Plenary Sessions
Virtual Reality Center as a Campus-wide Research Resource
Wednesday, September 21 – 9:15am-10:10am
Carolina Cruz-Neira. Ph.D.
Virtual reality Applications Center
Iowa State University
The Virtual Reality Applications Center (VRAC) was created in 2000 as an evolution from the Iowa Center for Emerging Manufacturing Technologies to centralize research activities in the area of virtual reality and interactive graphics. VRAC started as a small group of five faculty and twelve students; today it houses forty-five faculty from all colleges on campus and over 200 students. Its current funding is from federal and private sources in the form of grants, contracts, and donations. This talk will discuss the successes and struggles to create, grow and sustain VRAC within the university environment as well as the challenges faced along the way.
Top of page
The OptIPuter
Wednesday, September 21, 1:10-2:05pm
Jason Leigh, Ph.D.
Director, Electronic Visualization Laboratory
Associate Professor, Computer Science
University of Illinois at Chicago
The OptIPuter is a project examines a new model of computing in which ultra high-speed networks form the backplane of a planetary-scale computer. OptIPuter research focuses on developing technology to enable the real-time collaboration and visualization of very large time-varying data sets for the earth sciences and the biosciences. This presentation describes the synergistic international network, middleware, visualization and collaboration research that is needed to provide an end-to-end solution for research scientists.
Top of page
Imaging and Visualization: Cell to Organism
Thursday, September 22, 8:40-9:35am
Gregory M. Jones, Ph.D.
Associate Director
Scientific Computing and Imaging Institute
University of Utah
State Science Advisor
State of Utah
The next decade will see an explosion in the use and the scope of medical imaging and the fuel for this fire will be the combination of our increasing abilities in the study of genetics, new imaging techniques, and advances in biomedical computing and visualization. From Leonardo Da Vinci's anatomical drawings, to Wilhelm Roentgen's first X-ray of the human hand, to today's use of computer graphics and virtual reality to "fly through" three-dimensional reconstructions of magnetic resonance imaging (MRI) data, researchers have for centuries used visualization in their quest to understand human physiology. Each new visualization technique has brought researchers closer to capturing the complexity and beauty of human anatomy and physiology. Though the complexity of the human body still outstrips the capabilities of even the largest computational systems, and will do so for some time to come, computer scientists working with physicians and imaging researchers will continue to produce higher resolution and more informative interactive visualizations of physiologic systems. The advances in genetic research, computation sophistication, hardware performance, and imaging techniques will continue to feed one another bringing the overall field closer to the end goals of improving diagnosis and treatment while effectively decreasing medical costs.
Top of page
The NSF's Evolving Cyberinfrastructure Program
Thursday, September 22, 12:30-1:30pm
Guy Almes, Ph.D.
Program Director, Directorate for Computer and Information Science and Engineering
Shared Cyberinfrastructure Division
National Science Foundation
With origins in the NSF Supercomputer Center program and NSFnet program of the late 1980s, NSF is now enabling science and engineering users through its evolving Cyberinfrastructure program. Current resources supported by that program, including the TeraGrid, and representative science enabled through it will be described. Recent organizational and programmatic changes in NSF's Cyberinfrastructure program and the significance of performance and security will be discussed.
Top of page
Parallel Session 1 – Wednesday, September 21 – 10:40am-11:55am
Engineering Applications
3-D Stereo of Fluid Dynamics and Heat Transfer in a Blast Furnace Hearth
Chenn Zhou
Department of Mechanical Engineering
Purdue University Calumet
qzhou@calumet.purdue.edu
Dave Roldan
Department of Mechanical Engineering
Purdue University Calumet
Predrag Milovac
Department of Mechanical Engineering
Purdue University Calumet
A blast furnace which converts iron oxide represents the predominant iron-producing process in the U.S. Maximizing the campaign life of blast furnaces is critical to the economic vitality of an integrated steel mill. The campaign life of an iron blast furnace depends on hearth wear. Conditions inside the furnace are generally inferred from the response of thermocouples / heat flux sensors embedded in the refractory. Since measurements inside an operating hearth are not possible, computational fluid dynamics (CFD) has become a useful too to elucidate the internal condition. In this work, a 3-D comprehensive computational fluid dynamics (CFD) model has been developed for simulating the actual blast furnace hearth in which both hot metal flow and the conjugate heat transfer thorough the refractories have been included. The model has been validated using measurement data from Ispat Inland No. 7 blast furnace with good agreement between measured and predicted refractory temperature profiles. This CFD model is being utilized to 1) visualize the flow, temperature and erosion patterns; 2) understand the impact of internal conditions of the hearth on wear patterns and on the true nature of on-line measurements; and 3) design monitoring and controlling strategies for prolonging the campaign life.
Top of page
Parallel Sessions
Advanced Virtual Manufacturing Laboratory (AVML) for Training, Education, and Research
Hazim El-Mounayri
Mechanical Engineering
Indiana University-Purdue University Indianapolis
helmouna@iupui.edu
Tamer Wasfy
Advanced Science and Automation Corporation
tamer@ascience.com
Daniel Aw
Mechanical Engineering
Indiana University-Purdue University Indianapolis
daw@iupui.edu
Ayman Wafsy
Advanced Science and Automation Corporation
ayman@ascience.com
A high fidelity web-based virtual environment – Advanced Virtual Manufacturing Laboratory (AVML) – that uses advanced visualization techniques is presented. The dynamic virtual reality system is interactive and allows for the simulation of advanced manufacturing. Currently, the system supports CNC (Computer Numerically Controlled) milling, a versatile machining process that is widely used in industry.
The system provides a safe, cost-effective, and highly flexible and accessible tool for training and education as well as product realization. More specifically, the AVML provides the following training functions for users via the Internet: (a) interaction with a fully-functional virtual CNC milling machine, including real-time machining of parts, (b) training on key operating procedures of modern CNC machines, and (c) intelligent virtual tutor that gives a lecture describing the concepts of CNC milling and the components of the milling machine, as well as assistance with hands on training on the virtual machine.
The environment is driven by three software modules that communicate with each other using a TCP/IP network socket interface; these modules are: (1) a CNC Milling machine simulator, (2) a virtual-environment display engine, and (3) an intelligent-agent engine. The modules run on a single computer in a seamless web-based framework. The AVML can run on desktop or laptop personal computers. It can also run on more sophisticated systems such as a CAVE (for fully-immersive virtual reality visualization and simulation).
The AVML enables colleges and universities to easily and inexpensively provide students with effective, safe, and highly accessible training on advanced machine tools. It can also be used by machine tool manufacturers to provide online training, reducing or eliminating the need for onsite and/or live training classes for customers.
The work, which is still underway, has been presented in several conferences and symposia, including ASME, ASEE, ICCIE, and Solutions Conference 2005.
The methodology and its implementation, and the resulting virtual system for CNC machining will be presented. On the other hand, the current system�s capabilities and functionality will be demonstrated in the CAVE, which is currently available in the Advanced Visualization Laboratory at IUPUI.
Top of page
Parallel Sessions
Using Information Visualization as a Vital Component of an Assembly Process Simulation
Craig Miller
Computer Graphics Technology
Purdue University
miller02@purdue.edu
Isaac Chang
Computer Graphics Technology
Purdue University
yihsiang@purdue.edu
This paper discusses a course redesigned to meet the industry's demand of Product Lifecycle Management (PLM) literate workforce. The objective of this interdisciplinary course is to introduce students to manufacturing engineering theories coupled with an industry-sponsored project. One of the main requirements of this course is the building of an assembly line simulation. One of the main outcomes of the assembly line simulation is its ability to allow for design visualization and verification. This project exposes students to topics including design visualization, process design, process verification, and workspace ergonomics. Moreover, practices of project management and critical chain are built into this industry project. The end goal is to prepare students not only with the knowledge of PLM but also the capability of problem solving, communication, self-motivated teamwork, and leadership.
Top of page
Parallel Sessions
Parallel Session 1 – Wednesday, September 21 – 10:40am-11:55am
Scientific Visualization I
Multi-Scale Planetary Dynamics Aided by Stereo Visualization
Scott King
Earth and Atmospheric Sciences
Purdue University
sking@purdue.edu
Understanding the origin and evolution of planetary bodies requires integrating numerical modeling with observations and experiments from diverse fields. The calculations of creeping flow in planetary interiors are becoming increasingly complex because we find it necessary to include physical processes and interactions between processes that occur on multiple length and time scales. While the calculations computationally challenging, presenting the results in a coherent fashion has become an even greater challenge. We have found enormous benefit from using passive stereo animations. For collaborative work, these animations are also an ideal form of data compression. This talk will illustrate some of the discoveries made possible with stereo animation.
For more information, see: http://web.ics.purdue.edu/~sking/ED31E-04.htm
Top of page
Parallel Sessions
Interactive Visual Analytics of Multi-million Atom Nemo-3D Quantum Dot Simulations
Wei Qiao
Computer Science
Purdue University
qiaow@purdue.edu
Shaikh Ahmed
Electrical and Computer Engineering
Purdue University
ssahmed@purdue.edu
Gerhard Klimeck
Electrical and Computer Engineering
Purdue University
gecko@purdue.edu
David Ebert
Electrical and Computer Engineering
Purdue University
ebertd@purdue.edu
Marek Korkusinski
Electrical and Computer Engineering
Purdue University
marekk@purdue.edu
In this work we present a graphics hardware-accelerated direct volume rendering system for visualizing multivariate wave functions in semiconducting quantum dot (QD) simulations. The simulation data contains wave functions of multiple electron orbitals for up to tens of millions of atoms, computed by the NEMO3-D quantum device simulator software run on large-scale cluster architectures. These atoms form two interpenetrating crystalline Face Centered Cubic lattices (FCC). We have developed compact representation techniques for the FCC lattice within PC graphics hardware texture memory, hardware-accelerated linear and cubic reconstruction schemes, and new multi-field rendering techniques utilizing logarithmic scale transfer functions. Our system also enables the user to drill down through the simulation data and execute statistical queries using general-purpose computing on the GPU (Graphics Processing Unit).
Top of page
Parallel Sessions
Real-time Visualization of Large Particle Simulation
Michael Lasinski
E-Enterprise Center
Purdue University
lasinski@purdue.edu
Joseph F. Pekny
Chemical Engineering
Purdue University
pekny@purdue.edu
Jennifer S. Curtis
Chemical Engineering
Purdue University
jlds@ecn.purdue.edu
Raj Arangarasan
Envision Center for Data Perceptualization
Purdue University
arangarasan@purdue.edu
In this research presentation we discuss the simulations and visualization systems that are used to study rapid granular flows of particles that are tens of microns and larger. Specifically results from these kinds of simulations are used to investigate particle-phase phenomena and to obtain parameters for large-scale continuum models of industrial processes such as risers and dilute phase pneumatic conveying lines. Simulations of these kinds usually contain several hundreds of thousands of particles to millions of particles. Visualization system allows to visualize complex features such as sequence of 3D couette shear flow system with thousands to millions of hard sphere particles with a particular solids volume fraction and coefficient of restitution. The shear is maintained with Lees-Edwards Periodic Boundary Conditions and is in the absence of gravity. It is also assumed that fluid effects are negligible. Due to large number of particles and the complexity of its movements, we used several display systems such as ultra-high resolution IBM T221 LCD screen, and tiled wall display. The 22� LCD screen allows visualizing complex detailed rendering, such as this particle simulation, in a resolution of 3840x2400 pixels (9.2 million pixels) at 200 pixels per inch density. Still for very large particle simulation single workstation and single display system is not sufficient either for the computation or for the display quality. So other possibilities such as running this simulation in a cluster based high resolution 3D stereoscopic tiled display wall is used.
