BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240214T070242Z
LOCATION:Meeting Room C4.8\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T162500
DTEND;TZID=Australia/Melbourne:20231212T164000
UID:siggraphasia_SIGGRAPH Asia 2023_sess160_tog_103@linklings.com
SUMMARY:SparsePoser: Real-time Full-body Motion Reconstruction from Sparse
  Data
DESCRIPTION:Technical Papers, TOG\n\nJose Luis Ponton and Haoran Yun (Univ
 ersitat Politècnica de Catalunya (UPC)); Andreas Aristidou (University of 
 Cyprus, CYENS Centre of Excellence); and Carlos Andujar and Nuria Pelechan
 o (Universitat Politècnica de Catalunya (UPC))\n\nAccurate and reliable hu
 man motion reconstruction is crucial for creating natural interactions of 
 full-body avatars in Virtual Reality (VR) and entertainment applications. 
 As the Metaverse and social applications gain popularity, users are seekin
 g cost-effective solutions to create full-body animations that are compara
 ble in quality to those produced by commercial motion capture systems. In 
 order to provide affordable solutions though, it is important to minimize 
 the number of sensors attached to the subject's body. Unfortunately, recon
 structing the full-body pose from sparse data is a heavily under-determine
 d problem. Some studies that use IMU sensors face challenges in reconstruc
 ting the pose due to positional drift and ambiguity of the poses. In recen
 t years, some mainstream VR systems have released 6-degree-of-freedom (6-D
 oF) tracking devices providing positional and rotational information. Neve
 rtheless, most solutions for reconstructing full-body poses rely on tradit
 ional inverse kinematics (IK) solutions, which often produce non-continuou
 s and unnatural poses. In this paper, we introduce SparsePoser, a novel de
 ep learning-based solution for reconstructing a full-body pose from a redu
 ced set of six tracking devices. Our system incorporates a convolutional-b
 ased autoencoder that synthesizes high-quality continuous human poses by l
 earning the human motion manifold from motion capture data. Then, we emplo
 y a learned IK component, made of multiple lightweight feed-forward neural
  networks, to adjust the hands and feet towards the corresponding trackers
 . We extensively evaluate our method on publicly available motion capture 
 datasets and with real-time live demos. We show that our method outperform
 s state-of-the-art techniques using IMU sensors or 6-DoF tracking devices,
  and can be used for users with different body dimensions and proportions.
 \n\nRegistration Category: Full Access\n\nSession Chair: Ioannis Karamouza
 s (Clemson University, University of California Riverside)
URL:https://asia.siggraph.org/2023/full-program?id=tog_103&sess=sess160
END:VEVENT
END:VCALENDAR