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:20260114T163709Z
LOCATION:Meeting Room C4.9+C4.10\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231213T153000
DTEND;TZID=Australia/Melbourne:20231213T154500
UID:siggraphasia_SIGGRAPH Asia 2023_sess168_papers_711@linklings.com
SUMMARY:Projective Sampling for Differentiable Rendering of Geometry
DESCRIPTION:Ziyi Zhang, Nicolas Roussel, and Wenzel Jakob (Ecole Polytechn
 ique Fédérale de Lausanne)\n\nDiscontinuous visibility changes at object b
 oundaries remain a persistent source of difficulty in the area of differen
 tiable rendering. Left untreated, they bias computed gradients so severely
  that even basic optimization tasks fail.\n\nPrior path-space methods addr
 essed this bias by decoupling boundaries from the interior, allowing each 
 part to be handled using specialized Monte Carlo sampling strategies. Whil
 e conceptually powerful, the full potential of this idea remains unrealize
 d since existing methods often fail to adequately sample the boundary prop
 ortional to its contribution.\n\nThis paper presents theoretical and algor
 ithmic contributions. On the theoretical side, we transform the boundary d
 erivative into a remarkably simple local integral that invites present and
  future developments.\n\nBuilding on this result, we propose a new strateg
 y that projects ordinary samples produced during forward rendering onto ne
 arby boundaries. The resulting projections establish a variance-reducing g
 uiding distribution that accelerates convergence of the subsequent differe
 ntial phase.\n\nWe demonstrate the superior efficiency and versatility of 
 our method across a variety of shape representations, including triangle m
 eshes, implicitly defined surfaces, and cylindrical fibers based on Bézier
  curves.\n\nRegistration Category: Full Access\n\nSession Chair: Soo-Mi Ch
 oi (Sejong University)\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_711&sess=sess168
END:VEVENT
END:VCALENDAR