Top of page
Parallel Sessions
Parallel Session 1 – Wednesday, September 21 – 10:40am-11:55am
Advanced Interfaces and Applications I
Rapid Building of Cheap and Scalable Passive, Stereoscopic, Tiled Display Wall Using Commodity Products
Raj Arangarasan
Envision Center for Data Perceptualization
Purdue University
arangarasan@purdue.edu
James Bartek
Envision Center for Data Perceptualization
Purdue University
bartek@seyet.com
In the last decade computation power has increased considerably and as a result so has the complexity of the scientific problems. Often visualization of these scientific data sets allows the researchers to understand the problem better. As the complexity of the problems increase the need for visualizing the large data sets increases considerably. These kind of scientific visualization often requires high resolution display to visualize large complex models. One way to achieve high-resolution display is through a tiled display wall. Tiled display wall contains several components and technical factors to build in-house with much difficulty. Commercially available tiled display walls are often very expensive since these kinds of displays are custom built. In this paper we discuss our research experience in designing and building an in-house, one of the largest passive stereoscopic display wall in the country (USA) in the academic environment, a 3x4 passive stereoscopic tiled display wall using commodity components. We discuss multiple components, different parameters and issues, step-by-step instructions to rapidly design, build and align a passive stereoscopic tiled display wall using standard, mid-range digital light processing (DLP) projectors and cheap off-the-shelf components. The goal of this paper is to provide a practical and affordable approach to rapidly build a scalable passive stereoscopic tiled display wall.
Top of page
Parallel Sessions
The John-e-Box
John N. Huffman
Advanced Visualization Lab
University Information Technology Services
Indiana University
jnhuffma@indiana.edu
Michael Boyles
Advanced Visualization Lab
University Information Technology Services
Indiana University
mjboyles@Iupui.edu
Recent advances in commodity graphics and projection hardware have motivated many notable research projects and community discussions about the potential of these technologies to make advanced visualization more broadly accessible. However, the actual realization of this promise on a significant scale is challenging, requiring strong institutional commitment, expert technical support, and a broader visualization context. This paper describes an ongoing effort at Indiana University (IU) to develop a commodity-based, large-format, 3D stereo display system and to deploy a collection of such systems to a range of classrooms, laboratories, galleries, and learning environments throughout the IU system and the State of Indiana. To date, these systems have been used in over 30 projects by investigators in 15 departments across four different IU campuses. In addition, this technology has been used to reach well over 3,000 individuals through a series of coordinated outreach efforts. This initiative is also notable for fostering new interpersonal collaborations and inter-departmental cooperation, for enabling non-traditional applications in education and artistic expression, and for providing an interface to other advanced information technology efforts.
For more information, see: http://www.avl.iu.edu/technology/affordable/
Top of page
Parallel Sessions
Location-Aware Multimedia Delivery in an Art Museum
Mary Pietrowicz
Visualization and Interactive Spaces Lab
Pervasive Technology Labs at Indiana University
maryp@iupui.edu
Polly Baker
Visualization and Interactive Spaces Lab
Pervasive Technology Labs at Indiana University
baker@iu.edu
A key theme in today's information age is the ability to get information where and when it is most relevant. Translated into an art museum, this suggests the need to deliver information to visitors as they wander through the galleries. Audio tours have been available in art museums for several years, and commonly include narration about the artist and the art, as well as quotes, music, or dialogue to augment the printed information available in the gallery and enhance the visitors� experience of the art work.
Today's technologies make it possible to deliver full multimedia tours, including visualizations, to the museum visitor. Handheld PDAs have advanced in power to the point where they are capable of displaying imagery, animations, and narration. Wireless networking makes it possible to stream content to the handheld in an on-demand fashion. Finally, location tracking strategies are emerging where it is possible to use existing wireless networking infrastructure to determine a visitor�s location within the museum. Once we know a visitor's location, we can stream the appropriate visualization to their handheld for playback.
This talk will describe our efforts to determine a visitor's location using 802.11 wireless networking, in a project carried out with the Indianapolis Museum of Art. We describe the use and performance of the Ekahau location identification software, and our enhancements to tune and improve accuracy in the reported position. We also describe an alternative model for positioning, currently in experimental use within the ICTC building.
Top of page
Parallel Sessions
Parallel Session 2 – Wednesday, September 21 – 2:35pm-3:50pm
Telecollaboration I
Effective Distance Learning Through Sustained Interactivity and Visual Realism
Voicu Popescu
Computer Science
Purdue University
popescu@cs.purdue.edu
We describe an ongoing project whose goal is to develop a distance learning system that extends the real classroom to seemlessly accommodate remotely located students. This way distance learning becomes an integral but unobtrusive part of proven conventional on-campus learning. The system will convey to both remote and local participants a strong sense that the remote participants are actually present in the classroom. The sense of presence will be enabled by:
Interactivity. The system will support high-bandwidth, low-latency, uninterrupted, and secure communication between the participants.
High-quality graphics. The system will provide realistic visual descriptions of the classroom, instructor, students, and of the lecture materials by combining segmented high-resolution real-time video with pre-downloaded photorealistic 3D models and with presentation materials.
Top of page
Parallel Sessions
AGJuggler
Dioselin Gonzalez
Envision Center for Data Perceptualization
Purdue University
dioselin@purdue.edu
Laura Arns
Envision Center for Data Perceptualization
Purdue University
arns@purdue.edu
AGJuggler is a project developed at the Envision Center for Data Perceptualization at Purdue University. The main objective in this study is to make collaborative experiences and interactions richer with the incorporation of Virtual Reality tools inside a collaborative framework. This is accomplished by offering: an improved sense of presence; synchronous distributed visualization; real time interaction and a customizable solution for implementing virtual reality within a collaboration environment.
The Access Grid™ is the collaboration environment used for the study; while the API used for programming virtual reality applications is VRJuggler.
For more information, see: http://people.envision.purdue.edu/~dioselin/AGJuggler/
Top of page
Parallel Sessions
Enabling Collaborative Science with Cooperative Computing
Douglas Thain
Computer Science and Engineering
University of Notre Dame
dthain@cse.nd.edu
Despite the advent of cheap, plentiful computing power, even the most skilled of users struggle against a rising tide of complexity. Scientists in a variety of fields depend upon large scale computing to produce and share new results, yet are hamstrung by the limited capabilities for sharing and the immense number of failure modes in today's computing systems. To attack this problem, we must attack several fundamental problems in the design of operating systems, distributed systems, and programming languages. Our goal is to make it easy for non-specialists to efficiently share computing resources and data products across the wide area, while maintaining security and control of their own resources. We call this model "cooperative computing" and are developing a wide array of tools to assist the practicing scientist in his/her work. In this talk, we will present a variety of novel computing tools and describe how they have been applied to problems in bioinformatics, molecular dynamics, and high-energy physics.
For more information, see: http://www.cse.nd.edu/~ccl/
Top of page
Parallel Sessions
Parallel Session 2 – Wednesday, September 21 – 2:35pm-3:50pm
Information Visualization I
Bringing the Lab into the Market and the Market into the Lab: Using Technology to Measure and Manage the Customer Experience
Raymond Burke
Kelley School of Business
Indiana University
rayburke@indiana.edu
In recent years, there have been several innovations in measuring how customers shop retail stores and respond to changes at the point of purchase. These include using virtual reality simulations to recreate the store environment in the laboratory and measure consumers� decision processes, and using video cameras and other location sensing devices to track customer shopping patterns in the store. When these methods are combined with conventional research techniques, such as exit interviews and the analysis of shopping basket data, they provide a rich understanding of how consumer goals interact with the shopping environment to determine behavior.
Two studies will be presented demonstrating how these research techniques can be used to analyze the shopping process and measure the impact of changes in the store environment on consumer behavior. The first study used computer-aided video observation and transactional data to track how customers shopped in two consumer electronics stores over the entire 2003 holiday season. The research revealed that purchase conversion rates on "Black Friday" (the day after Thanksgiving) were significantly lower than other days of the holiday season due to long lines, crowding, and out-of-stock conditions. The second study used virtual reality simulations to test the impact of making changes to a store's layout and signage on product sales and price sensitivity. The study found that store design can have a substantial impact on shoppers� likelihood of penetrating the store and making a purchase, and that consumers are less price sensitive when the psychological (time/effort) costs of shopping are lower.
The presentation will conclude with a demonstration of how video-based customer tracking can be combined with virtual reality simulations to remotely monitor and manage the customer shopping experience.
Top of page
Parallel Sessions
Development and Understanding of Consumer Demand in Retailing with Visual Technology: A Demand-Side Research and Analysis
Sandra S. Liu
Consumer Sciences and Retailing
Purdue University
liuss@purdue.edu
Robert D. Melara
Psychological Sciences
Purdue University
melara@psych.purdue.edu
Raj Arangarasan
Envision Center for Data Perceptualization
Purdue University
arangarasan@purdue.edu
Retailing is an industry that works with lean profit margins and constant changes in consumer demand. Portfolio management of product categories is becoming critical for stores to ensure the profitability of their operations. Currently, one of the important issues that major retail chains and, on a more macro scale – shopping malls, encounter is how to generate and direct consumer traffic from one product category or a section of the mall to another in order to maximize the overall profit. The aim of this research is to understand, fundamentally and objectively, the cognitive and psychophysical responses of the individual consumer to stimuli such as store layout, signage, and color. To achieve these goals, we developed and exploited a new research tool – an immersive virtual environment for retailing and shopping. This research tool enables the systematic study of consumer responses, on an individual consumer basis, to the atmospheric factors of signage, color, and layout.
The novelty of the proposed project for demand-side analysis is an unprecedented ability to investigate real-time consumer behavior in a realistic retail setting while still exerting strict laboratory control, thereby permitting careful, objective psychophysical measurement. This project consists of the following segments: (1) A photorealistic virtual store through stereoscopic rendering technology; (2) a software system that is able to manipulate different atmospheric factors (signage, color and layout), while allowing real-time navigation through the environment; (3) psychophysical measures for monitoring consumers' response time and accuracy in selecting the cued product categories; and (4) two objective indices of consumer behavior, namely sensitivity and subjective bias in the selection of a product or a product category, based on the psychophysical responses of consumers to these atmospheric stimuli.
Top of page
Parallel Sessions
Analysis and Visualization of Technology Data in the US
Colin Murray
School of Library and Information Science
Indiana University Bloomington
colin.murray@nicta.com.au
Weimao Ke
School of Library and Information Science
Indiana University Bloomington
wke@indiana.edu
Hana Milanov
Kelley School of Business
Indiana University Bloomington
hmilanov@indiana.edu
Mark Meiss
Computer Science
Indiana University Bloomington
mmeiss@steinbeck.ucs.indiana.edu
Katy Börner
School of Library and Information Science
Indiana University Bloomington
katy@indiana.edu
Shravan Rajagopal
School of Library and Information Science
Indiana University Bloomington
shrajago@indiana.edu
We apply diverse data analysis and visualization techniques to a dataset of technology companies in the US from 1989 to 2003. Our goals are to understand (1) the birth, growth, and death of companies over the 15 year time span, (2) the geospatial distribution of different industries in US, (3) the correlation of the increase in number of employees and sales for companies that change their geographical locations at least six times, and (4) sudden increases in the number of employees and in sales for the different industries in the complete time span.
To understand the first goal we display nodes representing companies on a base map of the US. We generate a 15-frame animation, one frame for each year. Node coloring is used to identify new or dying companies. This provides a good overview of the dataset. For the second goal we generate similar maps for each industry with every year on the same map. This allows us to compare the size and geographical distribution of the different industries.
To address the third goal, for individual companies we plot the sales and employees on a line chart and add bars to indicate when moves occur. We also place a map of the company�s trajectory in the background. This visualization allows us to see the relationship between the movements of the company and its sales and number of employees.
