BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//jEvents 2.0 for Joomla//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:STANDARD
DTSTART:20231105T010000
RDATE:20240310T030000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:America/New_York EST
END:STANDARD
BEGIN:STANDARD
DTSTART:20241103T010000
RDATE:20250309T030000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:America/New_York EST
END:STANDARD
BEGIN:STANDARD
DTSTART:20251102T010000
RDATE:20260308T030000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:America/New_York EST
END:STANDARD
BEGIN:STANDARD
DTSTART:20261101T010000
RDATE:20270314T030000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:America/New_York EST
END:STANDARD
BEGIN:STANDARD
DTSTART:20271107T010000
RDATE:20280312T030000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:America/New_York EST
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20230501T150000
RDATE:20231105T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:America/New_York EDT
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20240310T030000
RDATE:20241103T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:America/New_York EDT
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20250309T030000
RDATE:20251102T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:America/New_York EDT
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20260308T030000
RDATE:20261101T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:America/New_York EDT
END:DAYLIGHT
BEGIN:DAYLIGHT
DTSTART:20270314T030000
RDATE:20271107T010000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:America/New_York EDT
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:3ddf46346dcc46b5dbfeb2313c411705
CATEGORIES:PhD Defense
CREATED:20240415T111719
SUMMARY:Maximize Utilization of Support-Set for Few-shot segmentation
DESCRIPTION:Speaker: Seonghyeon Moon\nLocation: CBIM 22 Multipurpose Room\nCommittee Co
 mposition:\n\n - Professor Mubbasir Kapadia from Rutgers University\n - Pro
 fessor Vladimir Pavlovic from Rutgers University\n - Professor Mridul Aanja
 neya from Rutgers University\n - Professor Xia Haifeng from Southeast Unive
 rsity (external)Abstract:\nIn Few-shot segmentation(FSS), the support set p
 lays a critical role as it provides target information to segment the targe
 t object in a given image. However, previous works focused on improving net
 work architecture to get performance improvements neglecting the importance
  of how to utilize target features from the support set. We observed there 
 were performance bottlenecks because of the limited utilization of the supp
 ort set. We investigated information loss and proposed new approaches to re
 trieve this loss to get more meaningful information from the support set. W
 e validated the effectiveness of our approach by instantiating it into thre
 e recent and strong FSS methods. Experimental results on several publicly a
 vailable FSS benchmarks show that our proposed methods consistently improve
  performance by visible margins and lead to faster convergence.\nZoom:\nhtt
 ps://rutgers.zoom.us/j/6756059091?pwd=M0RoNjY4YWN1OWptNys5WlRqMFhmdz09&
 omn=91204206398\nJoin by SIP\nRegularLabs.EmailProtector.unCloak("ep_3dd92c
 52"); (javascript:/* This email address is being protected from spambots.*/
 )RegularLabs.EmailProtector.unCloak("ep_9a141f5f", true);\nMeeting ID: 675 
 605 9091\nPassword: 077449\n
X-ALT-DESC;FMTTYPE=text/html:<p><strong>Speaker:</strong>&nbsp;Seonghyeon Moon</p><p><strong>Location:</
 strong>&nbsp;CBIM 22 Multipurpose Room</p><p><strong>Committee Composition:
 </strong></p><ul><li>Professor Mubbasir Kapadia from Rutgers University</li
 ><li>Professor Vladimir Pavlovic from Rutgers University</li><li>Professor 
 Mridul Aanjaneya from Rutgers University</li><li>Professor Xia Haifeng from
  Southeast University (external)</li></ul><p><strong>Abstract:</strong></p>
 <p>In Few-shot segmentation(FSS), the support set plays a critical role as 
 it provides target information to segment the target object in a given imag
 e. However, previous works focused on improving network architecture to get
  performance improvements neglecting the importance of how to utilize targe
 t features from the support set. We observed there were performance bottlen
 ecks because of the limited utilization of the support set. We investigated
  information loss and proposed new approaches to retrieve this loss to get 
 more meaningful information from the support set. We validated the effectiv
 eness of our approach by instantiating it into three recent and strong FSS 
 methods. Experimental results on several publicly available FSS benchmarks 
 show that our proposed methods consistently improve performance by visible 
 margins and lead to faster convergence.</p><p><strong>Zoom:</strong></p><p>
 <a href="https://rutgers.zoom.us/j/6756059091?pwd=M0RoNjY4YWN1OWptNys5WlRqM
 Fhmdz09&amp;omn=91204206398" target="_blank" rel="noopener">https://rutgers
 .zoom.us/j/6756059091?pwd=M0RoNjY4YWN1OWptNys5WlRqMFhmdz09&amp;omn=91204206
 398</a><br />Join by SIP<br /><!-- This email address is being protected fr
 om spambots. --><a href="https://www.cs.rutgers.edu/javascript:/* This emai
 l address is being protected from spambots.*/" target="_blank"><span class=
 "cloaked_email ep_3dd92c52"><span data-ep-a="&#54;&#55;&#53;6" data-ep-b="&
 #111;m"><span data-ep-a="0&#53;9&#48;" data-ep-b="&#114;&#99;&#46;&#99;"><s
 pan data-ep-a="&#57;&#49;&#64;z" data-ep-b="&#111;o&#109;&#99;"></span></sp
 an></span></span><script>RegularLabs.EmailProtector.unCloak("ep_3dd92c52");
 </script><span class="cloaked_email ep_9a141f5f" style="display:none;"><spa
 n data-ep-b="om" data-ep-a="67&#53;&#54;"><span data-ep-a="05&#57;&#48;" da
 ta-ep-b="&#114;&#99;.&#99;"><span data-ep-a="&#57;&#49;&#64;&#122;" data-ep
 -b="o&#111;&#109;c"></span></span></span></span></a><script>RegularLabs.Ema
 ilProtector.unCloak("ep_9a141f5f", true);</script><br />Meeting ID: 675 605
  9091<br />Password: 077449</p>
DTSTAMP:20260827T040712
DTSTART;TZID=America/New_York:20240501T150000
SEQUENCE:0
TRANSP:OPAQUE
END:VEVENT
END:VCALENDAR