We are going to talk about techniques that create physically based material models from photographs that we can use in our light simulation programs. In an earlier work, two photographs are required for high-quality reconstruction. It seems that working from only one photograph doesn’t seem possible at all. However, with the power of deep learning…
___________________________________
The paper “Two-Shot SVBRDF Capture for Stationary Materials” is available here:
The paper “Reflectance Modeling by Neural Texture Synthesis” is available here:
NVIDIA has implemented the two-shot model! Have a look:
Our earlier episode on Gradient Domain Light Transport is available here:
The light transport course at the Technical University of Vienna is available here:
WE WOULD LIKE TO THANK OUR GENEROUS PATREON SUPPORTERS WHO MAKE TWO MINUTE PAPERS POSSIBLE:
David Jaenisch, Sunil Kim, Julian Josephs, Daniel John Benton, Dave Rushton-Smith, Benjamin Kang.
We also thank Experiment for sponsoring our series. –
Image credits:
Sherrie Thai –
Giulio Bernardi –
Andy Beatty –
Open Grid Scheduler –
liz west –
Subscribe if you would like to see more of these! –
Music: Dat Groove by Audionautix is licensed under a Creative Commons Attribution license (
Artist:
Splash screen/thumbnail design: Felícia Fehér –
Károly Zsolnai-Fehér’s links:
Facebook →
Twitter →
Web →
source