For the final goal we perform a simple burst analysis calculating the number of companies in each industry that have a burst for sales and for employees. We show the bursts of each industry on line charts, one for sales and one for employees. From these charts we can see when the different industries experienced major growth.
While first insights could be gained, the dataset appears to be too limited to provide definite answers to questions that business experts would ask. The complementary web page with details on the presented analysis and supplementary material is available at:
http://iv.slis.indiana.edu/ref/iv05contest/.
This work was also submitted to the IEEE InfoVis 2005 Contest.
Top of page
Parallel Sessions
Parallel Session 2 – Wednesday, September 21 – 2:35pm-3:50pm
Systems Development
Developing Reliable Complex Software Systems in a Research Environment
Christopher Mueller
Open Systems Lab/Computer Science Department
Indiana University Bloomington
chemuell@cs.indiana.edu
Andrew Lumsdaine
Open Systems Lab/Computer Science Department
Indiana University Bloomington
lums@cs.indiana.edu
As computational science projects become more ambitious, many research projects rely on complex custom software to help process data and generate results. This software is often developed in an ad-hoc manner over the lifetime of the project by a combination of the PI, students, collaborators, and consultants. While effective at small scales, as the size and complexity of the software and data increases, this approach can lead to software that is difficult to use and maintain. Eventually, development and maintenance tasks can consume a disproportionate amount of time and resources, at which point many researchers become software developers first and scientists second. In this talk, we describe the software engineering challenges inherent in modern academic research labs and provide practical approaches to solving them. These techniques are based on our experience developing large software systems with the Biocomplexity Institute (CompuCell3D), Randy Heiland's Scientific Data Analysis Lab (data visualization), Katy Börner's InfoVis Lab (ICV, Network Bench), the Boost C++ project (Boost Graph Library), Open MPI, and other academic and industrial projects. We will present the software engineering and resource management strategies we applied to these projects that led to increased research productivity and a marked improvement in the quality of the software. Many of the core ideas have their foundation in industrial software engineering, but required adjustments to be effective in academic settings. Throughout the presentation, we will highlight these differences and demonstrate many unique aspects of academic software development. We will also discuss strategies for funding software development and provide details on the organizational structures we have found most effective. Ultimately, we will demonstrate how understanding the intricacies of managing academic software development can help research groups use custom software to enhance rather than hinder their programs.
Top of page
Parallel Sessions
High Performance Real-Time Interactive Visualization Technologies Using PC Clusters
Raj Arangarasan
Envision Center for Data Perceptualization
Purdue University
arangarasan@purdue.edu
Traditionally desktop-based graphics rendering has been done using a single personal computer (PC) on a single monitor. With the latest developments in graphics hardware and computing power, a single PC is used to drive multiple displays at the same time. But the complexity of scientific visualization applications increased tremendously so that a single PC could not render complex scenes in real-time even to single display, not to mention multiple displays. This single PC, single/multi display configurations have several limitations, such as insufficient computing power, trade-off between rendering quality vs. display area, especially to render complex scenes in real-time. In this paper, the author presents an approach that describes the use of commodity cluster to render complex scenes to a single monitor, and/or to different tiled display configurations in real-time. Our approach provides more flexibility and ability to customize the computing power, rendering quality and display area as required. Several hardware configurations, software applications, and specialized graphics libraries will be discussed that utilizes the latest hardware and software components. System bottlenecks due to hardware and software limitations will also be elaborated along with possible solutions and workarounds to achieve maximum performance. Based on this performance study, we will suggest multiple solutions and configurations to provide better insight to what hardware and software is vital and should be obtained as technology progresses in the future.
Top of page
Parallel Sessions
Developing and Deploying a Robust System for 3D Surface Scanning and Analysis
Jeff Rogers
Advanced Visualization Lab
University Information Technology Services
Indiana University
jlrogers@iupui.edu
Eric Wernert
Research and Academic Computing
University Information Technology Services
Indiana University
ewernert@indiana.edu
3D surface scanning is an evolving field that encompasses a wide variety of technologies and applications. For the past three years, the UITS Advanced Visualization Lab has been working with researchers in the IU School of Medicine, the Anthropology and Computer Science departments at IUPUI, and an international consortium of collaborators to develop an integrated system of hardware, software, and methodology to enable the reliable and accurate scanning of human heads and faces.
This NIH-funded collaboration is utilizing 3D surface scanning technology to supplement traditional physical measurements in the diagnosis and study of fetal alcohol spectrum disorders (FASD). Craniofacial anthropometry has been used to assess and describe abnormal craniofacial variation and the facial phenotype in many syndromes. It has also been used in the clinical setting as a diagnostic aid and as a means of objectifying clinical descriptions of individual patients. Through the use of 3D scanning, the project seeks to develop novel methods for assessing facial variations to provide more accurate diagnoses of the level of alcohol exposure. Moreover, by organizing the surface scans of subjects and control individuals in a database, researchers will be able to perform novel queries, comparisons, and longitudinal studies. This project has developed three portable 3D scanning systems that have been deployed to a variety of locations throughout the US as well as sites in Europe and Africa.
This talk will focus on our experiences in developing a robust and accurate 3D surface scanning system for the FASD project, including the evaluation of competing technologies, the development of software procedures for stitching, measurement, and analysis of models, and the challenges of international technology deployment. We will also highlight some of the outgrowth applications of this technology, including scanning of objects for forensics, cultural heritage, and digital art.
Top of page
Parallel Sessions
Parallel Session 3 – Wednesday, September 21 – 4:15pm-5:30pm
Educational Applications
Stereoscopic Visualization of Scientific and Medical Content
Albert William
School of Informatics
Indiana University-Purdue University Indianapolis
almwilli@iupui.edu
The use of stereoscopic display medium is now in use at the Ruth Lilly Health Education Center (RLHEC) in Indianapolis to help educate nearly 100,000 students annually. The stereoscopic display is helping to revolutionize a subject area to students K-12 in a subject matter that traditionally has been a very difficult area to learn. Content created by the research team at the IUPUI School of Informatics has been implemented to fit into the curriculum of the RLHEC. Future plans of the RLHEC include distributing this content to other health centers around the state, country, and globally, through high speed networks.
This presentation will display on a portable John-E-Box, or other portable stereo display system available to the School of Informatics, the display of a number of stereoscopic animations created for the RLHEC, including subjects such as The Cell, The Circulatory System, The Nervous System, The Immune System and The Respiratory System. Also displayed will be content created for the simulation of Hip and Knee Replacement Surgery, and simulations of real world data of biological macro-molecules.
Presentation will also include a description of content research procedures, pre-visualization methods, production methods, stereoscopic creation techniques, production issues and post-production methods. Considerations of the limitations and benefits of the stereoscopic medium will be discussed, as well as these considerations regarding the production of the stereoscopic content. A PowerPoint presentation can accompany the John E Box demonstration to further explain the production pipeline.
Top of page
Parallel Sessions
The GameSeum
Thom Gullespie
Telecommunications
Indiana University Bloomington
thom@indiana.edu
The Gameseum project is a proposal to use game engine technology to allow a Museum to exhibit itself and the engage patrons through game technologies and networks in ways not currently possible.
Game technologies have developed over the past 30 years to the point where extremely robust game technologies are available in a variety of low cost forms. World Forge is open source game technology which can be used to create massive 3D online role playing games on the web through campus and home personal computers connected to internet via DSL and cable modems. The Unreal Engine, which makes the extremely popular and profitable game Unreal, is available for free to develop non-commercial games of incredible complexity and beauty. The Torque game engine from GarageGames can be purchased for $100 with few restrictions on commercial use and development.
Game engine technology, while inexpensive, is not as ubiquitous and easy to use as Microsoft Word but the technology is moving in that direction. It is not far fetched to imagine a time within the next 5 years when students and community members will be as likely to have operating versions of game engines on their laptops, broadband accessible, as they are today to have word processing, graphic programs and interactive design tools such as Flash and Director from Macromedia.
How this would this work?
Most games begin with a game world, which can be explored. This space used to be 2 dimensional but today most games present the player with a 3 dimensional space where the game is played. The GameSeum project is a game project in the sense that there is a game world and there are rules for how the game is played just as basketball has a game world of the court and the rules for how the game is played fairly and how the game is won. The Museum will be the game master of this project controlling what the player has access to and what the rules of play are. This is not a proposal where anything will be destroyed or defaced in any fashion. This is not a proposal where players will be shooting anything. This is a proposal where the players come to the Museum to learn and appreciate art just as they do in the real world. That said, there would be players, which means there will be possibilities to learn in new ways.
Since this is a virtual space and not limited by physics it would not be necessary for any exhibit to actually come down. Depending upon the needs of the Museum staff exhibits could remain up indefinitely. Since it is a virtual space the Museum staff could exhibit material not normally exhibited in the museum due to size and fragility. Museum staff can also explore structural changes to the Museum before they are implemented. But, perhaps the must interesting possibilities are present from the patron�s point of view. Since it is a virtual space patrons could customize their person views of the Museum, as they need. Patrons could create their own personal commentary on art works both texturally or with audio or video. Patrons could also collaborate with the Museum staff.
At this stage we are imagining the possibilities but only with the GameSeum project as a working game simulation of the Museum can we realize possibilities we can�t even imagine.
This presentation will be a visual essay of the possibilities the GameSeum would offer for entertainment, education, research and funding.
Top of page
Parallel Sessions
Mathematical Models and Graphics Using Squeak
Randy Heiland
Scientific Data Analysis Lab
Pervasive Technology Labs at Indiana University
heiland@indiana.edu
David Milsho
Scientific Data Analysis Lab
Pervasive Technology Labs at Indiana University
dmilsho@yahoo.com
Charles Moad
Scientific Data Analysis Lab
Pervasive Technology Labs at Indiana University
cmoad@indiana.edu
We present an open source educational software environment called Squeak (squeakland.org) and describe a Squeak project we developed that is intended to enhance mathematics appreciation and learning. Squeak can be described as a programming environment for young people (K-8) or, for that matter, for a person of any age who may feel intimidated by computers and computer programming languages. Dr. Alan Kay, a renowned computer scientist and educator, is the driving force behind Squeak. Using a simple, but powerful, graphical user interface, a child begins by painting a two-dimensional object and then creates a script (an object-oriented program) that affects the behavior of that object. Creating a script is easily performed by dragging & dropping tiles that change the attributes or actions of the object. By creating multiple objects and scripts, a child can quickly create a complex, dynamic artificial world. We have constructed a Squeak project which introduces children to the concept of mathematical modeling – using the language of mathematics to describe (model) various phenomena in the world around us. We initially focus on modeling things found in nature (e.g., mountains, trees, plants, and clouds.)
Top of page
Parallel Sessions
Parallel Session 3 – Wednesday, September 21 – 4:15pm-5:30pm
Scientific Visualization II
Visualization for Discovery of Knowledge from Scientific Data
James Caruthers
Chemical Engineering
Purdue University
caruther@ecn.purdue.edu
Leif Delgass
Computer Graphics Technology
Purdue University
ldelgass@purdue.edu
Michael McLennan
Information Technology at Purdue
Purdue University
mclenna@purdue.edu
Laura Arns
Envision Center for Data Perceptualization
Purdue University
larns@purdue.edu
The discovery of knowledge from scientific data is quite challenging because of the vast amounts of typically disparate data that are needed for insight. Because the visual channel of humans has a very high bandwidth, our approach is to present the data in an advanced visualization environment utilizing a high-resolution display with 3-dimensional (3D) interaction. However, the information needs to be presented to the domain expert in a form that is natural for the scientific domain expert, without burdening the expert with issues concerning data representation and management.
