BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Seoul
X-LIC-LOCATION:Asia/Seoul
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:KST
DTSTART:18871231T000000
DTSTART:19881009T020000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20230103T035307Z
LOCATION:Room 324\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221206T153000
DTEND;TZID=Asia/Seoul:20221206T170000
UID:siggraphasia_SIGGRAPH Asia 2022_sess155_papers_420@linklings.com
SUMMARY:Efficient Differentiation of Pixel Reconstruction Filters for Path
 -Space Differentiable Rendering
DESCRIPTION:Technical Communications, Technical Papers\n\nEfficient Differ
 entiation of Pixel Reconstruction Filters for Path-Space Differentiable Re
 ndering\n\nYu, Zhang, Dong, Nowrouzezahrai, Zhao\n\nPixel reconstruction f
 ilters play an important role in physics-based rendering and have been tho
 roughly studied. In physics-based differentiable rendering, however, the p
 roper treatment of pixel reconstruction filters has remained largely under
 -explored. We present a new technique to efficiently differentiate pixel r
 econstruction filters based on the path-space formulation. Specifically, w
 e formulate the pixel boundary integral that models discontinuities in pix
 el filters and introduce new antithetic sampling methods that support diff
 erentiable path sampling methods, such as adjoint particle tracing and bid
 irectional path tracing. We demonstrate both the need and efficacy of anti
 thetic sampling when treating this integral, and we evaluate the effective
 ness of our method across several differentiable- and inverse-rendering se
 ttings.\n\nRegistration Category: FULL ACCESS, ON-DEMAND ACCESS\n\nLanguag
 e: ENGLISH\n\nFormat: IN-PERSON, ON-DEMAND
URL:https://sa2022.siggraph.org/en/full-program/?id=papers_420&sess=sess15
 5
END:VEVENT
END:VCALENDAR