Call for Nominations

Agenda

Invited Participants and Presentations

• Jatin Chhugani, Johns Hopkins University
  vLOD: A System for High-Fidelity Walkthroughs of Large Virtual Environments
  
  Two effective methods for efficient rendering of large polygonal models are the conservative culling of invisible geometry early in the pipeline and the removal of unnecessary detail not discernible from the current view-point. We present vLOD, a system for performing high-fidelity walkthroughs of large 3D geometric datasets. Our walkthrough system performs work proportional to the required detail in visible geometry. To accomplish this, we use a pre-computation phase to determine from-region visibility as well as level-of-detail. This pre-computation generates per-region vLOD, the geometry visible from a view cell at the right level of detail. We encode changes between neighboring cells' vLODs, which are not required to be memory resident. At rendering time, we incrementally construct the vLOD for the current view cell and render it. We have a small CPU and memory requirement for rendering and are able to display models with tens of million polygons at interactive frame rates with less than one pixel screen space deviation and correct visibility.
  

• Oana Cula, Rutgers, The State University of New Jersey
  Bidirectional Imaging and Modeling of Skin Texture
  
  Human skin is a complex surface, with fine scale geometry and local optical properties that make its appearance difficult to model. Also, the conditions under which the skin surface is viewed and illuminated greatly affect its appearance. In this work, we capture the dependency of skin appearance on imaging parameters using bidirectional imaging. We construct a new skin texture database, containing bidirectional measurements of normal skin and of skin affected by various disorders. The complete database contains more than 3500 images, and is made publicly available for further research. Furthermore, we present two computational models for use in skin texture recognition. Both models are image-based representations of skin appearance that account for the varied appearance of the skin with changes in illumination and viewing direction. We employ these models in two contexts: discrimination between different skin disorders (e.g., psoriasis vs. acne), and classification of facial locations based on facial skin texture (e.g., forehead vs. chin). The classification experiments demonstrate the usefulness of the modeling and measurement methods.
  

• Antonio Haro, Georgia Institute of Technology
  Example Based Processing for Image and Video Synthesis
  
  The example based processing problem can be expressed as: "Given an example of an image or video before and after processing, apply a similar processing to a new image or video". There exist some problems where a single general algorithm can be used to create varieties of outputs, solely by presenting examples of what is desired to the algorithm. This is valuable if the algorithm to produce the output is non-obvious, e.g. an algorithm to emulate an example painting's style. The problem is first treated as a texture synthesis influenced sampling problem, where the idea is to form feature vectors representative of the data and then sample them coherently to synthesize a plausible output for the new image or video. It is then reduced to a labeling problem to perform example based processing in a more generalized manner than earlier techniques. This allows a different estimation of what the output should be by approximating the optimal (and possibly unknowable) solution through a different approach.
  

• Jing Hua, Stony Brook University
  Modeling, Manipulating, and Visualizing Continuous Volumetric Data: A Novel Spline-Based Approach
  
  In this talk, I will articulate a novel spline-based approach for modeling, manipulating, and visualizing continuous volumetric data. We have systematically developed a new interactive modeling and simulation framework for volumetric data and/or solid geometry with heterogeneous attributes. Our system is founded upon spline-based volumetric implicit functions and powerful physics-based modeling. This new paradigm affords representing, direct modeling, and rendering of multi-dimensional, physical attributes across any volumetric objects. In particular, we employ the trivariate B-splines to model volumes of structured grids, and we use the trivariate simplex splines to represent a tetrahedral decomposition of any 3D domain with complicated geometry and arbitrary topology. The multiresolution capability is achieved by interactively subdividing any region of interest and allocating more knots and control coefficients accordingly. I will show that our feature-sensitive fitting algorithms can reconstruct a more compact, continuous representation for real data. In addition, our spline-based models are governed by the principles of dynamics, hence responding to sculpting forces in a natural and predictable manner. We augment our framework with a 3D haptic interface, further facilitating the realistic manipulation of any modeled dataset. Within our framework, users can easily manipulate the geometry and topology of modeled objects, as well as their physical properties through an intuitive, haptic means with force feedback. If time permits, I will demonstrate our spline-based approach also leads to a new Scalar-Field Deformation (SFD) technique for FFD. Throughout the talk, I will highlight diverse applications including solid modeling, shape design, haptic rendering, material editing and reconstruction, volume simplification, and medical data exploration and visualization.
  

