Ensemble Metropolis Light Transport

This article proposes a Markov Chain Monte Carlo ( MCMC ) rendering algorithm based on family of guided transition kernels. The kernels exploit properties ensembles light transport paths, which are distributed according to the lighting in scene, and utilize this information make informed decisions for guiding local path sampling. Critically, our approach does not require caching distributions world space, saving time memory, yet it is able sampling whole paths. We show how can be implemented efficiently by organizing paths each ensemble designing carefully chosen subset from ensemble. easy parallelize leads improvements variance when variety scenes.