Program

Wednesday:

08.00-13.45: Registration
08.00-09.00: Coffee
09.00-12.00: Workshop on Material Appearance Modeling
13.45-14.00: Opening + Welcome
14.00-15.00: Keynote Speaker Tomas Akenine-Möller
15.00-15.15: Break
15.15-16.45: Session Illustration & Stylization
16.45-17.00: Break
17.00-18.15: Session GPU rendering

19.00: Social event city tour

Thursday:

08.00-09.00: Coffee
09.00-10.30: Session Video & Textures
10.30-11.30: Break + Poster Session
11.30-12.30: Session Reconstruction & Denoising
12.30-14.00: Lunch
14.00-15.00: Keynote Speaker Art B. Owen
15.00-15.30: Break
15.30-17.00: Session Sampling
17.00-18.30: EGSR Townhall Meeting (Discussion on the future of EGSR, all attendees welcome)
19.45: Social event Dinner

Friday:

08.00-09.00: Coffee
09.00-10.30: Session Spectral & Wave rendering
10.30-11.00: Break
11.00-12.30: Session Light & Reflectance
12.30-14.00: Lunch
14.00-15.00: Session Visibility
15.00-15.30: Break
15.30-16:30: Invited Speaker Gregory J. Ward
16.30-17.30: Awards Test-of-Time award, best paper award, best student paper award + Closing

Awards:

Best Paper Award: Eric Heitz, Eugene d'Eon, "Importance Sampling Microfacet-Based BSDFs using the Distribution of Visible Normals"

Best Student Paper Award: Oskar Elek, Tobias Ritschel, Pablo Bauszat, Marcus Magnor, Hans-Peter Seidel, "Spectral Ray Differentials"


Keynote Speaker 1: Tomas Akenine-Möller (Lund University, Sweden)

Title: The ART of Stochastic Rasterization

Wednesday 14.00-15.00

Abstract: In this keynote presentation, I will present an overview of recent advances in stochastic rasterization, initially aimed at rendering correct motion blur and depth of field with rasterization-based approaches. One goal is to reveal bits of information about higher-dimensional rasterization that are not often easily extracted from papers. I will use an interactive presentation, and talk about possible hardware modifications as well as what can be done on contemporary graphics processors.

Bio: Tomas Akenine-Möller is a professor in computer graphics at Lund University since 2007. He is a co-author of Real-Time Rendering with Eric Haines and Naty Hoffman, and he has also co-authored a bunch of journal and conference papers. Currently, he is working part time in the Advanced Rendering Technology group at Intel and the other part at Lund University. His research interests include everything concerning graphics processors and ray tracing-based methods.

Keynote Speaker 2: Art B. Owen (Stanford University)

Title: Monte Carlo and quasi-Monte Carlo for image synthesis

Thusday 14.00-15.00

Abstract: This talk surveys ideas in Monte Carlo (MC) and quasi-Monte Carlo (QMC) sampling, with an eye to how they might be useful in image synthesis. The plain idea in MC sampling is to estimate the expectation of a random outcome by a corresponding average of independent draws. This seems a perfect match for rendering problems expressed as averages over photon paths. Two difficulties arise for MC. We may not have any techniques to generate independent samples from the needed distribution. Even when we can get those samples, the result may be too slow to converge. The first problem is addressed by methods such as Markov chain Monte Carlo (MCMC), Sequential Monte Carlo (SMC) and Multi-level Monte Carlo (MLMC). The second problem is addressed by QMC. There have been an enormous variety of MC and QMC techniques developed in recent decades. The path between a new sampling idea and utility in graphics is not always clear; there may be a blockage due to, lack of smoothness, aliasing artifacts, unbounded integrals or difficulty computing a normalizing constant. This talk will look specifically at MC, QMC, randomized QMC, multi-level MC and combinations of QMC with MCMC and SMC. There will also be some new new techniques for sampling within triangles, and some new improvements to multiple importance sampling.

Bio: Art Owen is a Professor of Statistics at Stanford University, He has a Batchelor of Mathematics in Statistics and Computer Science from the University of Waterloo and PhD in Statistics from Stanford. He is the inventor of empirical likelihood, a method of likelihood inference that does not require the user to specify a data distribution. He also developed scrambled nets, a form of randomized quasi-Monte Carlo sampling which supplies computable error estimates for QMC while also improving its performance. In addition to empirical likelihood and quasi-Monte Carlo, he works high dimensional transposable style data sets that come up in bioinformatics and Internet applications.

Download Slides

Invited Talk: Gregory J. Ward (Dolby Laboratories, Inc.)

Title: Rendering as the premier tool of human imagination

Friday 15.30-16.30

Abstract: During the past 25 years, computer graphics rendering has become the mainstay of both the design and entertainment worlds. Consumer and professional photography are nearly 100% digital today, with studios pushing for the same in movies. Modern architects and engineers could not perform their jobs without the CAD, simulation, and visualization tools that have computer graphics as a core technology. Special effects are longer "special," since they can be found in nearly every modern production. Our field has gone from a novelty to a staple, surpassing the original goal of realistic rendering to become a part of everyone's reality. Our success has been spectacular, and the impact of that success profound. Ultimately, CG rendering offers a new vehicle for exploration and the highest bandwidth channel for communicating new ideas to others. We will look at a few of the possibilities that CG enables, suggesting how rendering provides the key to imagining and navigating our future as a species.

