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:20250110T023312Z
LOCATION:Hall B5 (1)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241205T093400
DTEND;TZID=Asia/Tokyo:20241205T094600
UID:siggraphasia_SIGGRAPH Asia 2024_sess124_papers_852@linklings.com
SUMMARY:Deformation Recovery: Localized Learning for Detail-Preserving Def
 ormations
DESCRIPTION:Technical Papers\n\nRamana Sundararaman (Centre National de la
  Recherche Scientifique - Laboratoire d'informatique de l'École Polytechni
 que (LIX)); Nicolas Donati (Ansys); Simone Melzi (University of Milano-Bic
 occa); Etienne Corman (Université de Lorraine, CNRS); and Maks Ovsjanikov 
 (Centre National de la Recherche Scientifique - Laboratoire d'informatique
  de l'École Polytechnique (LIX))\n\nWe introduce a novel data-driven appro
 ach aimed at designing high-quality shape deformations based on a coarse l
 ocalized input signal. Unlike previous data-driven methods that require a 
 global shape encoding, we observe that detail-preserving deformations can 
 be estimated reliably without any global context in certain scenarios. Bui
 lding on this intuition, we leverage Jacobians defined in a one-ring neigh
 borhood as a coarse representation of the deformation. Using this as the i
 nput to our neural network, we apply a series of MLPs combined with featur
 e smoothing to learn the Jacobian corresponding to the detail-preserving d
 eformation, from which the embedding is recovered by the standard Poisson 
 solve. Crucially, by removing the dependence on a global encoding, every p
 oint becomes a training example, making the supervision particularly light
 weight. Moreover, when trained on a class of shapes, our approach demonstr
 ates remarkable generalization across different object categories. Equippe
 d with this novel network, we explore three main tasks: refining an approx
 imate shape correspondence, unsupervised deformation and mapping, and shap
 e editing.\n\nRegistration Category: Full Access, Full Access Supporter\n\
 nLanguage Format: English Language\n\nSession Chair: Yotam Gingold (George
  Mason University)
URL:https://asia.siggraph.org/2024/program/?id=papers_852&sess=sess124
END:VEVENT
END:VCALENDAR