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:20240214T070244Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231213T143500
DTEND;TZID=Australia/Melbourne:20231213T145000
UID:siggraphasia_SIGGRAPH Asia 2023_sess128_papers_914@linklings.com
SUMMARY:GANeRF: Leveraging Discriminators to Optimize Neural Radiance Fiel
 ds
DESCRIPTION:Technical Papers\n\nBarbara Roessle and Norman Müller (Technic
 al University of Munich); Lorenzo Porzi, Samuel Rota Bulò, and Peter Konts
 chieder (Meta Reality Labs); and Matthias Niessner (Technical University o
 f Munich)\n\nNeural Radiance Fields (NeRF) have shown impressive novel vie
 w synthesis results; nonetheless, even thorough recordings yield imperfect
 ions in reconstructions, for instance due to poorly observed areas or mino
 r lighting changes.\nOur goal is to mitigate these imperfections from vari
 ous sources with a joint solution: we take advantage of the ability of gen
 erative adversarial networks (GANs) to produce realistic images and use th
 em to enhance realism in 3D scene reconstruction with NeRFs. \nTo this end
 , we learn the patch distribution of a scene using an adversarial discrimi
 nator, which provides feedback to the radiance field reconstruction, thus 
 improving realism in a 3D-consistent fashion. \nThereby, rendering artifac
 ts are repaired directly in the underlying 3D representation by imposing m
 ulti-view path rendering constraints. \nIn addition, we condition a genera
 tor with multi-resolution NeRF renderings which is adversarially trained t
 o further improve rendering quality. \nWe demonstrate that our approach si
 gnificantly improves rendering quality, e.g., nearly halving LPIPS scores 
 compared to Nerfacto while at the same time improving PSNR by 1.4dB on the
  advanced indoor scenes of Tanks and Temples.\n\nRegistration Category: Fu
 ll Access\n\nSession Chair: Jianfei Cai (Monash University)
URL:https://asia.siggraph.org/2023/full-program?id=papers_914&sess=sess128
END:VEVENT
END:VCALENDAR