Robust and Efficient Adaptive Direct Lighting Estimation
  Yu-Chi Lai     Hsuan-Ting Chou     Kuo-Wei Chen     Hao-Yu Chang     Shaohua Fan  
Lai, Y.-C., Chou, H.-T., Chen, K.-W.,and Fan, S., Robust and Efficient Adaptive Direct Lighting Estimation, the visual computer, Vol. 31, No. 1, PP 83-91 Submitted material web page

Hemispherical integral is important for the estimation of direct lighting which has major impact on the results of global illumination. This work first proposes the
population Monte Carlo hemispherical integral (PMC-HI) sampler to improve the efficiency of hemispherical integral estimation. The sampler is unbiased and derived from the population Monte Carlo framework. The hemispheric integral sampler works on a population of samples and learns to be a better sampling function over iterations. Information found in one iteration can be used to guide subsequent
iterations by distributing more samples to important sampling techniques such as sampling from light and sampling from BSDF in order to focus more efforts on the sampling sub-domain which has larger contribution to the integral. In
addition, we also propose a cone sampling based on the previous sampling result to enhance the sampling success rate when complex occlusions exist. In addition, we also adapt the image-plane multidimensional adaptive sampling proposed by Hachisuka et al. [14] for multidimensional hemispherical adaptive sampling (MDHI) to estimate the hemispherical integrals deterministically in the random-number
direct lighting space. The images rendered with PMC-HI and MDHI are compared against those rendered with multiple importance sampling (MIS) [29] and adaptive light sample distributions (ALSD) [6]. Our PMC-HI sampler can significantly improve the rendering efficiency. Furthermore, we also show that PMC-HI can be implemented as a plugin to a modern global rendering system.


author = {Lai, Yu-Chi and Chou, Hsuan-Ting and Chen, Kuo-Wei and Fan, Shaohua},
title = {Robust and efficient adaptive direct lighting estimation},
journal = {The Visual Computer},
year = {2015},
volume = {31},
pages = {83--91},
number = {1},


 NSC 99-2218-E-011-005