Stavros Diolatzis, Inria & Université Côte d'Azur, France
Julien Philip, Inria & Université Côte d'Azur, France and Adobe Research, UK
George Drettakis, Inria & Université Côte d'Azur, France
Supplemental PDF
For each scene we provide:
(1) A ground truth rendered path
(2) The same rendered path with our method
(3) The same rendered path with uniform training
Additionally we provide:
(1) For Bedroom paths after 5 and 11 hours of training
(2) For Living Room, Glass Egg and Spaceship paths by Işık et al. 2021 (ANF)






