Presentation / Installation


Inverse Global Illumination Using a Neural Radiometric Prior
SessionInverse Rendering: Does Anybody Know How I Got Here?
DescriptionWe propose to recover scene parameters using a neural radiometric prior, which enforces a physical constraint on the optimization such that it accounts for the global illumination effects. Compared to methods based on building path integrals, we can recover scene parameters with comparable quality at considerably lower computation times.
Authors
Saeed Hadadan
University of Maryland, College Park
NVIDIA
Geng Lin
University of Maryland, University of Maryland, College Park
Jan Novák
NVIDIA
Fabrice Rousselle
NVIDIA
Matthias Zwicker
University of Maryland, University of Maryland, College Park
Event Type
Technical Paper
TimeTuesday, 8 August 202311:18am - 11:28am PDT
LocationPetree Hall C
ACM Digital Library Technical Paper PDF
Session Time & Location
Sunday, 6 August 20236pm - 8:30pm PDTWest Hall B
Tuesday, 8 August 202310:45am - 12:15pm PDTPetree Hall C
Interest Areas
Research & Education
Recordings
Livestreamed
Recorded
Keywords
Rendering
Registration Categories
FC
FCS
V
VS
EFC
Next PresentationNext Presentation
Parameter-space ReSTIR for Differentiable and Inverse Rendering