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:20260817T171533Z
LOCATION:Hall B5 (2)\, B Block\, Level 5
DTSTART;TZID=Asia/Tokyo:20241203T172800
DTEND;TZID=Asia/Tokyo:20241203T174000
UID:siggraphasia_SIGGRAPH Asia 2024_sess110_papers_1007@linklings.com
SUMMARY:Neural Global Illumination via Superposed Deformable Feature Field
 s
DESCRIPTION:Chuankun Zheng, Yuchi Huo, Hongxiang Huang, and Hongtao Sheng 
 (State Key Laboratory of CAD&CG, Zhejiang University); Junrong Huang (City
  University of Hong Kong); Rui Tang and Hao Zhu (Manycore Inc.); and Rui W
 ang and Hujun Bao (State Key Laboratory of CAD&CG, Zhejiang University)\n\
 nInteractive rendering of dynamic scenes with complex global illumination 
 has been a long-standing problem in computer graphics.\nRecent advances in
  neural rendering demonstrate new promising possibilities.\nHowever, while
  existing methods have achieved impressive results, complex rendering effe
 cts (e.g., caustics) remain challenging.\nThis paper presents a novel neur
 al rendering method that is able to generate high-quality global illuminat
 ion effects, including but not limited to caustics, soft shadows, and indi
 rect highlights, for dynamic scenes with varying camera, lighting conditio
 ns, materials, and object transformations.\nInspired by object-oriented tr
 ansfer field representations, we employ deformable neural feature fields t
 o implicitly model the impacts of individual objects or light sources on g
 lobal illumination.\nBy employing neural feature fields, our method gains 
 the ability to represent high-frequency details, thus supporting complex r
 endering effects.\nWe superpose these feature fields in latent space and u
 tilize a lightweight decoder to obtain global illumination estimates, whic
 h allows our neural representations to spontaneously adapt to the contribu
 tion of individual objects or light sources to global illumination in a da
 ta-driven manner, thus further improving the quality.\nOur experiments dem
 onstrate the effectiveness of our method on a wide range of scenes with co
 mplex light paths, materials, and geometry.\n\nRegistration Category: Full
  Access, Full Access Supporter\n\nLanguage Format: English Language\n\nSes
 sion Chair: Michael Wimmer (TU Wien, Technische Universität Wien (TU Wien)
 )\n\n
URL:https://asia.siggraph.org/2024/program/?id=papers_1007&sess=sess110
END:VEVENT
END:VCALENDAR