Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Engage Grants Program

Université de Montréal

$25,000.00

Interactive Covariance Tracing for Realistic Image Synthesis

FY 2014-15

Summary
In order to synthesize realistic images of virtual worlds, the steady-state energy balance of light in a scene mustbe simulated. The most general case, commonly referred to as the ""global illumination"" of a scene, considers allcomplex reflection, refraction, multi-bounce and shadowing effects. Generating images of virtual scenes withglobal illumination mandates comprehensive mathematical models of light transport, as well ascomputationally efficient techniques for resolving such models. Fabric Software Inc.'s FabricEngine graphicsplatform comprises, among other subsystems, an interactive global illumination solver; however, this kernel isnot capable of solving the general global illumination problem and, as such, the resulting images generated bythe platform appear clearly artificial.A recently developed ""covariance tracing"" approach proposes an alternative mathematical formulation to lighttransport that tracks, propagates, and warps the statistical properties of light packets as they travel and interactwith a virtual scene. In doing so, the computational requirements for generating true global illuminationsolutions is significantly reduced compared to traditional Monte Carlo solutions to the problem. Unfortunately,covariance tracing is not well-adapted to the readily available parallel processing capabilities of cloudcomputing services and commodity graphics hardware processors.The goal of this research collaboration is to adapt covariance tracing, devising new mathematicalrepresentations and algorithmic formulations, so as to permit GPU accelerated (ideally interactive) solutions tothe global illumination problem. Such a system would not only further the capabilities of the FabricEngineplatform, but will allow artists in feature-film production and video game content generation to more faithfullypreview and edit their final shots, leading to a more flexible and efficient authoring workflow.
Program
Engage Grants Program
Location
Quebec, Canada
Recipient Type
Academia
Type
Grant
Source
NSERC's Awards Data

As published

Application id
466450-2014
Name nom
Nowrouzezahrai, Derek
Department département
Informatique et recherche opérationnelle
Organization id
63
Institution établissement
Université de Montréal
Province en
Québec
Province fr
Québec
Country en
CANADA
Country fr
CANADA
Fiscal year exercice financier
2014
Competition year année de concours
2014
Award amount
25000
Program id
EGP
Program name en
Engage Grants Program
Program name fr
Programme de subventions d'engagement partenariat
Group en
Research partnerships programs
Group fr
Programmes de partenariat
Committee code
1552
Committee name en
Quebec Internal Decision Committee
Committee name fr
Comité de décision interne pour le Québec
Area of application code
801
Area of application group en
Information and communication services
Area of application group fr
Services d'information et de télécommunications
Area of application en
Computer software
Area of application fr
Logiciel
Research subject code
2707
Research subject group en
Information technology
Research subject group fr
Technologie de l'information
Research subject en
Graphics
Research subject fr
Infographie
Partie
1
Nb partie
1
Application title
Interactive Covariance Tracing for Realistic Image Synthesis
Keyword
NSERC
Application summary
In order to synthesize realistic images of virtual worlds, the steady-state energy balance of light in a scene mustbe simulated. The most general case, commonly referred to as the ""global illumination"" of a scene, considers allcomplex reflection, refraction, multi-bounce and shadowing effects. Generating images of virtual scenes withglobal illumination mandates comprehensive mathematical models of light transport, as well ascomputationally efficient techniques for resolving such models. Fabric Software Inc.'s FabricEngine graphicsplatform comprises, among other subsystems, an interactive global illumination solver; however, this kernel isnot capable of solving the general global illumination problem and, as such, the resulting images generated bythe platform appear clearly artificial.A recently developed ""covariance tracing"" approach proposes an alternative mathematical formulation to lighttransport that tracks, propagates, and warps the statistical properties of light packets as they travel and interactwith a virtual scene. In doing so, the computational requirements for generating true global illuminationsolutions is significantly reduced compared to traditional Monte Carlo solutions to the problem. Unfortunately,covariance tracing is not well-adapted to the readily available parallel processing capabilities of cloudcomputing services and commodity graphics hardware processors.The goal of this research collaboration is to adapt covariance tracing, devising new mathematicalrepresentations and algorithmic formulations, so as to permit GPU accelerated (ideally interactive) solutions tothe global illumination problem. Such a system would not only further the capabilities of the FabricEngineplatform, but will allow artists in feature-film production and video game content generation to more faithfullypreview and edit their final shots, leading to a more flexible and efficient authoring workflow.