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:20230103T035348Z
LOCATION:Room 324\, Level 3\, West Wing
DTSTART;TZID=Asia/Seoul:20221209T103000
DTEND;TZID=Asia/Seoul:20221209T120000
UID:siggraphasia_SIGGRAPH Asia 2022_sess173@linklings.com
SUMMARY:Sampling and Reconstruction
DESCRIPTION:Technical Communications, Technical Papers\n\nThe presentation
 s will be followed by a 30-min Interactive Discussion Session at Room 325-
 CD.\n\nTechnical Papers are published as a special issue of ACM Transactio
 ns on Graphics. In addition to papers selected by the SIGGRAPH Asia 2022 T
 echnical Papers Jury, the conference presents papers that have been publis
 hed in ACM Transactions on Graphics during the past year. Accepted papers 
 adhere to the highest scientific standards.\n\nThe Technical Communication
 s program is a premier forum for presenting the latest developments and re
 search still in progress. Leading international experts in academia and in
 dustry present work that showcase actual implementations of research ideas
 , works at the crossroads of computer graphics with computer vision, machi
 ne learning, HCI, VR, CAD, visualization, and many others.\n\nMarginal Mul
 tiple Importance Sampling\n\nWest, Georgiev, Hachisuka\n\nMultiple importa
 nce sampling (MIS) is a powerful tool to combine different sampling techni
 ques in a provably good manner. MIS requires that the techniques' probabil
 ity density functions (PDFs) are readily evaluable point-wise. However, th
 is requirement may not be satisfied when (some of) those PDFs ...\n\n-----
 ----------------\nNeural James-Stein Combiner for Unbiased and Biased Rend
 erings\n\nGu, Iglesias-Guitian, Moon\n\nUnbiased rendering algorithms such
  as path tracing produce accurate images given an infinite number of sampl
 es, but in practice, the techniques often leave visually distracting artif
 acts (i.e., noise) in their rendered images due to a limited time budget. 
 A favored approach for mitigating the noise ...\n\n---------------------\n
 Unbiased Caustics Rendering Guided by Representative Specular Paths\n\nLi,
  Wang, Tu, Xu, Holzschuch...\n\nCaustics are interesting patterns caused b
 y the light being focused when reflecting off glossy materials. Rendering 
 them in Computer Graphics is still challenging: they correspond to high lu
 minous intensity focused over a small area. Finding the paths that contrib
 ute to this small area is difficult,...\n\n---------------------\nScalable
  multi-class sampling via filtered sliced optimal transport\n\nSALAUN, Geo
 giev, Seidel, Singh\n\nWe propose a continuous domain formulation of Wasse
 rstein barycenters\nfor multi-class (-purpose) point set optimization. Our
  formulation is sys-\ntematically derived to handle hundreds to thousands 
 of classes for different\nsampling applications. We develop a practical op
 timization scheme that is\nclos...\n\n---------------------\nGaussian Blue
  Noise\n\nAhmed, Ren, Wonka\n\nAmong the various approaches for producing 
 point distributions with blue noise spectrum, we argue for an optimization
  framework using Gaussian kernels.\nWe show that with a wise selection of 
 optimization parameters, this approach attains unprecedented quality, prov
 ably surpassing the current state of...\n\n---------------------\nCombinin
 g GPU Tracing Methods within a Single Ray Query\n\nBartels, Harada\n\nWe s
 how that breaking up a ray query into multiple tracing calls and processin
 g a subset with distance field tracing results in high quality visual resu
 lts at low cost.\n\n---------------------\nDeep Adaptive Sampling and Reco
 nstruction using Analytic Distributions\n\nSalehi, Manzi, Roethlin, Schroe
 rs, Weber...\n\nWe propose an adaptive sampling and reconstruction method 
 for offline Monte Carlo rendering. Our method produces sampling maps const
 rained by a user-defined budget that minimize the expected future denoisin
 g error. Compared to other state-of-the-art methods, which produce the nec
 essary training data...\n\n\nRegistration Category: FULL ACCESS, ON-DEMAND
  ACCESS\n\nLanguage: ENGLISH\n\nFormat: IN-PERSON, ON-DEMAND
END:VEVENT
END:VCALENDAR