BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Learning Light Scattering Distributions
DTSTART:20180830T093000
DTEND:20180830T113000
DTSTAMP:20260406T185151Z
UID:9d5a1e76cb494c5337970797c100a37ab6b44d1672b609b9dbaf735b
CATEGORIES:Conferences - Seminars
DESCRIPTION:Delio Vicini\nEDIC candidacy exam\nExam president: Prof. Pasca
 l Fua\nThesis advisor: Prof. Wenzel Jakob\nCo-examiner: Prof. Mark Pauly\n
 \nAbstract\nRendering photorealistic imagery remains challenging\, as ever
  growing\ndemands on visual fidelity quickly make up for any gains in hard
 ware performance\nand algorithmic improvements. Our research objective is 
 to further\nreduce the time required to render complex light transport eff
 ects by using\nmachine learning to model the distribution of scattered lig
 ht. In this proposal\,\nwe summarize three related previous papers. First\
 , we describe a paper on\nthe efficient rendering of subsurface scattering
 . We then summarize a paper\nintroducing variational autoencoders\, a gene
 rative neural network architecture\,\nwhich allows to learn how to sample 
 from arbitrary distributions. The\nthird paper presents an online learning
  method to learn the radiance distribution\nduring rendering. We further p
 resent a novel approach to rendering\nsubsurface scattering using a variat
 ional autoencoder.\nIn the future\, we could apply a similar architecture 
 to more general light transport problems.\n\nBackground papers\nPhoton Bea
 m Diffusion: A Hybrid Monte Carlo Method for Subsurface Scattering\, by Ha
 bel et al.\nAuto-Encoding Variational Bayes\, by Kingma and Welling\nPract
 ical Path Guiding for Efficient Light-Transport Simulation\, by Müller et
  al.\n\n\n 
LOCATION:BC 329 https://plan.epfl.ch/?room==BC%20329
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR