BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Australia/Brisbane
X-LIC-LOCATION:Australia/Brisbane
BEGIN:STANDARD
DTSTART:19920301T030000
TZOFFSETFROM:+1100
TZOFFSETTO:+1000
TZNAME:AEST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20191203T080501Z
LOCATION:Plaza Meeting Room P2
DTSTART;TZID=Australia/Brisbane:20191119T115200
DTEND;TZID=Australia/Brisbane:20191119T121800
UID:siggraphasia_SIGGRAPH Asia 2019_sess112_papers_149@linklings.com
SUMMARY:Mitsuba 2: A Retargetable Forward and Inverse Renderer
DESCRIPTION:Technical Papers\n\nMitsuba 2: A Retargetable Forward and Inve
 rse Renderer\n\nNimier-David, Vicini, Zeltner, Jakob\n\nModern rendering s
 ystems are confronted with a dauntingly large and growing set of requireme
 nts: in their pursuit of realism, physically based techniques must increas
 ingly account for intricate properties of light, such as its spectral comp
 osition or polarization. To reduce prohibitive rendering times, vectorized
  renderers exploit coherence via instruction-level parallelism on CPUs and
  GPUs. Differentiable rendering algorithms propagate derivatives through a
  simulation to optimize an objective function, e.g., to reconstruct a scen
 e from reference images. Catering to such diverse use cases is challenging
  and has led to numerous purpose-built systems—partly, because retrofittin
 g features of this complexity onto an existing renderer involves an error-
 prone and infeasibly intrusive transformation of elementary data structure
 s, interfaces between components, and their implementations (in other word
 s, everything).\n\nWe propose Mitsuba 2, a versatile renderer that is intr
 insically retargetable to various applications including the ones listed a
 bove. Mitsuba 2 is implemented in modern C++ and leverages template metapr
 ogramming to replace types and instrument the control flow of components s
 uch as BSDFs, volumes, emitters, and rendering algorithms. At compile time
 , it automatically transforms arithmetic, data structures, and function di
 spatch, turning generic algorithms into a variety of efficient implementat
 ions without the tedium of manual redesign. Possible transformations inclu
 de changing the representation of color, generating a "wide" renderer that
  operates on bundles of light paths, just-in-time compilation to create co
 mputational kernels that run on the GPU, and forward/reverse-mode automati
 c differentiation. Transformations can be chained, which further enriches 
 the space of algorithms derived from a single generic implementation.\n\nW
 e demonstrate the effectiveness and simplicity of our approach on several 
 applications that would be very challenging to create without assistance: 
 a rendering algorithm based on coherent MCMC exploration, a caustic design
  method for gradient-index optics, and a technique for reconstructing hete
 rogeneous media in the presence of multiple scattering.\n\nRegistration Ca
 tegory: Platinum Pass, Full Conference Pass, Full Conference One-Day Pass
URL:https://sa2019.conference-program.com/presentation?id=papers_149&sess=
 sess112
END:VEVENT
END:VCALENDAR