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:20241204T105600
DTEND;TZID=Asia/Tokyo:20241204T110800
UID:siggraphasia_SIGGRAPH Asia 2024_sess112_papers_237@linklings.com
SUMMARY:Fast and Globally Consistent Normal Orientation based on the Windi
 ng Number Normal Consistency
DESCRIPTION:Technical Papers\n\nSiyou Lin (Department of Automation, Tsing
 hua University); Zuoqiang Shi (Yau Mathematical Sciences Center, Tsinghua 
 University; Yanqi Lake Beijing Institute of Mathematical Sciences and Appl
 ications); and Yebin Liu (Department of Automation, Tsinghua University)\n
 \nEstimating consistently oriented normals for point clouds enables a numb
 er of important applications in computer graphics such as surface reconstr
 uction. While local normal estimation is possible with simple techniques l
 ike principal component analysis (PCA), orienting these normals to be glob
 ally consistent has been a notoriously difficult problem.\nSome recent met
 hods exploit various properties of the winding number formula to achieve g
 lobal consistency with state-of-the-art performance.\nDespite their exciti
 ng progress, these algorithms either have high space/time complexity, or d
 o not produce accurate and consistently oriented normals for imperfect dat
 a.\nIn this paper, we propose a novel property from the winding number for
 mula, Winding Number Normal Consistency (WNNC), to tackle this problem. Th
 e derived property is based on the simple observation that the normals (ne
 gative gradients) sampled from the winding number field should be codirect
 ional to the normals used to compute the winding number field. Since the W
 NNC property itself does not resolve the inside/outside orientation ambigu
 ity, we further propose to incorporate an objective function from Parametr
 ic Gauss Reconstruction (PGR). We propose to iteratively update normals by
  alternating between WNNC-based normal updates and PGR-based gradient desc
 ents, which leads to an embarrassingly simple yet effective iterative algo
 rithm that allows fast and high-quality convergence to a globally consiste
 nt normal vector field.\nFurthermore, our proposed algorithm only involves
  repeatedly evaluating the winding number formula and its derivatives, whi
 ch can be accelerated and parallelized using a treecode-based approximatio
 n algorithm due to their special structures. Exploiting this fact, we impl
 ement a GPU-accelerated treecode-based solver. Our GPU (and even CPU) impl
 ementation can be significantly faster than the recent state-of-the-art me
 thods for normal orientation from raw points. Our code is integrated with 
 the popular PyTorch framework to facilitate further research into winding 
 numbers, and is publicly available at https://jsnln.github.io/wnnc/index.h
 tml.\n\nRegistration Category: Full Access, Full Access Supporter\n\nLangu
 age Format: English Language\n\nSession Chair: Michael Wimmer (TU Wien)
URL:https://asia.siggraph.org/2024/program/?id=papers_237&sess=sess112
END:VEVENT
END:VCALENDAR