We will report on the beginning development of a computer-based system for scientific discovery in the chemical sciences that enables the scientist/engineer to effectively manage and visualize widely disparate types of data. The types of data that are represented include spreadsheets, 2D and 3D graphs, chemical structures, molecular orbitals, etc., where the different representations of data can be changed at will. For example, a chemist that is developing a new catalyst for producing a cleaner diesel fuel or a new polymer or any of a number of other applications naturally thinks in terms of 2D chemical structures, 3D structures of molecules and molecular orbitals; thus, instead of examining a traditional 3D scatter plot of some catalyst performance data versus some descriptors of the catalyst structure that uses colored symbols, the chemist can select a point on the graph which will then display the chemical structure. If the expert desires additional information concerning that point, the expert can connect to the synthesis data (text description), experimental characterization data (2D plots) or molecular simulations (3D sterographic images). The details of these changes in data representation and fidelity of information are hidden from the domain expert so that they can focus their activities on scientific discovery not data management and/or representation.
The system that has been developed is scaleable so that it can be used on a dual monitor PC, a multi-head display or a large tiled wall � whatever device is most appropriate and available. Since the user interface is the same on this range of display devices, the scientific domain expert only has to learn how to use one system.
Top of page
Parallel Sessions
Zero-Crossing Iso-Surfaces in Volume Datasets
Shiaofen Fang
Computer and Information Science
Indiana University-Purdue University Indianapolis
sfang@cs.iupui.edu
Pooja Gupta
Computer and Information Science
Indiana University-Purdue University Indianapolis
pgupta@cs.iupui.edu
This paper describes a new surface extraction method for volume datasets by applying iso-surface extraction techniques to the zero-crossing edges in a volumetric domain. A volume dataset is first filtered using a Laplacian of Gaussian filter to generate a zero-crossing field. A marching cube process will then be able to extract the entire zero-crossing surface that may be viewed selectively based on various intensity ranges and gradient scales. This new technique provides a more efficient surface navigation and extraction mechanism, as well as more accurate surface details, than the traditional iso-surface techniques.
Top of page
Parallel Sessions
Use of Purdue's DRE for Creating Nanotechnology Visualizations
Carlos Morales
Envision Center for Data Perceptualization
Purdue University
morales@purdue.edu
This presentation focuses on the use of Purdue�s Distributed Rendering Environment (DRE) in the creation of animated visualizations for a nanotechnology exhibit. Purdue�s DRE allows users to distribute rendering jobs across campus and utilize idle cycles for creating 3D imagery. The net result is a dramatic reduction in the amount of time that it takes to render 3D visualizations. In this particular case, nine animated visualizations were created in less than three months. Without the DRE, there would not have been enough computational power available to the team to complete the task in the allotted time. Three-dimensional visualizations that portrayed the uses of nanotechnology in the following areas were created: agriculture, computation, sports, medical, manufacturing, and military.
Top of page
Parallel Sessions
Parallel Session 3 – Wednesday, September 21 – 4:15pm-5:30pm
Telecollaboration II
Instruments and Sensors on the Grid: The CIMA Middleware Project
Donald McMullen
Knowledge Acquisition and Projection Lab
Pervasive Technology Labs at Indiana University
mcmullen@indiana.edu
The integration of scientific instruments with computations and scientific data management is an important problem in creating usable cyberinfrastructure. Instruments are still largely disconnected from downstream analytical software and the scientific productivity of instruments can be greatly enhanced using existing Grid services for location, allocation, scheduling, and access control. This problem is becoming critical as three issues continue to grow in importance in research: 1) investments in geographically extended (e.g., international) collaborations organized around large shared instrument resources; 2) increasingly real-time use of instruments by remote researchers both for "first look" activities and pipelined data acquisition and reduction; and 3) sensor networks with hundreds to thousands of nodes being deployed.
The Common Instrument Middleware Architecture (CIMA) project aims at �Grid enabling� instruments as real-time data sources to improve accessibility of instruments and to facilitate their integration into the Grid. A primary challenge addressed by this research program is to develop a generalized approach to instrument middleware that allows existing and new instruments to be integrated into Grid computing environments.
CIMA middleware is based on current Grid implementation standards, and is accessible through platform independent standards such as Web Services. In keeping with the design goal of general applicability, CIMA interfaces are being developed for a variety of instrument and controller types from large shared facilities (synchrotron X-ray sources, robotic optical telescopes) to small embedded industrial controllers and sensor nets and micro-sensor packages such as MOTEs, developed for the DARPA SmartDust program.
Other issues being explored include extending the accessibility of instruments to new classes of users (e.g. student remote access to advanced materials science facilities at national labs), use of instruments by software agents, and increasing the longevity, flexibility and durability of software systems for instruments and sensors.
Within Indiana CIMA implementations are linking X-ray diffraction facilities at IU and Purdue over the iLight network. This provides a uniform infrastructure for researchers in the state to find a facility with capabilities that meet their needs and to interact with the technicians and instruments there remotely. This talk will include a demonstration of the remote access capabilities of a CIMA-enabled X-ray diffraction crystallography laboratory.
For more information see: http://www.instrument-middleware.org/
Top of page
Parallel Sessions
Virtual-Reality 3D Interactive Interface for the Teleoperation of Robotic Systems
Xiaoli (Lucy) Yang
Electrical and Computer Engineering
Purdue University Calumet
lucyyxl@hotmail.com
Virtual reality is a high-end human-computer interface allowing user interaction with simulated environments in real-time and through multiple sensorial channels. The application of virtual reality technologies in Internet robot teleoperation is an emerging research field.
The objective of this research is to develop a more interactive and immersive 3D virtual reality interface for Internet robot teleoperation in order to improve the stability and alleviate the time delay between the user and robot. Updating 3D graphical models of robot and working environment in real-time can provide the user with a more interactive and friendly interface to �see� and control the robot in the real world. Force, size and shape feedbacks will help the user get more immersed into the simulated virtual reality. A high-level supervisory script language will generate commands for different kinds of robots and be flexible for both robot programmers and regular users. A multi-user collaborative environment will be developed for complex experiments.
The implementation of the research can apply Internet robot teleoperation to some specific medical and biological experiments. While dealing with highly contagious biological viruses, some medical and biological experiments are hazardous for researchers. By applying this virtual reality teleoperation system to carry out these dangerous tasks, researchers can be protected to stay away from the harmful work. Some designed medical experiments will be conducted with this system and the performance of the system will be analyzed on the stability, the response time, discrepancy of virtual and real world to further improve the stability of the system.
Top of page
Parallel Sessions
Loose Minds in a Box at Purdue
Dioselin Gonzalez
Envision Center for Data Perceptualization
Purdue University
dioselin@purdue.edu
Timothy Rogers
Envision Center for Data Perceptualization
Purdue University
tjrogers@purdue.edu
Carol Cunningham
Visual and Performing Arts
Purdue University
ccunning@cla.purdue.edu
Laura Arns
Envision Center for Data Perceptualization
Purdue University
larns@purdue.edu
As explained by Performance Director Jimmy Miklavcic, InterPlay: Loose Minds in a Box (LMIB) is a multi-site distributed performance where six geographically remote sites utilize Access Grid� technology to integrate each local performance into a global work. The concept of LMIB �revolves around the idea of being constrained in a box. The box is a metaphor for the physical, social, political or psychological constraints that we and/or others place upon us. The box also represents a sense of place in the realm of the virtual as well as in our conscience.� (Miklavcic, 2005)
The Envision Center for Data Perceptualization at Purdue University participates in Loose Minds in a Box with a local piece that includes a motion capture (mocap) segment and a virtual reality (VR) segment. During the performance, a dancer moves between the mocap area and the VR area several times. The mocap segment allows the dancer to control a character composed of boxes; it also uses the Shared MIDI Service (Ayromlou, 2005) to remotely control sound being generated at Montana. The VR segment includes an interactive application running in an immersive Flex� virtual reality system where the dancer uses tracking devices to interact with the world; it also includes the same application running collaboratively by using AGJuggler (Gonzalez and Arns, 2005) to allow remote audiences to interact with each other.
Top of page
Parallel Sessions
Parallel Session 4 – Thursday, September 22 – 9:50am-11:05am
Biomedical Applications
Intelligent Biomedical Literature Mining and Knowledge Visualization
Mathew Palakal
Computer and Information Science
Indiana University-Purdue University Indianapolis
mpalakal@cs.iupui.edu
Snehasis Mukhopadhyay
Computer and Information Science
Indiana University-Purdue University Indianapolis
smukhopadhyay@cs.iupui.edu
Shielly Hartanto
Computer and Information Science
Indiana University-Purdue University Indianapolis
shartant@cs.iupui.edu
Jagannathan Lakshmipathy
Advanced Visualization Laboratory
University Information Technology Services
Indiana University
jlakshmi@iupui.edu
Biological literature databases continue to grow rapidly with vital information that is important for conducting sound biomedical research. As data and information space continue to grow exponentially, the need for rapidly surveying the published literature, synthesizing, and discovering the embedded "knowledge" is becoming critical to allow the researchers to conduct "informed" work, avoid repetition, and generate new hypotheses. Knowledge, in this case, is defined as one-to-many and many-to-many relationships among biological entities such as gene, protein, drug, disease, etc. The knowledge discovery process basically involves identification of biological object names, reference resolution, ontology and synonym discovery, and finally extracting object-object relationships. We have developed a system called BioMap, using the entire MEDLINE collection of (over 12 million) bibliographic citations and author abstracts from over 4600 biomedical journals worldwide and to develop an interactive knowledge network for users to access this secondary knowledge.
The BioMap "knowledgebase" may significantly enhance the ability of biological researchers with diverse objectives to efficiently utilize on-line resources, generate methods for analysis of biological data such as identifying biological pathways, and provide computerized support for disease target and new drug discovery. The extracted knowledge is further mined to discover directional associations, transitive associations, and hyper associations among biological entities. These relationships form large knowledge graphs and we have developed information visualization tools to display and to interact with such knowledge spaces.
Top of page
Parallel Sessions
Dynamic Volume Rendering and Life Cycle Planning for Intensity Modulated Radiation Therapy (IMRT) Treatment
Seza Orcun
E-Enterprise Center
Purdue University
sorcun@purdue.edu
Sun Geun Kim
Envision Center for Data Perceptualization
Purdue University
sungeun@purdue.edu
Raj Arangarasan
Envision Center for Data Perceptualization
Purdue University
arangarasan@purdue.edu
For some decades, radiation therapy has been proved successful in cancer treatment. The major task of radiation therapy is to impose a maximum dose of radiation to the tumor cells. The IMRT technology makes it possible to deliver radiation more precisely by dividing the accelerator head into smaller units called "beamlets" that can be manipulated independently. This treatment planning requires a time consuming iterative work between a physician and an IMRT technician. The state of the art current technique is to determine the IMRT treatment plan at the beginning and use it without changing it in the course of the treatment. However, the assumption of fixed 'target volume' throughout the IMRT treatment, is very limiting given that the tumor shrinks in response to the radiation therapy. In this research, we apply time-varying volume rendering technique to the IMRT treatment and develop a proof-of-concept prototype system that enables capturing and relating the time dependent changes of the irradiated volume throughout the course of treatment. This prototype system will enable the researchers to explore different what-if scenarios, such as determining the 'Target Treatment Volume' depending on the delivered dose of radiation.