• Aravind Kalaiah, University of Maryland, College Park
  Context-Aware Samples for Efficient Transmission and Rendering
  
  The rapid growth in the complexity of geometry models is leading a sustained revision of several techniques of computer graphics. At the heart of this trend is the representation of geometry by independent samples. However, raw sample primitives lead to data representation by locally constant approximations. This generally leads to a higher sampling rate and thus a high cost of representation, transmission, and rendering. We advocate an alternate approach involving context-aware samples that capture the local variation of the geometry. We detail two approaches; one, based on differential geometry and the other based on statistics. Our differential-geometry-based approach captures the context of the local geometry using an estimation of the local Taylors series expansion. We render such samples using programmable Graphics Processing Unit (GPU) by fast approximation of the geometry at the screen space. The benefits of this representation can also be seen in other applications such as simulation of light transport. In our statistics-based approach we capture the context of the local geometry using Principal Component Analysis (PCA). This allows us to achieve hierarchical detail by modeling the geometry in a non-deterministic fashion as a hierarchical probability distribution. We approximate the geometry and its attributes using quasi-random sampling. Our results show significant speedup in comparison to current approaches.
  

• Misha Kazhdan, Princeton University
  Shape Matching and Model Alignment
  
  In this talk I will discuss the challenge of efficiently retrieving shapes from a 3D model database. I will focus on the role that alignment plays in the retrieval and discuss methods, both practical and impractical, for addressing this challenge. In particular, I will talk about how spherical harmonics can be used to obtain an efficient lower bound for the similarity of two optimally aligned models.
  

• Aaron Lefohn, University of California, Davis
  A Streaming Narrow-Band Algorithm: Interactive Computation and Visualization of Level-Set Surfaces
  
  Deformable isosurfaces, implemented with level-set methods, have demonstrated a great potential in visualization and computer graphics for applications such as segmentation, surface processing, and physically-based modeling. Their usefulness has been limited, however, by their high computational cost and reliance on significant parameter tuning. This paper presents a solution to these challenges by describing graphics processor (GPU) based algorithms for solving and visualizing level-set solutions at interactive rates. The proposed solution is based on a new, streaming implementation of the narrow-band algorithm. The new algorithm packs the level-set isosurface data into 2D texture memory via a multi-dimensional virtual memory system. As the level-set moves, this texture-based representation is dynamically updated via a novel GPU-to-CPU message passing scheme. By integrating the level-set solver with a real-time volume renderer, a user can visualize and intuitively steer the level-set surface as it evolves. We demonstrate the capabilities of this technology for interactive volume segmentation and visualization.
  

• Wei Li, Stony Brook University
  Accelerating Simulation and Visualization on Graphics Hardware
  
  Simulation and visualization typically involve complex computations and large amount of data, hence consume huge computation power and memory bandwidth. Fortunately, most of these applications exhibit high parallelism, hence are suitable for acceleration on modern graphics hardware, that has evolved into a powerful and highly programmable graphics processing unit (GPU), composed of multi-pipelined SIMD processors and high memory bandwidth. In addition, if the simulation results reside directly within the graphics memory, they are easily visualized and steered without incurring expensive transfer overhead.   Since GPU was originally designed for graphics applications, there are certain obstacles to overcome before it can be easily and efficiently utilized for non-graphics purposes. The first part of this proposal presents a framework, namely the Programmable Lattice (PL), that encapsulates multiple physical models for simulation. Specifically, we present our GPU-based Lattice Boltzmann computation as an example, as well as the applications developed on top of it.   To visualize the simulation results, which are usually represented as volumetric datasets, texture-based volume rendering is employed. Traditional texture-based volume rendering processes all the voxels, which is unnecessary, since a large amount of voxels are invisible, either due to the transfer function or occlusion by other voxels. The second part of the proposal is focused on accelerating the visualization on GPUs by skipping the invisible voxels, without any loss of image quality.
  

• Aidong Lu, Purdue University
  Illustrative Interactive Stipple Rendering
  
  Simulating hand-drawn illustration can succinctly express information in a manner that is communicative and informative. We present a framework for an interactive direct stipple rendering of volume and surface-based objects. By combining the principles of artistic and scientific illustration, we explore several feature enhancement techniques to create effective, interactive visualizations of scientific and medical datasets. We also introduce a rendering mechanism that generates appropriate point lists at all resolutions during an automatic preprocess and modifies rendering styles through different combinations of these feature enhancements. The new system is an effective way to interactively preview large, complex volume and surface data sets in a concise, meaningful, and illustrative manner.
  

• Alejandro Troccoli, Columbia University
  Texture Registration and Processing for Photo Realistic 3D Modeling
  
  Recent developments of 3D scanning technology have opened new areas of active research, with numerous applications in digital archiving of historic monuments or archeological sites. However, the whole process is not yet fully automated, sometimes requiring intense user intervention. At the Robotics Lab at Columbia University we are creating tools for automating the acquisition and 3D modeling of two different domains: urban environments and archeological sites. I will present my contributions to this work in the area of texture acquisition and processing for the creation of photorealistic models, a task even more challenging for outdoor scenes due to the changing conditions of the environment. I will talk about the image-geometry registration process discussing two different scenarios we have addressed: first, fixing a camera with respect to the range sesor and second, allowing a camera to move freely. I will also present my work in progress on color correction, a step for fixing the color discontinuity between adjacent texture patches that were captured by different images. Finally, I will show the results of modeling Mt. Polizzo, a 6th century BC site in Sicily.