The SIGGRAPH 2012 Technical Papers program is the premier international forum for disseminating new scholarly work in computer graphics and interactive techniques. The 39th International Conference and Exhibition on Computer Graphics and Interactive Techniques, 5–9 August 2012 at the Los Angeles Convention Center, features Technical Papers that detail new advances across many fronts, including 3D display technology, photographic prints for HDR images, textile design, and more.
“This year’s program features papers that propel the field forward as rapidly and vibrantly as possible, while at the same time keeping us well-grounded academically, ensuring that SIGGRAPH remains a breeding ground for significant new areas of research,” said Hanspeter Pfister, SIGGRAPH 2012 Technical Papers Chair from Harvard University.
The papers to be presented were chosen by a distinguished committee of academia and industry experts. Out of 449 submissions, 94 papers were accepted to SIGGRAPH 2012, representing an acceptance rate of 21 percent.
This year’s Technical Papers program also includes conference presentations for 38 papers published this year in the journal ACM Transactions on Graphics (TOG). A complete listing of all selected Technical Papers will be available on the SIGGRAPH 2012 web site in late May.
Highlights from the SIGGRAPH 2012 Technical Papers program:
Tensor Displays: Compressive Light Field Synthesis Using Multilayer Displays with Directional Backlighting
Authors: Gordon Wetzstein, Douglas Lanman, Matthew Hirsch, and Ramesh Raskar, MIT Media Lab
This paper presents a new display technology for glasses-free stereo viewing. Combining multiple layers of LCD, directional back-lighting, temporal modulation, and a new mathematical formulation allows for greater depths of field, wider fields of view, and a thinner display.
Practical application as suggested by the Technical Papers Chair: Glasses-free stereo display.
Printing Spatially-Varying Reflectance for Reproducing HDR Images
Authors: Yue Dong, Microsoft Research Asia; Xin Tong, Microsoft Research Asia; Fabio Pellacini, Dartmouth College and Sapienza University of Rome; and Baining Guo, Microsoft Research Asia
When traditional photographs are printed, the range of brightness can be heavily compressed, and the result can look flat. This paper presents a solution for viewing HDR images using a reflective sheet of paper, glossy ink, and a torch light illuminating the paper. With the proposed technique, one can get a better sense of the range of brightness in the scene and adjust it by moving the light or the paper.
Practical application as suggested by the Technical Papers Chair: Printing high-dynamic-range images, keeping brightness fidelity.
Structure-Aware Synthesis for Predictive Woven Fabric Appearance
Authors: Shuang Zhao, Wenzel Jakob, Steve Marschner, and Kavita Bala, Cornell University
3D scans of real fabrics to design new 3D fabrics for rendering. The fabrics are rendered in a physically based way, which allows textile designers to predict how a given weave pattern would look if it were fabricated, resulting in highly realistic results for textile design, ecommerce, entertainment, and apparel visualization.
Practical application as suggested by the Technical Papers Chair: Designing real fabrics by predicting their appearance.
Precomputed Acceleration Noise for Improved Rigid-Body Sound
Authors: Jeffrey N. Chadwick, Changxi Zheng, and Doug L. James, Cornell University
Colliding objects produce impact sounds, usually described by modal synthesis. This paper introduces the effect of acceleration in this model, thus allowing for rendering more realistic, crisper collision sounds, or making a sound audible when the modal component only would be inaudible by humans.
Practical application as suggested by the Technical Papers Chair: Improving collision-sound realism in virtual environments.
Design of Self-Supporting Surfaces
Authors: Etienne Vouga, Columbia University/King Abdullah University of Science and Technology; Mathias Höbinger, Evolute/TU Wien, Johannes Wallner, TU Graz/TU Wien; Helmut Pottmann, King Abdullah University of Science and Technology
Finding architectural shapes that are self-supporting is a major challenge in masonry. Discrete differential geometry allows the authors to propose a non-linear optimization process approximating a given surface by a self-supporting one. They also produce a quad mesh with planar faces guiding steel/glass constructions.
Practical application as suggested by the Technical Papers Chair: A tool for architects to design constructions that have not been seen before.
Eulerian Video Magnification for Revealing Subtle Changes in the World
Authors: Hao-Yu Wu, Michael Rubinstein, Massachusetts Institute of Technology CSAIL; Eugene Shih, Quanta Research Cambridge, Inc.; Frédo Durand, William Freeman, Massachusetts Institute of Technology, CSAIL
Using an Eulerian formulation and signal-processing principles, the authors amplify subtle changes in videos that could not be perceived otherwise.
Practical application as suggested by the Technical Papers Chair: Easier monitoring of adult and infant patients.
CrossShade: Shading Concept Sketches Using Cross-Section Curves
Authors: Cloud Shao, University of Toronto; Adrien Bousseau, REVES - INRIA Sophia Antipolis; Alla Sheffer, The University of British Columbia; Karan Singh, University of Toronto
Cross-section curves are often used by artists to depict man-made objects. This paper exploits perceptual principles to derive a framework for shading these sketches, reconstructing appropriate normals and rendering them in a 3D-like fashion.
Practical application as suggested by the Technical Papers Chair: 3D-like rendering of sketches for improved surface depiction.
What Makes Paris Look Like Paris?
Authors: Carl Doersch, Saurabh Singh, and Abhinav Gupta, Carnegie Mellon University; Josef Sivic, INRIA/ Ecole Normale Sup´erieure, Paris; Alexei A. Efros, Carnegie Mellon University, INRIA/Ecole Normale Sup´erieure, Paris
Particular features (such as windows, lamps, trees) in photographs indicate where the images were captured. Using machine learning techniques and large collections of geotagged images, this technology extracts such elements from photographs and guesses where they were taken.
Practical application as suggested by the Technical Papers Chair: Automatic geolocalization of photographs.