one second from the start of a testing action.
-->
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C3-Action-Anticipation">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C3-Action-Anticipation">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/702#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/702#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/702#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/702#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/aa.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="challenge-domain-adaptation">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Unsupervised Domain Adaptation for Action Recognition</h5>
<p class="text-muted">
<b>Task. </b> Assign a (verb, noun) label to a trimmed segment, following the Unsupervised Domain Adaptation paradigm:
a labelled source domain is used for training, and the model needs to adapt to an unlabelled target domain. <br/>
<b>Training input. </b> A set of trimmed action segments, each annotated with a (verb, noun) label. <br/>
<b>Testing input. </b> A set of trimmed unlabelled action segments. <br/>
<b>Splits. </b> Videos recorded in 2018 (EPIC-KITCHENS-55) constitute the source domain,
while videos recorded for EPIC-KITCHENS-100's extension constitute the unlabelled target domain.
This challenge uses custom train/validation/test splits, which you can find
<a href="https://github.com/epic-kitchens/epic-kitchens-100-annotations#unsupervised-domain-adaptation-challenge" target="_blank">here</a>. <br/>
<b>Evaluation metrics. </b> Top-1/5 accuracy for verb, noun and action (verb+noun), on the target test set.
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C4-UDA-for-Action-Recognition">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C4-UDA-for-Action-Recognition">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/1241#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/1241#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/1241#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/1241#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/uda.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
<section class="challenge" id="challenge-action-retrieval">
<div class="row">
<div class="col-md-12">
<h5 class="subheading">Multi-Instance Retrieval</h5>
<p class="text-muted">
<b>Tasks. </b> <i>Video to text</i>: given a query video segment, rank captions such that those with a higher rank are
more semantically relevant to the action in the query video segment.
<i>Text to video:</i> given a query caption, rank video segments such that those with a higher rank are more semantically relevant
to the query caption. <br/>
<b>Training input. </b> A set of trimmed action segments, each annotated with a caption.
Captions correspond to the narration in English from which the action segment was obtained. <br/>
<b>Testing input. </b> A set of trimmed action segments with captions. Important: You are not allowed to use the known correspondence in the Test set <br/>
<b>Splits. </b> This challenge has its own custom splits, available <a href="https://github.com/epic-kitchens/epic-kitchens-100-annotations/tree/master/retrieval_annotations">here</a>. <br/>
<b>Evaluation metrics. </b> normalised Discounted Cumulative Gain (nDCG) and Mean Average Precision (mAP).
You can find more details in <a href="https://arxiv.org/pdf/2006.13256.pdf" target="_blank">our paper</a>.
</p>
</div>
</div>
<div class="details row text-center">
<div class="col-md">
<a href="https://github.com/epic-kitchens/C5-Multi-Instance-Retrieval">
<i class="text-center centered fas fa-4x fa-seedling"></i>
</a>
<br />
<p class="text-muted">
<a href="https://github.com/epic-kitchens/C5-Multi-Instance-Retrieval">Get started</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/617#learn_the_details-submission-format">
<i class="text-center centered fas fa-4x fa-info"></i>
</a>
<br />
<p class="text-muted">Learn about the
<a href="https://codalab.lisn.upsaclay.fr/competitions/617#learn_the_details-submission-format">submission format details</a>
</p>
</div>
<div class="col-md">
<a href="https://codalab.lisn.upsaclay.fr/competitions/617#results">
<i class="text-center centered fas fa-4x fa-trophy"></i>
</a>
<p class="text-muted">Submit your results on
<a href="https://codalab.lisn.upsaclay.fr/competitions/617#results">CodaLab website</a>
</p>
</div>
</div>
<figure class="text-muted">
<figcaption>Sample qualitative results from the challenge's baseline</figcaption>
<img src="{{ site.baseurl }}/static/img/challenges-epic-100/aret.png" style="width: 100%" class="img-responsive"/>
</figure>
</section>
</div>
</div>
</div>
</div>
</section>
<!-- Team -->
<section class="bg-light" id="team">
<div class="container">
<div class="row">
<div class="col-lg-12">
<h2 class="section-heading text-uppercase">The Team</h2>
<div class="text-muted">
<p> We are a group of researchers working in computer vision
from the <a href="http://www.bristol.ac.uk/">University
of Bristol</a> and <a href="http://www.unict.it/">University of
Catania</a>. The original dataset was collected in collaboration with <a href="https://www.cs.utoronto.ca/~fidler/">Sanja Fidler, University of Toronto</a>
</p>
</div>
<div class="row">
<div class="col-md-6">
<a href="http://www.bristol.ac.uk/">
<img src="{{ site.baseurl }}/static/img/universities/bristol-min.png" alt="" style="width:90%; margin-top:15px; margin-bottom:15px;">
</a>
</div>
<div class="col-md-6">
<a href="https://www.unict.it/en/">