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:20240214T070241Z
LOCATION:Darling Harbour Theatre\, Level 2 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231212T093000
DTEND;TZID=Australia/Melbourne:20231212T124500
UID:siggraphasia_SIGGRAPH Asia 2023_sess209_papers_629@linklings.com
SUMMARY:From Skin to Skeleton : Towards Biomechanically Accurate 3D Digita
 l Humans
DESCRIPTION:Technical Papers\n\nMarilyn Keller (Max Planck Institute for I
 ntelligent Systems), Keenon Werling (Stanford University), Soyong Shin (Ma
 x-Planck-Institut für Informatik), Scott Delp (Stanford), Sergi Pujades (I
 NRIA), Karen Liu (Stanford University), and Michael Black (Max Planck Inst
 itute for Intelligent Systems)\n\nGreat progress has been made in estimati
 ng 3D human pose and shape from images and video by training neural networ
 ks to directly regress the parameters of parametric human models like SMPL
 .\nHowever, existing body models have simplified kinematic structures that
  do not correspond to accurate joint locations and articulations in the hu
 man skeletal system, limiting their potential use in biomechanics. On the 
 other hand, methods for estimating biomechanically accurate skeletal motio
 n typically rely on complex motion capture systems and expensive optimizat
 ion methods.\nWhat is needed is a parametric 3D human model with a biomech
 anically accurate skeletal structure that can be easily regressed from ima
 ges.\nTo that end, we develop SKEL,  which re-rigs the SMPL body model wit
 h a biomechanics skeleton. To enable this, we need training data of skelet
 ons inside SMPL meshes in diverse poses. We build such a dataset by optimi
 zing biomechanically accurate skeletons inside SMPL meshes from AMASS sequ
 ences. We then learn a regressor from SMPL mesh vertices to the true joint
  locations and bone rotations. Finally, we re-parametrize the SMPL mesh wi
 th the new kinematic parameters.\nThe resulting SKEL model is animatable l
 ike SMPL but with fewer, and biomechanically-realistic, degrees of freedom
 . We also train a regressor from SMPL meshes to the skeleton enabling us t
 o ``upgrade" existing datasets that are in SMPL format. We show that SKEL 
 has more biomechanically accurate joint locations than SMPL, and the bones
  fit inside the body surface better than previous methods.\nSKEL provides 
 a new tool to enable biomechanics in the wild, while also providing vision
  and graphics researchers with a better constrained and more realistic mod
 el of human articulation.\n\nRegistration Category: Full Access, Enhanced 
 Access, Trade Exhibitor, Experience Hall Exhibitor
URL:https://asia.siggraph.org/2023/full-program?id=papers_629&sess=sess209
END:VEVENT
END:VCALENDAR