BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Seoul
X-LIC-LOCATION:Asia/Seoul
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:KST
DTSTART:18871231T000000
DTSTART:19881009T020000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230103T035312Z
LOCATION:Room 325-AB\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221209T140000
DTEND;TZID=Asia/Seoul:20221209T153000
UID:siggraphasia_SIGGRAPH Asia 2022_sess178_papers_170@linklings.com
SUMMARY:UmeTrack: Unified multi-view end-to-end hand tracking for VR
DESCRIPTION:Technical Communications, Technical Papers\n\nUmeTrack: Unifie
 d multi-view end-to-end hand tracking for VR\n\nHan, Wu, Zhang, Liu, Zhang
 ...\n\nReal-time tracking of 3D hand pose in world space is a challenging 
 problem\nand plays an important role in VR interaction. Existing work in t
 his space are\nlimited to either producing root-relative (versus world spa
 ce) 3D pose or rely\non multiple stages such as generating heatmaps and ki
 nematic optimization\nto obtain 3D pose. Moreover, the typical VR scenario
 , which involves multiview\ntracking from wide field of view (FOV) cameras
  is seldom addressed by\nthese methods. In this paper, we present a unifie
 d end-to-end differentiable\nframework for multi-view, multi-frame hand tr
 acking that directly predicts\n3D hand pose in world space. We demonstrate
  the benefits of end-to-end\ndifferentiabilty by extending our framework w
 ith downstream tasks such\nas jitter reduction and pinch prediction. To de
 monstrate the efficacy of our\nmodel,we further present a newlarge-scale e
 gocentric hand pose dataset that\nconsists of both real and synthetic data
 . Experiments show that our system\nhandles various challenging interactiv
 e motions, and has been successfully\napplied to real-time VR applications
 .\n\nRegistration Category: FULL ACCESS, ON-DEMAND ACCESS\n\nLanguage: ENG
 LISH\n\nFormat: IN-PERSON, ON-DEMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_170&sess=sess17
 8
END:VEVENT
END:VCALENDAR