Technical Papers Fast-Forward :
Variance-Aware Multiple Importance Sampling
Technical Papers Fast-Forward
TimeSunday, 17 November 201918:01 - 18:02
LocationGreat Hall 1&2
DescriptionMany existing Monte Carlo methods rely on multiple importance sampling (MIS) to achieve robustness and versatility. Typically, the balance or power heuristics are used, mostly thanks to the seemingly strong guarantees regarding their variance. We show that these MIS heuristics are oblivious to the effect of certain variance reduction techniques like stratification. This shortcoming is particularly pronounced when unstratified and stratified techniques are combined (e.g., in a bidirectional path tracer). We propose to enhance the balance heuristic by injecting variance estimates. We achieve substantial variance reduction for combinations of stratified and unstratified techniques, as well as defensive sampling applications. The proposed method is simple to implement and introduces little overhead.