Bio: Greg Ward has been involved in 3-D and 2-D computer graphics since 1985, and is the principal author of the widely referenced Radiance lighting simulation and rendering system. His research includes global illumination, reflectance models, high dynamic range image processing and color perception. He has been employed by the Lawrence Berkeley National Laboratory, EPFL Switzerland, SGI, Shutterfly, Exponent, and BrightSide Technologies. Greg holds a bachelor's in Physics from UC Berkeley and a master's in Computer Science from SF State University. He is currently employed by Dolby Laboratories, Inc. and consults for the Lawrence Berkeley National Laboratory.

Session Illustration & Stylization Wednesday 15.15-16.45

Session Chair: George Drettakis

  • Color Me Noisy: Example-based Rendering of Hand-colored Animations with Temporal Noise Control.
    Jakub Fišer, Michal Lukáč, Ondřej Jamriška, Martin Cadik, Yotam Gingold, Paul Asente, Daniel Sykora
  • Vectorising Bitmaps into Semi-Transparent Gradient Layers.
    Christian Richardt, Jorge Lopez-Moreno, Adrien Bousseau, Maneesh Agrawala, George Drettakis
  • AutoStyle: Automatic Style Transfer from Image Collections to Users' Images.
    Yiming Liu, Michael Cohen, Matt Uyttendaele, Szymon Rusinkiewicz.

Session GPU rendering Wednesday 17.00-18.00

Session Chair: Samuli Laine

  • Bilinear Accelerated Filter Approximation.
    Josiah Manson, Scott Schaefer.
  • C-LOD: Context-aware Material Level-of-Detail applied to Mobile Graphics.
    George Alex Koulieris, George Drettakis, Douglas Cunningham, Katerina Mania
  • Massively Parallel Hierarchical Scene Processing with Applications in Rendering.
    Marek Vinkler, Jiří Bittner, Vlastimil Havran, Michal Hapala

Session Video & Textures Thursday 9:00-10.30

Session Chair: Szymon Ruskinkiewicz

  • Semi-Automated Video Morphing.
    Jing Liao, Diego Nehab, Hugues Hoppe, Pedro Sander
  • User-Assisted Video Stabilization.
    Jiamin Bai, Aseem Agarwala, Maneesh Agrawala, Ravi Ramamoorthi
  • Interactive Parameter Retrieval for Two-Tone Procedural Textures.
    Lena Gieseke, Sebastian Koch, Jens Uwe Hahn, Martin Fuchs

Session Reconstruction & Denoising Thursday 11:30-12.30

Session Chair: Eric Tabellion

  • Layered Reconstruction for Defocus and Motion Blur.
    Jacob Munkberg, Karthik Vaidyanathan, Jon Hasselgren, Petrik Clarberg, Tomas Akenine-Möller
  • Robust Image Denoising using a Virtual Flash Image for Monte Carlo Ray Tracing.
    Bochang Moon, Jong Yun Jun, Jong Hyeob Lee, Kunho Kim, Toshiya Hachisuka, Sung-Eui Yoon

Session Sampling Thursday 15.30-17.00

Session Chair: Wenzel Jakob

  • Convergence analysis of direct illumination estimators that use combinations of stochastic sampling strategies.
    Kartic Subr, Derek Nowrouzezahrai, Wojciech Jarosz, Jan Kautz, Kenny Mitchell
  • Spherical Fibonacci Point Sets for Illumination Integrals.
    R. Marques, C. Bouville, M. Ribardière, L. P. Santos, K. Bouatouch
  • Importance Sampling Microfacet-Based BSDFs using the Distribution of Visible Normals.
    Eric Heitz, Eugene d'Eon

Session Spectral & Wave rendering Friday 9.00-10.30

Session Chair: Per Christensen

  • Spectral Ray Differentials.
    Oskar Elek, Tobias Ritschel, Pablo Bauszat, Marcus Magnor, Hans-Peter Seidel
  • Full Wave Modeling of Light Propagation and Reflection.
    A. Musbach, G.W. Meyer, F. Reitich, S.H. Oh
  • Hero Wavelength Spectral Sampling.
    Alexander Wilkie, Sehera Nawaz, Marc Droske, Andrea Weidlich, Johannes Hanika

Session Light & Reflectance Friday 11.00-12.30

Session Chair: Greg Ward

  • A Physically-Based BSDF for Modeling the Appearance of Paper.
    Marios Papas, Krystle de Mesa, Henrik Wann Jensen
  • Robust Fitting on Poorly Sampled Data for Surface Light Field Rendering and Image Relighting.
    K. Vanhoey, B. Sauvage, O. Génevaux, F. Larue, J.M. Dischler
  • Multiple light source estimation in a single image.
    Jorge Lopez-Moreno, Elena Garces, Sunil Hadap, Erik Reinhard, Diego Gutierrez

Session Visibility Friday 14.00-15.00

Session Chair: Naty Hoffman

  • Visibility Silhouettes for Semi-Analytic Spherical Integration.
    Derek Nowrouzezahrai, Ilya Baran, Kenny Mitchell, Wojciech Jarosz
  • Probabilistic visibility evaluation using geometric proxies.
    Niels Billen, Ares Lagae, Philip Dutre


EGSR town hall meeting:

Discussion on the future of EGSR, all attendees welcome

Thursday 17.00-18.30

The EGSR town hall meeting will be broadcast to the community. If you want to watch, please follow this Youtube live stream: https://www.youtube.com/watch?v=fCjCALaflTE

Google+ users can also follow this event: https://plus.google.com/events/crq0lai22m9baenj41hltta9ja0