Top of page
Parallel Sessions
Mining Digital Geometry Data for Medical Diagnosis
Jiandong Fang
Computer and Information Science
Indiana University-Purdue University Indianapolis
jfang@cs.iupui.edu
Shiaofen Fang
Computer and Information Science
Indiana University-Purdue University Indianapolis
sfang@cs.iupui.edu
Jeffrey Huang
Computer and Information Science
Indiana University-Purdue University Indianapolis
This paper describes a new approach of mining polygon mesh surface datasets of human faces for a medical diagnosis application. The goal is to explore the natural patterns and 3D facial features that can provide diagnostic information for Fetal Alcohol Syndrome (FAS) disorder. Our approach is based on a digital geometry analysis framework that applies data mining and pattern recognition techniques to digital geometry (polygon mesh) datasets from 3D laser scanners and other sources. Novel 3D geometric features are extracted and analyzed to determine the most discriminatory features that best represent FAS facial characteristics. As part of a collaboration with the NIH International Consortium for FASD, the techniques developed here are being applied and tested on real patient datasets collected by the NIH Consortium both within and outside the US.
Top of page
Parallel Sessions
Parallel Session 4 – Thursday, September 22 – 9:50am-11:05am
Information Visualization II
Mapping Science
Katy Börner
School of Library and Information Science
Indiana University Bloomington
katy@indiana.edu
Kevin W. Boyack
Sandia National Laboratories
kboyack@sandia.gov
About 40 years ago, Derek J. deSolla Price suggested studying science using the scientific methods of science. Today, research on 'Mapping Knowledge Domains' – see PNAS 101 (Suppl. 1) Apr 6, 2004 -- aims to develop methods to analyze, model, and visualize the structure and evolution of science.
When mapping sciences, knowledge and expertise is extracted from digitally available literature, patent, grant and other data as well as from information about the producers and consumers of those data sets. Advanced data analysis techniques in combination with spatial metaphors, geographic principles, and cartographic methods are applied to organize, visualize, and communicate the semantic relationships inherent in the data.
The ultimate goal of mapping sciences might be an interactive cartographic map of all of science, with continents representing the major research areas such as, e.g., biology or physics, dots denoting major authors, PIs, papers or news, dynamically evolving research frontiers, blinking 'hot' papers and topics, etc. This map could be used to teach and understand the evolving structure of all of science, to identify major experts, to find and read the most relevant papers and patents, to see the effects of resource allocation decisions, to study social networks, etc. Last but not least, it would provide a unique bird's eye view of major experts in specific areas and mankind's knowledge in general.
This talk will present recent results on mapping science on a small and large scale. It will discuss the cyberinfrastructure needed to store and process large scale scholarly datasets and to serve the resulting maps to diverse user groups. We will also introduce the 'Places and Spaces' science exhibit (http://vw.indiana.edu/places&spaces/) that was created to demonstrate the power of maps for the navigation and management of physical places and abstract semantic spaces.
Top of page
Parallel Sessions
Table-Based Navigation of the Indianapolis Museum of Art Collection
Rob Stein
Visualization and Interactive Spaces Lab
Pervasive Technology Labs at Indiana University
rjstein@iupui.edu
Polly Baker
Visualization and Interactive Spaces Lab
Pervasive Technology Labs at Indiana University
baker@iu.edu
Rhonda Winter
Directory of Technology
Indianapolis Museum of Art
RWinter@ima-art.org
This talk describes the etx, an installation created to allow visitors to the Indianapolis Museum of Art interact with the collection. The etx installation includes a table-top display where users interact with the application. Imagery is also projected to three walls in the room, creating a sense of immersion in the application. Visitors interact with the etx application by pointing to and selecting from digital images of works that float on a tabletop. Through interaction with the application, participants can request details about the works of art, such as time period or artist. Participants can also explore how one item from the collection relates to other items in the collection.
The etx is based on the PercepTable technology, developed at the Visualization and Interactive Spaces Lab, part of the Pervasive Technology Labs at Indiana University. At the PercepTable, users interact by gesturing with small tools that can be moved on the table. Camera-based tracking, using algorithms developed in the Lab, is used to track users� movement.
This talk will present an overview of the etx installation, including the history of the application, details of its implementation, and the collaboration involved to create the exhibit. We will also show what we have learned from studying patterns of user interaction with the etx installation.
Top of page
Parallel Sessions
Distribution Services for 2005 Indiana Orthophotography Project
Anna Radue
University Information Technology Services
Indiana University
aradue@indiana.edu
Nathan Eaton
University Information Technology Services
Indiana University
neaton@indiana.edu
Michael Halla
University Information Technology Services
Indiana University
A first-of-its-kind statewide orthophotography project is rapidly expanding Indiana�s geographic information systems (GIS) data resources and will support new ways to visualize Indiana. This landmark collaboration involving the Department of Homeland Security, the Geographic Information Council (IGIC), and Indiana�s ninety-two counties, is generating terabytes of seamless, accurate statewide aerial photography and elevation data. This project will support the GIS needs of non-profits, state agencies, university researchers, and the general citizens of Indiana in such diverse activities as watershed management, economic development, and homeland security.
Data Management Support (DMS), a unit within University Information Technology Services, will discuss the 2005 Orthophotography Project's data distribution needs and challenges of creating enterprise database services. DMS will present an overview of the Indiana Spatial Data Service (ISDS), an enterprise web and database service supported by Indiana University's high performance computing and networking infrastructure. DMS will also present information regarding the Indiana Spatial Data Portal (ISDP), a geographic archival service supported by Indiana University's massive data storage system.
For more information, see: http://gis.iu.edu for access to the ISDS and ISDP.
Top of page
Parallel Sessions
Parallel Session 4 – Thursday, September 22 – 9:50am-11:05am
Advanced Interfaces and Applications II
The Solar Journey Project
Andrew Hanson
Computer Science
Indiana University Bloomington
hanson@cs.indiana.edu
The Solar Journey Project is 6-year, multi-institution collaboration focused on the development and application of advanced visualization, modeling, virtual reality and analysis tools to virtual astronomy. While the specific scientific objective of this research is a better understanding of the local galactic neighborhood of our Sun, this collaboration has also resulted in the development of a range of new visualization technologies, including: multi-scale techniques for overcoming the precision limitations of graphics hardware; navigational techniques for effectively traversing many orders of spatial magnitude, methods for physically-based star rendering, and a software framework to seamlessly integrate realtime analysis with high-quality batch rendering. Specific productions highlighted in this talk include several high-quality animated shorts (including one shown in the SIGGRAPH 2000 Electronic Theater), an interactive CAVE application, and a 20-minute production for the multi-projector laser dome at the Beijing Planetarium.
For more information, see: http://www.cs.indiana.edu/~soljourn/
Top of page
Parallel Sessions
Motion Capture Visualization for Ergonomic Analysis and Safety Design
Vincent Duffy
Industrial Engineering
Purdue University
Motion capture data can be used as the input to computer-based ergonomic analysis tools. A key objective is to improve safety and reduce the need for prototyping. New design concepts such as the Virtual Build and justification of redesign in automotive assembly are facilitated by this motion capture capability. Once more is known about the transfer of information from the visualizations from virtual to real, the interaction with some new designs can be simulated, so as to reduce the need for prototyping. Until these digital human modeling simulation capabilities are more developed, motion capture visualization will be an important tool for digital human model development. In the future, motion capture can still be an effective tool in digital human model validation. Improved safety through motion capture and related virtual reality visualization has the potential for reducing costs of workplace injury and insurance in manufacturing, transportation, government and health care.
Top of page
Parallel Sessions
Using mixed physical/virtual interfaces to enhance social aspects of educational games
Erik Pukinskis
School of Informatics
Indiana University
epukinsk@indiana.edu
Yvonne Rogers
School of Informatics/School of Library and Information Science
Indiana University
yrogers@indiana.edu
Interactive tabletop displays, like MERL's DiamondTouch provide a large touch-sensitive surface that allows for simultaneous multi-user finger-based interaction. Each user's touch completes a low current circuit with the table and a discrete pad placed on the floor or on the user's chair. A fine grid of antennas embedded in the surface detect the precise position of the touch and can differentiate between each user, allowing up to four users to interact with the table simultaneously and independently. The table can detect detailed motion of the fingers and hands and can approximate pressure sensitivity using the strength of signal.
We have used the DiamondTouch table along with tangible interface elements – physical objects with embedded sensors – to support a novel form of collaborative learning where collocated groups of children interact with a virtual world through a mixed digital/physical interface. We hope to use this application to study how children learn together in this new environment and how their path to understanding differs from traditional desktop interfaces and non-colocated learning situations.
In order to focus the learning experience on group learning and sharing, we have designed a game which allows children to explore the "Tragedy of the Commons" – the situation that arises when a small number of greedy participants spoil a shared resource for themselves and for everyone else. The game allows players to place sheep out to graze in a shared pasture, competing to see who can harvest wool the fastest without destroying the pasture. This requires players to come to an agreement with other players about how to allocate resources, encouraging players to think about their relationships with each other, not just their performance in the virtual world.
Our implementation of this game allows players use the DiamondTouch table to manipulate the game world directly by touching and dragging sheep through the environment while using a tangible interface – the "Sheep Shaker" – to add their sheep to the pasture. The mixed physical/virtual interface has been designed to give players a greater sense of their personal role in the game and prevent them from dissociating from their virtual identity.
The next stage of our research plan is to explore how small groups of students use the prototype system to play the collaborative game and to investigate whether a shared tabletop surface like the DiamondTouch table and tangible technologies can support new kinds of collaborative learning.
Top of page
Parallel Sessions
Grid Systems & Applications – Thursday, September 22 – 1:50pm-4:55pm
Grid Systems & Applications I
ORNL and Teragrid: An Overview
John Cobb
Oak Ridge National Lab
The Oak Ridge National Laboratory (ORNL) has recently undergone a renaissance with new programs and facility construction connected with the Department of Energy's Office of Science National Leadership Computing Facility (NLCF) at the National Center for Computational Sciences and the Spallation Neutron Source (SNS.) Additionally an aggressive wide area dark fiber initiative called FutureNet has allowed the connection of these infrastructures to national high performance networks. One component of these new activities is ORNL's Neutron Science TeraGrid Gateway (NSTG) which is a project partner in the National Science Foundation's TeraGrid project. The SNS expects orders of magnitude larger data and analysis needs than current neutron sources. The NSTG is focused on providing advanced TeraGrid cyber-infrastructure to meet the new scale of needs of SNS users. The NSTG is physically co-located with the NLCF resources and utilizes FutureNet infrastructure to connect to the TeraGrid backplane. Its user facing interface is a Science Gateway portal providing data management, application services, and SNS user identity management.
Top of page
Parallel Sessions
Grid Technology for Real Time Streams
Geoffrey Fox
Community Grids Lab
Pervasive Technology Labs at Indiana University
gcf@indiana.edu
Marlon Pierce
Community Grids Lab
Pervasive Technology Labs at Indiana University
mpierce@cs.indiana.edu
Shrideep Pallickara
Community Grids Lab
Pervasive Technology Labs at Indiana University
spallick@indiana.edu
Wenjun Wu
Community Grids Laboratory
Pervasive Technology Labs at Indiana University
wewu@indiana.edu
A growing number of applications involve real-time streams of information that need to be transported in a dynamic high-performance reliable secure fashion. Examples include sensor nets ranging from science to the military, mobile devices on ad-hoc networks and collaboration with at the simplest a multitude of audio-video streams. Further it is attractive to view the sources, sinks and filters of these streams as Web or Grid services so we can realize the pervasive interoperability of a service oriented architecture. We have been developing this idea systematically using an open source base infrastructure NaradaBrokering that forms a distributed set of message brokers that implement a publish-subscribe software overlay network.
This environment supports multiple protocols (including UDP, TCP, parallel TCP) and provides reliable message delivery with a scalable architecture. The infrastructure can supply message oriented middleware support for Web services with support of WS-Eventing, WS-Notification, WS-Reliable Messaging and WS-Reliability. Conventional support of SOAP will be augmented by "fast XML/SOAP schemes" that will transport the SOAP infoset efficiently and will allow Web services to exchange streams
(sets of messages) with high performance. We describe four applications that illustrate the capabilities of this system; collaboration with audio video and shared display streams, integration of hand-held devices to a Grid, and the linkage of real-time GPS sensors to Geographical Information Systems implemented as Web services. The final example links annotations to video streams showing how composite streams can be supported for real-time annotation. Challenges here are efficient archiving and metadata systems allowing "instant" rewinding and replay of any stream. This latter capability is demonstrated for remote sports coaching.
