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:20240214T070247Z
LOCATION:Meeting Room C4.8\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231214T105000
DTEND;TZID=Australia/Melbourne:20231214T110500
UID:siggraphasia_SIGGRAPH Asia 2023_sess149_papers_210@linklings.com
SUMMARY:Learning the Geodesic Embedding with Graph Neural Networks
DESCRIPTION:Technical Papers\n\nBo Pang (Peking Unversity); Zhongtian Zhen
 g (Peking University); Guoping Wang (Peking Unversity); and Peng-Shuai Wan
 g (Peking University, Wangxuan Institute of Computer Technology)\n\nWe pre
 sent GeGnn, a learning-based method for computing the approximate geodesic
  distance between two arbitrary points on discrete polyhedra surfaces with
  constant time complexity after fast precomputation. Previous relevant met
 hods either focus on computing the geodesic distance between a single sour
 ce and all destinations, which has linear complexity at least, or require 
 long precomputation time. Our key idea is to train a graph neural network 
 to embed an input mesh into a high-dimensional embedding space and compute
  the geodesic distance between a pair of points using the corresponding em
 bedding vectors and a lightweight decoding function. To facilitate the lea
 rning of the embedding, we propose novel graph convolution and graph pooli
 ng modules that incorporate local geodesic information and are verified to
  be much more effective than previous designs. After training, our method 
 requires only one forward pass of the network per mesh as precomputation. 
 Then, we can compute the geodesic distance between a pair of points using 
 our decoding function, which requires only several matrix multiplications 
 and can be massively parallelized on GPUs. We verify the efficiency and ef
 fectiveness of our method on ShapeNet and demonstrate that our method is f
 aster than existing methods by orders of magnitude while achieving compara
 ble or better accuracy. Additionally, our method exhibits robustness on no
 isy and incomplete meshes and strong generalization ability on out-of-dist
 ribution meshes.\n\nRegistration Category: Full Access\n\nSession Chair: M
 arco ATTENE (Institute for Applied Mathematics and Information Technologie
 s (IMATI), CNR)
URL:https://asia.siggraph.org/2023/full-program?id=papers_210&sess=sess149
END:VEVENT
END:VCALENDAR