BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260817T171536Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241206T092300
DTEND;TZID=Asia/Tokyo:20241206T093400
UID:siggraphasia_SIGGRAPH Asia 2024_sess140_papers_281@linklings.com
SUMMARY:Neural Differential Appearance Equations
DESCRIPTION:Chen Liu and Tobias Ritschel (University College London (UCL))
 \n\nWe propose a method to reproduce dynamic appearance textures with spac
 e-stationary but time-varying visual statistics.\nWhile most previous work
  decomposes dynamic textures into static appearance and motion, we focus o
 n dynamic appearance that results not from motion but variations of fundam
 ental properties, such as rusting, decaying, melting, and weathering.\nTo 
 this end, we adopt the neural ordinary differential equation (ODE) to lear
 n the underlying dynamics of appearance from a target exemplar.\nWe simula
 te the ODE in two phases.\nAt the ``warm-up'' phase, the ODE diffuses a ra
 ndom noise to an initial state.\nWe then constrain the further evolution o
 f this ODE to replicate the evolution of visual feature statistics in the 
 exemplar during the generation phase.\nThe particular innovation of this w
 ork is the neural ODE achieving both denoising and evolution for dynamics 
 synthesis, with a proposed temporal training scheme.\nWe study both religh
 table (BRDF) and non-relightable (RGB) appearance models.\nFor both we int
 roduce new pilot datasets, allowing, for the first time, to study such phe
 nomena:\nFor RGB we provide 22 dynamic textures acquired from free online 
 sources;\nFor BRDF, we further acquire a dataset of 21 flash-lit videos of
  time-varying materials, enabled by a simple-to-construct setup.\nOur expe
 riments show that our method consistently yields realistic and coherent re
 sults, whereas prior works falter under pronounced temporal appearance var
 iations.\nA user study confirms our approach is preferred to previous work
  for such exemplars.\n\nRegistration Category: Full Access, Full Access Su
 pporter\n\nLanguage Format: English Language\n\nSession Chair: Seungyong L
 ee (POSTECH)\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_281&sess=sess140
END:VEVENT
END:VCALENDAR