For more information, see:
http://www.naradabrokering.org
http://www.globalmmcs.org
Top of page
Parallel Sessions
Building and Deploying Community Nanotechnology Software Tools on nanoHUB.org – Atomistic Simulations of Multimillion-Atom Quantum Dot Nanostructures
Shaikh Ahmed
Electrical and Computer Engineering
Purdue University
ssahmed@purdue.edu
Gerhard Klimeck
Electrical and Computer Engineering
Purdue University
gekco@purdue.edu
Marek Korkusinski, Purdue University
Haiying Xu, Purdue University
Seungwon Lee, Purdue University
Sebastien Goasguen, Purdue University
Faisal Saied, Purdue University
The Network for Computational Nanotechnology (NCN) is a multi-university, NSF-funded initiative with a mission to lead in research, education, and outreach to students and professionals while at the same time deploying a unique web-based infrastructure to serve the nation's National Nanotechnology Initiative. The primary NCN outreach vehicle is our web site http://nanoHUB.org, which currently provides interactive on-line simulation and educational resources such as tutorials, seminars, and on-line courses. In 2004, over 3,200 users used the educational and outreach resources. Over 1,000 users performed over 60,000 on-line simulations. The raw web page hits exceeded 3.7 million. Around 30 research codes are available ranging from toy-models to sophisticated simulation engines. The NCN provides the resource for models, simulation and computation without any software installation for users via web-delivery. All the NCN services are free of charge and reach a broad audience.
One facet of this infrastructure involves the development of new "community codes" that provide the nanoscience research community with new capabilities and that lay a foundation for a new generation of CAD tools. Each research team has a research mission to move its field ahead and an equally important mission to contribute resources to the NCN's infrastructure. One such community tool is NEMO-3D (dynamo.ecn.purdue.edu/~gekco) which has been publicly released and recently expanded in its capabilities through detailed numerical performance analysis.
The growth of self-assembled semiconductor quantum dots (QDs) is driven by strain, induced by the mismatch of the lattice constants of the QD material (in this work, InAs) and that of the barrier material (GaAs). The resulting long-range strain field strongly modifies the energy diagram of the system, and has to be accounted for in realistic simulations of QD electronic properties. The nanoelectronic modeling tool NEMO-3D is designed to provide quantitative estimates of QD-bound electron and hole states by treating the system on the atomistic level. It captures the strain by adjusting the positions of constituent atoms so as to minimize the total elastic energy computed in the VFF Keating model. The displacements of atoms from their bulk positions are incorporated into the 20-band nearest-neighbor sp3d5s* tight-binding Hamiltonian.
A single dome-shaped InAs QD, with height of 1.8 nm and diameter of 19.2 nm, positioned on a 0.6-nm-thick wetting layer (WL), embedded in a GaAs barrier material is considered. Since strain is a long-range phenomenon, it has to be computed in a domain much larger than the quantum dot itself, while the domain for the electronic calculation can be smaller, since the electronic wave function is confined.
Scaling to tens of millions of atoms: NEMO-3D uses the conjugate-gradient algorithm to find equilibrium positions of atoms, and the Lanczos algorithm to diagonalize the sparse Hamiltonian matrix. In order to extend the treatment to multi-million atom systems, both algorithms are parallelized using MPI. For example, on 16 64-bit Itanium2 CPUs with memory of 11.5 GB/CPU (NCSA Teragrid) it is possible to compute strain in 200-million-atom, and electronic structure of 35-million-atom devices. NEMO-3D exhibits a linear scaling, both of computational time and memory, as a function of the system size. This work presents a systematic study of the strain calculated within a domain consisting of up to 64 million atoms, followed by an electronic calculation on a subdomain containing up to 21 million atoms. Unique and targeted eigenstates can be extracted from system matrices of order 4Å~108.
Top of page
Parallel Sessions
Grid Systems & Applications II
Parallel Processing and Large Data Visualization in ParaView
Berk Geveci, Kitware, Inc.
ParaView is an open source scientific visualization tool specifically designed to handle large datasets in distributed computing environments. It is based on the popular open source library VTK (Visualization Toolkit). This presentation will describe the historical evolution of ParaView and introduce its architecture and capabilities.
Top of page
Parallel Sessions
Distributed and Parallel Scientific Visualization with ParaView on TeraGrid
Resat Umit Payli
Mechanical Engineering
Indiana University-Purdue University Indianapolis
rpayli@iupui.edu
Erdal Yilmaz
Mechanical Engineering
Indiana University-Purdue University Indianapolis
eyilmaz@iupui.edu
Hasan U. Akay
Mechanical Engineering
Indiana University-Purdue University Indianapolis
hakay@iupui.edu
Akin Ecer
Mechanical Engineering
Indiana University-Purdue University Indianapolis
aecer@iupui.edu
TeraGrid (http://www.teragrid.org) is an NSF-supported national effort to build and deploy the world's largest, most comprehensive, distributed infrastructure for open scientific research. It includes 20 teraflops of computing power distributed at nine sites with facilities capable of managing and storing nearly 1 petabyte of data, high-resolution visualization environments, and toolkit for grid computing. All the sites are tightly integrated and connected through a network that operates 40 gigabits per second. Indiana and Purdue Universities are connected to TeraGrid via I-Light.
ParaView (http://www.paraview.org) is an extensible, open source multi-platform application for visualizing large data sets. ParaView operates on a wide range of data formats including structured, unstructured grids, and adaptive mesh refinement data sets. ParaView includes suit of visualization algorithms, including contouring, clipping, streamlines, streamribbons, glyphing, and animations. In addition to these features, ParaView offers an advantage of running visualization processes in distributed and parallel fashion for processing large data sets. It runs parallel on distributed and shared memory systems using Message Passing Interface (MPI). ParaView uses the data parallel model. In this model, data is broken into the pieces to be processed by different processors which allows to visualize large data sets. This capability makes Paraview suitable visualization application for TeraGrid environment.
In this presentation, we will show how large data sets, produced by the parallel Computational Fluid Dynamics (CFD) applications can be visualized with distributed and parallel modes of the ParaView on the TeraGrid resources.
Acknowledgement. TeraGrid access was supported by the National Science Foundation (NSF) under the following programs: Partnerships for Advanced Computational Infrastructure, Distributed Terascale Facility (DTF) and Terascale Extensions: Enhancements to the Extensible Terascale Facility, with Grant Number: TG-CTS050003T.
Top of page
Parallel Sessions
Browsing Molecules in GEMS
Justin Wozniak
Computer Science and Engineering
University of Notre Dame
jwozniak@nd.edu
Paul Brenner
Computer Science and Engineering
University of Notre Dame
pbrenne1@nd.edu
Doug Thain
Computer Science and Engineering
University of Notre Dame
dthain@nd.edu
Aaron Striegel
Computer Science and Engineering
University of Notre Dame
striegel@nd.edu
Jesus Izaguirre
Computer Science and Engineering
University of Notre Dame
izaguirr@nd.edu
Biomolecular simulations produce more output data than can be managed effectively by traditional computing systems. Researchers need distributed systems that allow the pooling of resources, the sharing of simulation data, and the reliable publication of both tentative and final results. To address this need, we have designed GEMS, a system that enables biomolecular researchers to store, search, and share large-scale simulation data. The primary design problem is striking a balance between generosity and gluttony. On one hand, storage providers wish to be generous and share resources with their collaborators. On the other hand, an unchecked data producer can be gluttonous and easily replicate data unnecessarily until it fills all available space. To balance generosity and gluttony, GEMS allows both storage providers and data producers to state and enforce policies on the consumption of storage and the replication of data. By taking advantage of known properties of simulation data, the system is able to distinguish between high value final results that must be preserved and low value intermediate results that can be deleted and regenerated if necessary. We have built a prototype of GEMS on a cluster of workstations and demonstrate its ability to store new data, to replicate within policy limits, and to recover from failures.
A recent addition to the GEMS system is a GUI client that may be used to search the metadata catalog, and display 3D molecular structures all in the same client. We would like to perform a demonstration of GEMS' ability to catalog new datasets and display their scientific content on the fly.
Top of page
Parallel Sessions
Web Services in Life Science
Charles Moad
Scientific Data Analysis Lab
Pervasive Technology Labs at Indiana University
cmoad@indiana.edu
Randy Heiland
Scientific Data Analysis Lab
Pervasive Technology Labs at Indiana University
heiland@indiana.edu
Web Services have become a standard in application to application communication over the Internet. As the amount of data in the Life Sciences grows exponentially, researchers turn towards Web Services to provide them with the most up-to-date data, without requiring large in-house databases. We will discuss our approach of utilizing Web Services in several aspects of proteomics including non-synonymous single nucleotide polymorphisms (SNPs), mutations, and structural homology. Using open-source protein visualization applications and integrated Web Services, a researcher can interactively explore large remote databases or submit structures of their own for processing.
For more, see:
http://www.mutdb.org#Web Interface
http://www.lifescienceweb.org/beta#Web Service Interface
Top of page
Parallel Sessions
Posters and Poster Session Demonstrations – Wednesday, September 21 – 10:10-10:40am, 2:05-2:35pm, 3:50-4:14pm
Places & Spaces: Cartography of the Physical and the Abstract
Katy Börner
School of Library and Information Science
Indiana University Bloomington
Deborah MacPherson
Accuracy & Aesthetics
The Places & Spaces exhibit has been created to demonstrate the power of maps. An initial theme of this exhibit is to compare and contrast first maps of our entire planet with the first maps of all of science as we know it. Come see with your own eyes the extent to which maps can be employed to help make sense of the flood of information we are confronted with and how domain maps can be used to locate complex and beautiful information.
This exhibit presents a selected series of maps and their makers along with detailed explanations of why these maps work. The physical prints support the close inspection of high quality reproductions for display at conferences and education centers. It is meant to inspire cross-disciplinary discussion on how to best track and communicate human activity and scientific progress on a global scale. The online part of the exhibit is viewable at: http://vw.indiana.edu/places&spaces/
This exhibit was supported in part by the School of Library and Information Science, Indiana University and a National Science Foundation CAREER grant under IIS-0238261.
Top of page
Parallel Sessions
Pedigree Visualization and Navigation
Jagannathan Lakshmipathy
Advanced Visualization Lab
University Information Technology Services
Indiana University
jlakshmi@iupui.edu
Eric Wernert
Research and Academic Computing
University Information Technology Services
Indiana University
ewernert@indiana.edu
Michael Boyles
Advanced Visualization Lab
University Information Technology Services
Indiana University
mjboyles@iupui.edu
Traditionally pedigree data has been very useful to present and document complex information of an ethnic or a population subgroup. Various analysis tools were developed around the pedigree data to extract useful information. Recently these pedigree data has posed unprecedented challenges to the scientific community due to their increasing size. Some fundamental analysis tools have become a burden and have strained the conventional workflow. It is important to revisit those tools with the present state of technology and enhance the usability of the software and the efficiency of the workflow. In this poster, we present the design and implementation details of PViN (Pedigree Visualization and Navigation), a scalable and flexible software system that enables visualization, analysis and printing of pedigree database, and compare the present and the past workflow. The following are some compelling reasons for the PViN development: (1) Indiana university's center of hereditary diseases has accumulated very large amounts of hereditary information from various populations for ongoing research projects and has difficulty managing and effectively printing the associated pedigree trees with legacy FORTRAN software; (2) The size of some of these databases (over 40,000 entries covering seven generations) is too large for existing commercial pedigree software to handle; (3) Our researchers and support staff need more effective ways to perform visual analysis tasks, such as the comparison of multiple pedigrees and the cross-referencing of individuals that appear in multiple families (through re-marriage.) The PViN system addresses these fundamental problems while also providing a number of additional features and functions including: context-free drawing routines that enable rendering onto screen and printer contexts interchangeably; a generic framework that allows the system to interface with multiple databases and database servers; a multiple view user interface that provides side-by-side comparisons and "focus+context" rendering; and advanced node searching and cross-referencing capabilities.
