Semi-Procedural Textures using Point Process Texture Basis Functions
EGSR 2020 (Eurographics Symposium on Rendering)
CFG (Computer Graphics Forum)
Submission ID: 1018
PPTBF Appearance Space
This supplemental material provides additional results.
Please click on the images to enlarge them.
References
[Zsolnai-Fehér et al. 2018] Károly Zsolnai-Fehér, Peter Wonka, and Michael Wimmer. 2018. Gaussian material synthesis. ACM Trans. Graph. 37, 4, Article 76 (July 2018), 14 pages. DOI: https://doi.org/10.1145/3197517.3201307
Appearance Space Pipeline
Appearance Space Pipeline
Given an input exemplar, the user selects the number of best matches for nearest neighbors queries. Our implementation uses the FLANN library [Muja and Lowe 2009] for these queries. Based on these primarily results, we generate more results by Monte Carlo sampling. These final results are used to feed the GPR and GPLVM [Lawrence 2003], that combine a Gaussian non-linear regression with a dimensionality reduction technique. This approach has been recently applied for complex material editing [ZsolnaiFehér et al. 2018]. See the accompanying video for a demo.
| Appearance Space navigation |
Additional Results
Please click on the images to enlarge them.