Home | english  | Impressum | Sitemap | KIT

Local quasi-Monte Carlo Exploration

Lorenzo Tessari, Johannes Hanika, and Carsten Dachsbacher

Proceedings of Eurographics Symposium on Rendering 2017 (EI&I track)

Equal time comparison (15m): Regular quasi-Monte Carlo light transport (Halton points, left) can be inefficient at exploring local features such as the caustic in on the iris or the subsurface scattering in the sclera of this eye. Markov chain methods such as Kelemen Metropolis light transport (KMLT, middle) explore such features better but suffer from clumping artifacts leading to temporal inconsistency in animation. We propose a local quasi-Monte Carlo integration scheme (LQMC, right) which uses stratified point sets for local exploration of lighting features, leading to more even convergence.


In physically-based image synthesis, the path space of light transport paths is usually explored by stochastic sampling. The two main families of algorithms are Monte Carlo/quasi-Monte Carlo sampling and Markov chain Monte Carlo. While the former is known for good uniform discovery of important regions, the latter facilitates efficient exploration of local effects. We introduce a hybrid sampling technique which uses quasi-Monte Carlo points to achieve good stratification in both stages: we use the Halton sequence to generate initial seed paths and rank-1 lattices for local exploration. This method avoids the hard problem of introducing QMC sequences into MCMC while still stratifying samples both globally and locally. We propose perturbation strategies that prefer dimensions close to the camera, facilitating efficient reuse of transport path suffixes. This framework provides maximum control of the sampling scheme by the programmer, which can be hard to achieve with Markov chain-based methods. We show that local QMC exploration can generate results on par with state of the art light transport sampling methods, while providing more uniform convergence, improving temporal consistency.



Temporal stability video


  title = {Local Quasi-{Monte Carlo} Exploration},
  author = {Lorenzo Tessari and Johannes Hanika and Carsten Dachsbacher},
  year = 2017,
  booktitle = {Proceedings of the Eurographics Symposium on Rendering: Experimental Ideas \& Implementations},
  month = jun,
  pages = {to appear}