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:20260114T163711Z
LOCATION:Meeting Room C4.11\, Level 4 (Convention Centre)
DTSTART;TZID=Australia/Melbourne:20231213T145000
DTEND;TZID=Australia/Melbourne:20231213T150500
UID:siggraphasia_SIGGRAPH Asia 2023_sess128_papers_227@linklings.com
SUMMARY:CamP: Camera Preconditioning for Neural Radiance Fields
DESCRIPTION:Keunhong Park, Phillip Henzler, Ben Mildenhall, Jonathan T. Ba
 rron, and Ricardo Martin-Brualla (Google Research)\n\nNeural Radiance Fiel
 ds (NeRF) can be optimized to obtain high-fidelity 3D scene reconstruction
 s of objects and large-scale scenes. However, NeRFs require accurate camer
 a parameters as input --- inaccurate camera parameters result in blurry re
 nderings. Extrinsic and intrinsic camera parameters are usually estimated 
 using Structure-from-Motion (SfM) methods as a pre-processing step to NeRF
 , but these techniques rarely yield perfect estimates. Thus, prior works h
 ave proposed jointly optimizing camera parameters alongside a NeRF, but th
 ese methods are prone to local minima in challenging settings.\n\nIn this 
 work, we analyze how different camera parameterizations affect this joint 
 optimization problem, and observe that standard parameterizations exhibit 
 large differences in magnitude with respect to small perturbations, which 
 can lead to an ill-conditioned optimization problem. We propose using a pr
 oxy problem to compute a whitening transform that eliminates the correlati
 on between camera parameters and normalizes their effects, and we propose 
 to use this transform as a preconditioner for the camera parameters during
  joint optimization.\n\nOur preconditioned camera optimization significant
 ly improves reconstruction quality on scenes from the Mip-NeRF 360 dataset
 : we reduce error rates (RMSE) by 67% compared to state-of-the-art NeRF ap
 proaches that do not optimize for cameras like Zip-NeRF, and by 29% relati
 ve to state-of-the-art joint optimization approaches using the camera para
 meterization of SCNeRF. Our approach is easy to implement, does not signif
 icantly increase runtime, can be applied to a wide variety of camera param
 eterizations, and can straightforwardly be incorporated into other NeRF-li
 ke models.\n\nRegistration Category: Full Access\n\nSession Chair: Jianfei
  Cai (Monash University)\n\n
URL:https://asia.siggraph.org/2023/full-program?id=papers_227&sess=sess128
END:VEVENT
END:VCALENDAR