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:20230103T035306Z LOCATION:Auditorium\, Level 5\, West Wing DTSTART;TZID=Asia/Seoul:20221206T100000 DTEND;TZID=Asia/Seoul:20221206T120000 UID:siggraphasia_SIGGRAPH Asia 2022_sess153_papers_420@linklings.com SUMMARY:Efficient Differentiation of Pixel Reconstruction Filters for Path -Space Differentiable Rendering DESCRIPTION:Technical Papers\n\nEfficient Differentiation of Pixel Reconst ruction Filters for Path-Space Differentiable Rendering\n\nYu, Zhang, Dong , Nowrouzezahrai, Zhao\n\nPixel reconstruction filters play an important r ole in physics-based rendering and have been thoroughly studied. In physic s-based differentiable rendering, however, the proper treatment of pixel r econstruction filters has remained largely under-explored. We present a ne w technique to efficiently differentiate pixel reconstruction filters base d on the path-space formulation. Specifically, we formulate the pixel boun dary integral that models discontinuities in pixel filters and introduce n ew antithetic sampling methods that support differentiable path sampling m ethods, such as adjoint particle tracing and bidirectional path tracing. W e demonstrate both the need and efficacy of antithetic sampling when treat ing this integral, and we evaluate the effectiveness of our method across several differentiable- and inverse-rendering settings.\n\nRegistration Ca tegory: FULL ACCESS, EXPERIENCE PLUS ACCESS, EXPERIENCE ACCESS, TRADE EXHI BITOR\n\nLanguage: ENGLISH\n\nFormat: IN-PERSON URL:https://sa2022.siggraph.org/en/full-program/?id=papers_420&sess=sess15 3 END:VEVENT END:VCALENDAR