BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260805T205503Z
LOCATION:Room 408 A
DTSTART;TZID=America/Los_Angeles:20260722T140000
DTEND;TZID=America/Los_Angeles:20260722T153000
UID:siggraph_SIGGRAPH 2026_sess139@linklings.com
SUMMARY:Shading, Relighting and Volumes
DESCRIPTION:Relit-LiVE: Relight Video by Jointly Learning Environment Vide
 o\n\nRelit-LiVE produces physically consistent and temporally stable relig
 hting results without requiring prior knowledge of the camera pose. It uti
 lizes raw reference images to ensure realistic lighting effects and introd
 uces a joint prediction framework generating relighting videos with frame-
 level war...\n\n\nWeiqing Xiao (Nanjing University, BAAI); Hong Li (Beihan
 g University, BAAI); Xiuyu Yang (Tsinghua University); Houyuan Chen (HKUST
 , BAAI); Wenyi Li (University of Chinese Academy of Sciences China); Tianq
 i Liu (Huazhong University of Science and Technology); Shaocong Xu (BAAI);
  Chongjie Ye (The Chinese University of Hongkong); Hao Zhao (Tsinghua Univ
 ersity, BAAI); and Beibei Wang (Nanjing University)\n---------------------
 \nGR3EN: Generative Relighting for 3D Environments\n\nGR3EN enables religh
 ting 3D environments to user-specified lighting conditions. It provides fu
 ll control over every light source in a 3D scene — enabling adding new lig
 ht sources, changing the color of existing ones or turning them off entire
 ly. Shadows, reflections, and interreflections all u...\n\n\nXiaoyan Xing 
 (University of Amsterdam); Philipp Henzler (Google Research); and Junhwa H
 ur, Runze Li, Jonathan T. Barron, Pratul P. Srinivasan, and Dor Verbin (De
 epMind)\n---------------------\nTaming optimization variance in compact ne
 ural shading networks\n\nWe present a training strategy for small networks
  that improves stability, reducing run-to-run variance by 88% and average 
 loss by 38%. Multiple instances train in parallel on different batch slice
 s, while weaker ones are periodically removed until only the best-performi
 ng instance remains, improvin...\n\n\nBenedikt Bitterli, Petrik Clarberg, 
 Chris Cummings, and Aaron Lefohn (NVIDIA); Steve Marschner (Cornell Univer
 sity, NVIDIA); and Jan Novák, Fabrice Rousselle, Andrea Weidlich, and Tizi
 an Zeltner (NVIDIA)\n---------------------\nMulti-feature Radiance Baking 
 Neural Networks for Instant Volumetric Rendering\n\nWe propose MRBNN that 
 delivers real-time photorealistic volume rendering. We leverage reformulat
 ion of in-scattering integral and design an efficient neural representatio
 n and sampling strategy. Our method achieves rendering in a few millisecon
 ds with the ability to handle complex cases such as spat...\n\n\nJiaming L
 iang (Peking University); Hongliang Yuan (XiaoMi company); and Meng Gai, G
 uoping Wang, and Sheng Li (Peking University)\n---------------------\nBoun
 ding Stratified Bernoulli Impulses for Ray Marching Gaussian Process Impli
 cit Surfaces\n\nIn this paper, we present an optimized algorithm for rende
 ring Gaussian Process Implicit Surfaces (GPISes). By using stratified Bern
 oulli impulses, we precompute point-level bounds stored in a bound map. We
  also represent region-level bounds with a spatial data structure. Combini
 ng these bounds, we ...\n\n\nJunjie Chen and Zhimin Fan (Nanjing Universit
 y); Ling-Qi Yan (Mohamed bin Zayed University of Artificial Intelligence);
  Junqiu Zhu (Shandong University); Yanwen Guo (Nanjing University); Kun Zh
 ou (State Key Lab of CAD&CG, Zhejiang University; Institute of Hangzhou Ho
 lographic Intelligent Technology); and Jie Guo (Nanjing University)\n-----
 ----------------\nAdaptive Ray Marching for Rendering Gaussian Process Imp
 licit Surfaces\n\nThis paper accelerates Gaussian Process Implicit Surface
  (GPIS) rendering by introducing an online algorithm for incrementally sam
 pling conditional Gaussian distributions along rays and an adaptive ray ma
 rching scheme that minimizes steps. Together, these techniques reduce rend
 ering error by up to 4...\n\n\nZhiqian Zhou (University of California Irvi
 ne), Dario Seyb (Valve), and Shuang Zhao (University of Illinois Urbana-Ch
 ampaign)\n---------------------\nShading, Relighting and Volumes - Interac
 tive Discussion\n\nAfter the summary presentations, attendees will partici
 pate in an interactive discussion. Outside the room will be a series of po
 ster boards for authors to gather around with the audience. Authors are in
 vited to bring any material related to their paper that could instigate fu
 rther conversation such...\n\n\nInterest Area: Research & Education\n\nKey
 word: Artificial Intelligence/Machine Learning, Lighting, Modeling, Real-T
 ime, Rendering\n\nRegistration Category: Full Conference Supporter, Full C
 onference\n\nSession Chair: Yang Zhou (Meta, University of California Sant
 a Barbara)
END:VEVENT
END:VCALENDAR