Abstract overview
Traverse’s published NeuBTF work investigates neural material representation, focusing on the quality, efficiency and memory behaviour of NeuMIP-based bidirectional texture functions. The technique is demonstrated in Mitsuba 2 and in the Breda framework.
The poster was published at SIGGRAPH Asia 2022 Posters, held 6–9 December 2022. The linked project page is the source of the paper, its results and bibliographic details.
Implementation
The public source repository accompanies the work. Read the repository’s licence and implementation notes before using it in a project.
Citation
Luca Quartesan and Carlos Pereira Santos. 2022. Neural Bidirectional Texture Function Compression and Rendering. SIGGRAPH Asia 2022 Posters. DOI: 10.1145/3550082.3564188.
The publication presents the authors’ material-quality and compression comparisons.