Amortizing Samples in Physics-Based Inverse Rendering Using ReSTIR
ACM Transactions on Graphics · 05.12.2023 · 42 (6) · 1-17
DOI: 10.1145/3618331 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Graphics, online erschienen am 05.12.2023. Verfasst von Yu-Chen Wang und 3 weiteren Autor:innen. Die Arbeit wurde bisher 49-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NSF.
Abstract (Originalsprache)
Recently, great progress has been made in physics-based differentiable rendering. Existing differentiable rendering techniques typically focus on static scenes, but during inverse rendering---a key application for differentiable rendering---the scene is updated dynamically by each gradient step. In this paper, we take a first step to leverage temporal data in the context of inverse direct illumination. By adopting reservoir-based spatiotemporal resampled importance resampling (ReSTIR), we introduce new Monte Carlo estimators for both interior and boundary components of differential direct illumination integrals. We also integrate ReSTIR with antithetic sampling to further improve its effectiveness. At equal frame time, our methods produce gradient estimates with up to 100× lower relative error than baseline methods. Additionally, we propose an inverse-rendering pipeline that incorporates these estimators and provides reconstructions with up to 20× lower error.
Förderung
- NSF
Zitieren
APA
Wang, Y., Wyman, C., Wu, L., & Zhao, S. (2023). Amortizing Samples in Physics-Based Inverse Rendering Using ReSTIR. ACM Transactions on Graphics, 42(6), 1-17. https://doi.org/10.1145/3618331
Vancouver
Wang Y, Wyman C, Wu L, Zhao S. Amortizing Samples in Physics-Based Inverse Rendering Using ReSTIR. ACM Transactions on Graphics. 2023;42(6):1-17. doi:10.1145/3618331
BibTeX
@article{Wang2023,
author = {Wang, Yu-Chen and Wyman, Chris and Wu, Lifan and Zhao, Shuang},
title = {Amortizing Samples in Physics-Based Inverse Rendering Using ReSTIR},
journal = {ACM Transactions on Graphics},
year = {2023},
volume = {42},
number = {6},
pages = {1--17},
doi = {10.1145/3618331}
}
EUVIMED