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 (1)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241205T131100
DTEND;TZID=Asia/Tokyo:20241205T132300
UID:siggraphasia_SIGGRAPH Asia 2024_sess130_papers_1233@linklings.com
SUMMARY:NASM: Neural Anisotropic Surface Meshing
DESCRIPTION:Hongbo Li, Haikuan Zhu, and Sikai Zhong (Wayne State Universit
 y); Ningna Wang (University of Texas at Dallas); Cheng Lin (University of 
 Hong Kong); Xiaohu Guo (University of Texas at Dallas); Shiqing Xin (Shand
 ong University); Wenping Wang (Texas A&M University); and Jing Hua and Zic
 hun Zhong (Wayne State University)\n\nThis paper introduces a new learning
 -based method, NASM, for anisotropic surface meshing. Our key idea is to p
 ropose a graph neural network to embed an input mesh into a high-dimension
 al (high-d) Euclidean embedding space to preserve curvature-based anisotro
 pic metric by using a dot product loss between high-d edge vectors. This c
 an dramatically reduce the computational time and increase the scalability
 . Then, we propose a novel feature-sensitive remeshing on the generated hi
 gh-d embedding to automatically capture sharp geometric features. We defin
 e a high-d normal metric, and then derive an automatic differentiation on 
 a high-d centroidal Voronoi tessellation (CVT) optimization with the norma
 l metric to simultaneously preserve geometric features and curvature aniso
 tropy that exhibit in the original 3D shapes. To our knowledge, this is th
 e first time that a deep learning framework and a large dataset are propos
 ed to construct a high-d Euclidean embedding space for 3D anisotropic surf
 ace meshing. Experimental results are evaluated and compared with the stat
 e-of-the-art in anisotropic surface meshing on a large number of surface m
 odels from Thingi10K dataset as well as tested on extensive unseen 3D shap
 es from Multi-Garment Network dataset and FAUST human dataset.\n\nRegistra
 tion Category: Full Access, Full Access Supporter\n\nLanguage Format: Engl
 ish Language\n\nSession Chair: Noam Aigerman (University of Montreal, Mila
 )\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_1233&sess=sess130
END:VEVENT
END:VCALENDAR