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:20250110T023309Z
LOCATION:Hall B5 (1)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241203T151900
DTEND;TZID=Asia/Tokyo:20241203T153100
UID:siggraphasia_SIGGRAPH Asia 2024_sess106_papers_466@linklings.com
SUMMARY:DiffCSG: Differentiable CSG via Rasterization
DESCRIPTION:Technical Papers\n\nHaocheng Yuan (University of Edinburgh); A
 drien Bousseau (Inria Sophia-Antipolis, Université Côte d’Azur); Hao Pan (
 Microsoft Research Asia); Quancheng Zhang (Nanjing University); Niloy J. M
 itra (University College London (UCL), Adobe Research); and Changjian Li (
 University of Edinburgh)\n\nDifferentiable rendering is a key ingredient f
 or inverse rendering and machine learning, as it allows to optimize scene 
 parameters (shape, materials, lighting) to best fit target images. Differe
 ntiable rendering requires that each scene parameter relates to pixel valu
 es through differentiable operations. While 3D mesh rendering algorithms h
 ave been implemented in a differentiable way, these algorithms do not dire
 ctly extend to Constructive-Solid-Geometry (CSG), a popular parametric rep
 resentation of shapes, because the underlying boolean operations are typic
 ally performed with complex black-box mesh-processing libraries. We presen
 t an algorithm, DiffCSG, to render CSG models in a differentiable manner. 
 Our algorithm builds upon CSG rasterization, which displays the result of 
 boolean operations between primitives without explicitly computing the res
 ulting mesh and, as such, bypasses black-box mesh processing. We describe 
 how to implement CSG rasterization within a differentiable rendering pipel
 ine, taking special care to apply antialiasing along primitive intersectio
 ns to obtain gradients in such critical areas. Our algorithm is simple and
  fast, can be easily incorporated into modern machine learning setups, and
  enables a range of applications for computer-aided design, including dire
 ct and image-based editing of CSG primitives. Code and data: https://yyyyy
 hc.github.io/DiffCSG/.\n\nRegistration Category: Full Access, Full Access 
 Supporter\n\nLanguage Format: English Language\n\nSession Chair: Yonghao Y
 ue (Aoyama Gakuin University)
URL:https://asia.siggraph.org/2024/program/?id=papers_466&sess=sess106
END:VEVENT
END:VCALENDAR