BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Melbourne
X-LIC-LOCATION:Australia/Melbourne
BEGIN:DAYLIGHT
TZOFFSETFROM:+1000
TZOFFSETTO:+1100
TZNAME:AEDT
DTSTART:19721003T020000
RRULE:FREQ=YEARLY;BYMONTH=4;BYDAY=1SU
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:19721003T020000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20240214T070245Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231213T181900
DTEND;TZID=Australia/Melbourne:20231213T182900
UID:siggraphasia_SIGGRAPH Asia 2023_sess127_papers_646@linklings.com
SUMMARY:LitNeRF: Intrinsic Radiance Decomposition for High-Quality View Sy
 nthesis and Relighting of Faces
DESCRIPTION:Technical Communications, Technical Papers\n\nKripasindhu Sark
 ar (Google Inc.); Marcel Bühler and Simon Li (ETH Zürich, Google Inc.); an
 d Daoye Wang, Delio Vicini, Jérémy Riviere, Yinda Zhang, Sergio Orts-Escol
 ano, Paulo Gotardo, Thabo Beeler, and Abhimitra Meka (Google Inc.)\n\nHigh
 -fidelity, photorealistic 3D capture of a human face is a long-standing pr
 oblem in computer graphics -- the complex material of skin, intricate geom
 etry of hair, and fine scale textural details make it challenging. Traditi
 onal techniques rely on very large and expensive capture rigs to reconstru
 ct explicit mesh geometry and appearance maps and require complex differen
 tiable path-tracing to achieve photorealistic results. More recent volumet
 ric methods (\eg, NeRFs) have enabled view-synthesis and sometimes relight
 ing by learning an implicit representation of the density and reflectance 
 basis, but suffer from artifacts and blurriness due to the inherent ambigu
 ities in volumetric modeling. These problems are further exacerbated when 
 capturing with few cameras and light sources. We present a novel technique
  for high-quality capture of a human face for 3D view synthesis and religh
 ting using a sparse, compact capture rig consisting of 15 cameras and 15 l
 ights. Our method combines a volumetric representation of the face reflect
 ance with traditional multi-view stereo based geometry reconstruction. The
  proxy geometry allows us to anchor the 3D density field to prevent artifa
 cts and guide the disentanglement of intrinsic radiance components of the 
 face appearance such as diffuse and specular reflectance, and incident rad
 iance (shadowing) fields. \nOur hybrid representation significantly improv
 es the state-of-the-art quality for arbitrarily dense renders of a face fr
 om desired camera viewpoint as well as environmental, directional, and nea
 r-field lighting.\n\nRegistration Category: Full Access\n\nSession Chair: 
 Jernej Barbic (University of Southern California)
URL:https://asia.siggraph.org/2023/full-program?id=papers_646&sess=sess127
END:VEVENT
END:VCALENDAR