Top of page
Parallel Sessions
SRS Browser: A Visual Interface to the Sequence Retrieval System
Ketan Mane
School of Library and Information Science
Indiana University Bloomington
kmane@indiana.edu
Katy Börner
School of Library and Information Science
Indiana University Bloomington
katy@indina.edu
This paper presents a novel approach to the visual exploration and navigation of complex association networks of biological data sets, e.g., published papers, gene or protein information. The generic approach was implemented in the SRS Browser as an alternative visual interface to the highly used Sequence Retrieval System (SRS) [1]. SRS supports keyword-based search of about 400 biomedical databases. While the SRS presents search results as rank-ordered lists of matching entities, the SRS Browser displays entities and their relations for interactive exploration. A formal usability study was conducted to examine the SRS Browser interface's capabilities to support knowledge discovery and management.
Top of page
Parallel Sessions
Third Generation Data Capture and Monitoring for High Interaction Honeynets
Ed Balas
Advanced Network Management Lab
Pervasive Technology Labs at Indiana University
edb@indiana.edu
Camilo Viecco
Advanced Network Management Lab
Pervasive Technology Labs at Indiana University
cviecco@indiana.edu
Until very recently researchers for high interaction honeynets faced the problem of how to integrate and relate the multiple data sources they have as network and host information into a unified and complete view. We will demo the current implementation of the third generation data architecture and model for high interaction honeynets which is part of the current honeynet project linux CD-ROM distribution "roo". This model allows intruders activity to be traced using process trees that also combine network information and IDS entries.
Top of page
Parallel Sessions
Scene Tunnel Acquisition for Cityscape Visualization
Yu Zhou
Computer and Information Science
Indiana University-Purdue University Indianapolis
yzhou@cs.iupui.edu
Jiang Yu Zheng
Computer and Information Science
Indiana University-Purdue University Indianapolis
jzheng@cs.iupui.edu
This paper proposes a visual representation named scene tunnel to archive and visualize urban scenes for Internet based virtual tour. We scan cityscapes using multiple cameras on a vehicle that moves along streets, and generate scene archive more complete than a route panorama. The scene tunnel can cover high architectures and various object aspects. It contains much less data than video, which is suitable for image transmission and rendering over the Internet. It has a uniformed resolution along the camera path and provides continuous views for virtual traversing of a real city. We have developed image acquisition methods from slit setting, view scanning, to image integration. We have also achieved city visualization with scene tunnels on the Internet by transforming view, streaming data, and providing interactions.
Top of page
Parallel Sessions
Zero-Crossing Iso-Surfaces in Volume Datasets
Shiaofen Fang
Computer and Information Science
Indiana University-Purdue University Indianapolis
sfang@cs.iupui.edu
Pooja Gupta
Computer and Information Science
Indiana University-Purdue University Indianapolis
pgupta@cs.iupui.edu
This paper describes a new surface extraction method for volume datasets by applying iso-surface extraction techniques to the zero-crossing edges in a volumetric domain. A volume dataset is first filtered using a Laplacian of Gaussian filter to generate a zero-crossing field. A marching cube process will then be able to extract the entire zero-crossing surface that may be viewed selectively based on various intensity ranges and gradient scales. This new technique provides a more efficient surface navigation and extraction mechanism, as well as more accurate surface details, than the traditional iso-surface techniques.
Top of page
Parallel Sessions
Indiana Center for Biological Microscopy – Interactive 3D Image Processing During Volume Rendering
Jeff Clendenon
Nephrology Division
Indiana University School of Medicine
jclende@iupui.edu
Jason Byars
Nephrology Division
Indiana University School of Medicine
jbyars@iupui.edu
Ken Dunn
Nephrology Division
Indiana University School of Medicine
kwdunn@iupui.edu
The Indiana Center for Biological Microscopy is a core facility at IUPUI providing university and commercial biomedical researchers with access to state-of-the-art confocal and two-photon laser-scanning microscopes, which can be used to collect multi-channel 3D and 4D (i.e. 3D time series) images of live or fixed specimens. Images are collected using the microscopes, then transferred over the network to PCs or Macs in either our facility or a researcher's office or lab, for processing and analysis.
But 3D optical microscopy depends heavily on interactive image processing, which often requires that microscopy-specific image processing techniques be incorporated into volume rendering software. Traditionally, 3D imaging programs use graphics processing units (GPUs) to perform real-time volume rendering, but do not use the GPU to perform image processing, due to the limited programmability and low-precision pixel arithmetic used in older GPUs.
We use an extended version of our Voxx volume rendering program and a
GeForce 6800 GPU to demonstrate how the improved programmability and
higher-precision pixel arithmetic available in newer GPUs can be used to perform real-time image processing during volume rendering of 3D/4D microscopy data.
The combination of our Voxx program, a low-cost GPU-based video board, and a PC or Mac makes a cost-effective 3D imaging workstation for microscopy-based research.
For more information, see; www.nephrology.iupui.edu/imaging/voxx/.
Top of page
Parallel Sessions
Real time rendering of route panoramas with GPU
Panayiotis Mili
Computer Science
Indiana University-Purdue University Indianapolis
pmili@iupui.edu
Jiang Zheng
Computer Science
Indiana University-Purdue University Indianapolis
jzheng@cs.iupui.edu
Our software will showcase the real-time route panorama rendering in 3D with the use of GPU (graphics processing unit). Our program uses special lens that modifies the vertices in real time with the result of a panorama lens effect. In order to achieve high performance real-time rendering, we implemented the special lens by programming the vertex pipeline of a video card. Our implementation is extremely fast on mainstream computers with video cards that support GPU programming.
For more information, see: http://www.cs.iupui.edu/~jzheng/RP
Top of page
Parallel Sessions
Applying Lagrangian-Eulerian Advection to Supernova Simulations
Edward Bachta
Visualization & Interactive Spaces
Pervasive Technology Labs at Indiana University
M. Pauline Baker
Visualization & Interactive Spaces
Pervasive Technology Labs at Indiana University
baker@iu.edu
Robert J. Stein
Visualization & Interactive Spaces
Pervasive Technology Labs at Indiana University
rjstein@iupui.edu
Simulations run by scientists collaborating on the TeraScale Supernova Initiative involve the complex mechanics of thermodynamics and particle physics and can generate enormous amounts of simulation data. How to represent this data and provide useful services to scientists are two of the challenges for visualization researchers on the project. One of the newer tools used in the exploration of vector fields is called Lagrangian-Eulerian Advection (LEA). This technique provides a rich amount of information compared to earlier methods such as arrow plots. Vector-scalar relationships can be shown by combining LEA with color and contours, while the relative orientation of vector fields can be portrayed by visualizing dot products. Experimentation with these techniques has helped provide insight into the science of these intriguing multivariate datasets.
Top of page
Parallel Sessions
Materials Designed by Discovery Informatics
James Caruthers
Purdue University
The building of models to describe significant physical processes is a key
activity in science and engineering. Model building involves the
intersection of experimental data with theory, where the objective is to
determine the relationship between experimentally measured properties and an
optimal set of descriptors of the process. The key intellectual challenge is
to determine the appropriate descriptors, which is ans evolving process as
the researcher gains insight about the phenomena.
High throughput experimentation, high throughput computation, dynamically
linked databases, and advanced visualization methods offer the possibility
of qualitatively changing the model building process. High-resolution
visualization with large screens enables the researcher to discover
relationships in the data. This is a process in which the researcher
modifies the model in real time using the information content embedded in
the data.
Top of page
Parallel Sessions
Purdue Portable Haptic Display for Large Immersive Virtual Environments
Enkhtuvshin Dorjgotov
Seungmoon Choi
Laura Arns
Raj Ararangarsan
Steven R. Dunlop
Gary R. Bertoline
Hong Z. Tan
Purdue University
The Purdue Portable Haptic Display is aimed at developing a
platform-independent haptic rendering system that can be easily integrated
to large immersive visual displays for multi-modal data perceptualization.
The Envision Center for Data Perceptualization at Purdue University has
several kinds of large immersive visual displays including a CAVE and a
tiled-wall display along with force-feedback haptic interfaces.
However, there exists significant incompatibility between both hardware and
software of the haptic interfaces and the visual displays. In order to
facilitate the use of the haptic interfaces with the large visual displays,
we have conceived the idea of a portable haptic display that has the
architecture of distributed rendering through network.
Top of page
Parallel Sessions
Evening Demonstrations in the ICTC – Wednesday, September 21 – 6:00-8:00pm
VR Theater
Cabinet of Dreams
Margaret Dolinsky
School of Fine Arts
Indiana University Bloomington
dolinsky@indiana.edu
Dimitrij Hmeljak
Org: Computer Science
Indiana University Bloomington
mitja@indiana.edu
Mathew A. Powers
School of Fine Arts
Indiana University Bloomington
matapowe@indiana.edu
Adam Schweigert
School of Music
Indiana University Bloomington
aschweig@indiana.edu
Craig Strubing
School of Fine Arts
Indiana University Bloomington
cstrubin@indiana.edu
"Cabinet of Dreams" is a 3D virtual reality showcasing highlights of the Chinese art collection at the Indianapolis Museum of Art (IMA). The objects range in date from 1000 BC to the mid-1800s and include wood, bronze, and earthenware ceremonial pieces. An inkstone is a small mountain landscape with a dragon. A brushpot is engraved with lacquers to show scenes of country pursuits. A ritual cooking vessel carries an ancient inscription of a man on a chariot. These magical items were studied both in the IMA archive vault and through digital technology as inspiration towards developing whimsical environments to re-exhibit objects in 3D interactive graphics.
Using VR, the items no longer remain in storage to sit frozen behind glass – they exist in another instance of time and space and are being displayed in real-time 3D stereo vision and audio using projection technology. In the virtual environment, the museum visitor is allowed to approach the precious object with a new sense of proximity, scale and viewing perspectives that real life cannot accommodate. It is a rare moment to be able to view all sides of an object as we did for our data. This type of intimacy could only occur for research purposes with the presence of a museum official.
One of the objects is a Qing dynasty cabinet made of cloisonné, glass, and zitan. The cabinet is the metaphoric center of the installation, reflected in the art and the display device as if it were a modern day Wunderkabinett. By combining the actual cabinet with the virtual dreams inspired by the real objects, computer graphic environments represent a structure of times past as well as a sense of virtual space. The actual cabinet object is on exhibition near the Asian galleries. Its presence hints at the collection of rarities that are hidden within the virtual environment and the museum itself.
The main virtual environment is a series of hallways, sliding doors and rooms. Each room houses a small version of the original cabinet. When approached, the cabinet doors open to display a floating image of another world. As the visitor enters the cabinet, time, sound and space fluctuate and one is transported to that alternate world. Each environment's audio is composed to establish a unique atmosphere based on the history of the object, its functions and its metaphoric environment in the computer graphics world. For example, an ancient wellhead depicts a shaman and tells the story of the eastern sky. In the virtual environment, as one moves closer to the wellhead, the audio changes with the sky, twinkling to night and revealing its constellations. Thirty objects from the museum collection comprise eleven virtual environments.
"Cabinet of Dreams" viewing occurs on 3D stereo display systems and visitors wear 3D glasses while navigating the environments.
