Neural Denoising for Spectral Monte Carlo Rendering

Abstract
Spectral Monte Carlo (MC) rendering is still to be largely adopted partially due to the specific noise, called color noise, induced by wavelength-dependent phenomenons. Motivated by the recent advances in Monte Carlo noise reduction using Deep Learning, we propose to apply the same approach to color noise. Our implementation and training managed to reconstruct a noise-free output while conserving high-frequency details despite a loss of contrast. To address this issue, we designed a three-step pipeline using the contribution of a secondary denoiser to obtain high-quality results.
Description

CCS Concepts: Computing methodologies --> Ray tracing; Neural networks; Image processing

        
@inproceedings{
10.2312:egp.20221011
, booktitle = {
Eurographics 2022 - Posters
}, editor = {
Sauvage, Basile
and
Hasic-Telalovic, Jasminka
}, title = {{
Neural Denoising for Spectral Monte Carlo Rendering
}}, author = {
Rouphael, Robin
and
Noizet, Mathieu
and
Prévost, Stéphanie
and
Deleau, Hervé
and
Steffenel, Luiz-Angelo
and
Lucas, Laurent
}, year = {
2022
}, publisher = {
The Eurographics Association
}, ISSN = {
1017-4656
}, ISBN = {
978-3-03868-171-7
}, DOI = {
10.2312/egp.20221011
} }
Citation