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:20240214T070248Z
LOCATION:Meeting Room C4.8\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231214T143500
DTEND;TZID=Australia/Melbourne:20231214T144500
UID:siggraphasia_SIGGRAPH Asia 2023_sess151_papers_637@linklings.com
SUMMARY:Neural Stress Fields for Reduced-order Elastoplasticity and Fractu
 re
DESCRIPTION:Technical Papers\n\nZeshun Zong and Xuan Li (University of Cal
 ifornia Los Angeles); Minchen Li (University of California Los Angeles, Ca
 rnegie Mellon University); Maurizio M. Chiaramonte (Meta Reality Labs Rese
 arch); Wojciech Matusik (MIT CSAIL); Eitan Grinspun (University of Toronto
 ); Kevin Carlberg (Meta Reality Labs Research); Chenfanfu Jiang (Universit
 y of California Los Angeles); and Peter Yichen Chen (MIT CSAIL)\n\nThe mat
 erial point method (MPM) is a versatile simulation framework for large-def
 ormation elastoplasticity and fracture. However, MPM's long runtime and la
 rge memory consumptions render it unsuitable for applications constrained 
 by computation time and memory usage, e.g., virtual reality. To overcome t
 hese barriers, we propose a reduced-order MPM framework. Our key innovatio
 n is training a low-dimensional manifold for the Kirchhoff stress field vi
 a an implicit neural representation. This low-dimensional neural stress fi
 eld (NSF) enables efficient evaluations of stress values and, correspondin
 gly, internal forces at arbitrary spatial locations. In addition, we also 
 train neural deformation and affine fields to build low-dimensional manifo
 lds for the deformation and affine momentum fields. These neural stress, d
 eformation, and affine fields share the same low-dimensional latent space,
  which uniquely embeds the high-dimensional MPM simulation state. After tr
 aining, we run new simulations by evolving in this single latent space, wh
 ich drastically reduces the computation time and memory consumption. Our g
 eneral continuum-mechanics-based reduced-order framework is applicable to 
 any phenomena governed by the elastodynamics equation. To showcase the ver
 satility of our framework, we simulate a wide range of material behaviors,
  including elastica, sand, metal, non-Newtonian fluids, fracture, contact,
  and collision. We demonstrate dimension reduction by up to 100,000X and t
 ime savings by up to 10X.\n\nRegistration Category: Full Access\n\nSession
  Chair: Tao Du (Tsinghua University, Shanghai Qi Zhi Institute)
URL:https://asia.siggraph.org/2023/full-program?id=papers_637&sess=sess151
END:VEVENT
END:VCALENDAR