Top of page
Parallel Sessions
Advanced Virtual Manufacturing Laboratory (AVML) for Training, Education, and Research
Hazim El-Mounayri
Mechanical Engineering
Indiana University-Purdue University Indianapolis
helmouna@iupui.edu
Tamer Wasfy
Advanced Science and Automation Corporation
tamer@ascience.com
Daniel Aw
Mechanical Engineering
Indiana University-Purdue University Indianapolis
daw@iupui.edu
Ayman Wafsy
Advanced Science and Automation Corporation
ayman@ascience.com
A high fidelity web-based virtual environment – Advanced Virtual Manufacturing Laboratory (AVML) – that uses advanced visualization techniques is presented. The dynamic virtual reality system is interactive and allows for the simulation of advanced manufacturing. Currently, the system supports CNC (Computer Numerically Controlled) milling, a versatile machining process that is widely used in industry.
The system provides a safe, cost-effective, and highly flexible and accessible tool for training and education as well as product realization. More specifically, the AVML provides the following training functions for users via the Internet: (a) interaction with a fully-functional virtual CNC milling machine, including real-time machining of parts, (b) training on key operating procedures of modern CNC machines, and (c) intelligent virtual tutor that gives a lecture describing the concepts of CNC milling and the components of the milling machine, as well as assistance with hands on training on the virtual machine.
The environment is driven by three software modules that communicate with each other using a TCP/IP network socket interface; these modules are: (1) a CNC Milling machine simulator, (2) a virtual-environment display engine, and (3) an intelligent-agent engine. The modules run on a single computer in a seamless web-based framework. The AVML can run on desktop or laptop personal computers. It can also run on more sophisticated systems such as a CAVE (for fully-immersive virtual reality visualization and simulation).
The AVML enables colleges and universities to easily and inexpensively provide students with effective, safe, and highly accessible training on advanced machine tools. It can also be used by machine tool manufacturers to provide online training, reducing or eliminating the need for onsite and/or live training classes for customers.
The work, which is still underway, has been presented in several conferences and symposia, including ASME, ASEE, ICCIE, and Solutions Conference 2005.
The methodology and its implementation, and the resulting virtual system for CNC machining will be presented. On the other hand, the current system's capabilities and functionality will be demonstrated in the CAVE, which is currently available in the Advanced Visualization Laboratory at IUPUI.
Top of page
Parallel Sessions
Interactive 3D visualization of kinetic sculpture
Nicoletta Adamo-Villani
Computer Graphics Technology
Purdue University
nadamovi@purdue.edu
Laura Arns
Envision Center for Data Perceptualization
Purdue University
larns@purdue.edu
The goal of this project is development of an interactive visualization of a sculpture to be placed on the east side of the College of Technology building on Purdue University West Lafayette campus. The 40 foot tall sculpture represents a torch with three rotating flames. The motion of the flames is initiated by the rotation of a metal disc at the base of the sculpture.
Traditionally, prior to the realization of the final sculpture, the sculptor builds a model to scale, which helps the artist visualize the form and three dimensionality of the object, but doesn't provide any visual feedback on how the sculpture will fit within the surrounding environment. Our visualization allows the sculptor to view and interact with a full scale model, placed in its surrounding environment. The ability to see the object from different points of view and in relation to the other campus buildings has provided the artist with the unique opportunity of truly testing the expressiveness of the art within the environment in which it will reside.
Because proper scale of the sculpture in comparison to the surrounding buildings was so important to the visualization of this project, satellite photos were used and the 3D models were built on top of this overhead reference, using Discrete's 3ds Max. Textures and colors were captured from digital photos of the buildings and were manipulated to maximize the possibility for repetition and reuse. Each model was exported separately into the OpenSceneGraph (www.openscenegraph.org) format for use with VRJuggler (www.vrjuggler.org), to display the virtual environment in immersive devices such as the Purdue Envision Center FLEX. A pair of Fakespace Labs' pinch gloves, along with a 6 degrees-of-freedom tracker by Intersense allows the user to interact directly with the application by detecting when the user makes gestures that can affect the virtual sculpture, such as "grabbing" the sculpture and performing constrained rotations of the metal disc.
Top of page
Parallel Sessions
Interactive 3D Sign Language in Immersive Environment
Nicoletta Adamo-Villani
Computer Graphics Technology
Purdue University
nadamovi@purdue.edu
Edward Carpenter
Computer Graphics Technology
Purdue University
ecarpen@purdue.edu
The general goal of this project (work in progress) is development of an immersive virtual environment in which deaf children can interact with fantasy 3D signers. The interactive signing avatars can be displayed in immersive devices such as the Purdue Envision Center FLEX.
Recently we have created a highly interactive computer animation program (Mathsigner©2004) for learning K-3 arithmetic skills, aimed at deaf children. The program includes 3D animated signers modeled as seamless low-poly meshes and rigged with a human-like skeletal deformation system. The virtual signers are animated using a library of signing clips recorded directly from a signer wearing a 19-markers motion capture optical suit and a pair of 18-sensors cybergloves. Smoothness of motion (in real time) and interactivity are achieved via programmable blending of the motion captured animation segments.
The specific objectives of this work in progress are: (1) display of the Mathsigner© animated 3D signers in an immersive environment (e.g., the FLEX); and (2) realization of a basic interaction between the user and the avatars. A pair of Fakespace Labs' pinch gloves, along with a 6 degrees-of-freedom tracker by Intersense will allow the user to interact directly with the application by detecting when the user makes gestures that can affect the motion of the virtual signer, such as "picking up" a virtual number (represented by a 3D block) thus causing the 3D avatar to come forward and produce the corresponding number sign.
The main challenge presented by this project is the display and interaction with seamless characters that deform organically during motion. In general, characters displayed in immersive environments are segmented characters made of rigid components which rotate without changing shape. In order to display the Mathsigner© 3D characters with realistic deformations during signing motion, we have used Cal3D character animation library (cal3d.sourceforge.net/), along with OsgCal. OsgCal is an adaptor for using Cal3D within OpenSceneGraph (www.openscenegraph.org), the format necessary for use with VRJuggler (www.vrjuggler.org).
The 3D rigged models were exported in different components using the four cal3D file formats: .cmf (Mesh file), .csf (Skeleton file), .caf (animation file), and .crf (texture file). The skeleton was exported first because exports of the mesh and animations need to reference the .csf file. All meshes required a material identifier in order to be exported, and every time a material was changed, the mesh had to be exported along with the new material. Once each separate file for cal3D animation was exported, the files were reassembled within the osg program using the osgCal libraries which allow to load the cal3D data into a model node within the scene graph. Each exported cal3D file can be read into the program through a configuration file, or can be hard coded directly into osgNav. After using osgCal to load the files, we are currently programming the interaction with instances of the model using calls to the osgCal library. We are using startLoop and stopLoop functions to cue animations; in addition, we are experimenting with un-looped animation, changes of animation speed, and blending of multiple animations by assigning them different weight values. Once all the animations are working within the osg program, interaction with the model as a scene graph node will be coded as well.
Top of page
Parallel Sessions
The Solar Journey Project
Andrew Hanson
Computer Science
Indiana University Bloomington
hanson@cs.indiana.edu
The Solar Journey Project is 6-year, multi-institution collaboration focused on the development and application of advanced visualization, modeling, virtual reality and analysis tools to virtual astronomy. While the specific scientific objective of this research is a better understanding of the local galactic neighborhood of our Sun, this collaboration has also resulted in the development of a range of new visualization technologies, including: multi-scale techniques for overcoming the precision limitations of graphics hardware; navigational techniques for effectively traversing many orders of spatial magnitude, methods for physically-based star rendering, and a software framework to seamlessly integrate realtime analysis with high-quality batch rendering. Specific productions highlighted in this talk include several high-quality animated shorts (including one shown in the SIGGRAPH 2000 Electronic Theater), an interactive CAVE application, and a 20-minute production for the multi-projector laser dome at the Beijing Planetarium.
For more information, see: http://www.cs.indiana.edu/~soljourn/
Top of page
Parallel Sessions
Passive Stereo/John-e-Box/Other Displays
Stereoscopic Visualization of Scientific and Medical Content
Albert William
School of Informatics
Indiana University-Purdue University Indianapolis
almwilli@iupui.edu
The use of stereoscopic display medium is now in use at the Ruth Lilly Health Education Center (RLHEC) in Indianapolis to help educate nearly 100,000 students annually. The stereoscopic display is helping to revolutionize a subject area to students K-12 in a subject matter that traditionally has been a very difficult area to learn. Content created by the research team at the IUPUI School of Informatics has been implemented to fit into the curriculum of the RLHEC. Future plans of the RLHEC include distributing this content to other health centers around the state, country, and globally, through high speed networks.
This presentation will display on a portable John-E-Box, or other portable stereo display system available to the School of Informatics, the display of a number of stereoscopic animations created for the RLHEC, including subjects such as The Cell, The Circulatory System, The Nervous System, The Immune System and The Respiratory System. Also displayed will be content created for the simulation of Hip and Knee Replacement Surgery, and simulations of real world data of biological macro-molecules.
Presentation will also include a description of content research procedures, pre-visualization methods, production methods, stereoscopic creation techniques, production issues and post-production methods. Considerations of the limitations and benefits of the stereoscopic medium will be discussed, as well as these considerations regarding the production of the stereoscopic content. A PowerPoint presentation can accompany the John E Box demonstration to further explain the production pipeline.
Top of page
Parallel Sessions
AGJuggler
Dioselin Gonzalez
Envision Center for Data Perceptualization
Purdue University
dioselin@purdue.edu
Laura Arns
Envision Center for Data Perceptualization
Purdue University
arns@purdue.edu
AGJuggler is a project developed at the Envision Center for Data Perceptualization at Purdue University. The main objective in this study is to make collaborative experiences and interactions richer with the incorporation of Virtual Reality tools inside a collaborative framework. This is accomplished by offering: an improved sense of presence; synchronous distributed visualization; real time interaction and a customizable solution for implementing virtual reality within a collaboration environment.
The Access Grid™ is the collaboration environment used for the study; while the API used for programming virtual reality applications is VRJuggler.
For more information, see: http://people.envision.purdue.edu/~dioselin/AGJuggler/
Top of page
Parallel Sessions
Developing and Deploying a Robust System for 3D Surface Scanning and Analysis
Jeff Rogers
Advanced Visualization Lab
University Information Technology Services
Indiana University
jlrogers@iupui.edu
Eric Wernert
Research and Academic Computing
University Information Technology Services
Indiana University
ewernert@indiana.edu
3D surface scanning is an evolving field that encompasses a wide variety of technologies and applications. For the past three years, the UITS Advanced Visualization Lab has been working with researchers in the IU School of Medicine, the Anthropology and Computer Science departments at IUPUI, and an international consortium of collaborators to develop an integrated system of hardware, software, and methodology to enable the reliable and accurate scanning of human heads and faces.
This NIH-funded collaboration is utilizing 3D surface scanning technology to supplement traditional physical measurements in the diagnosis and study of fetal alcohol spectrum disorders (FASD). Craniofacial anthropometry has been used to assess and describe abnormal craniofacial variation and the facial phenotype in many syndromes. It has also been used in the clinical setting as a diagnostic aid and as a means of objectifying clinical descriptions of individual patients. Through the use of 3D scanning, the project seeks to develop novel methods for assessing facial variations to provide more accurate diagnoses of the level of alcohol exposure. Moreover, by organizing the surface scans of subjects and control individuals in a database, researchers will be able to perform novel queries, comparisons, and longitudinal studies. This project has developed three portable 3D scanning systems that have been deployed to a variety of locations throughout the US as well as sites in Europe and Africa.v
This talk will focus on our experiences in developing a robust and accurate 3D surface scanning system for the FASD project, including the evaluation of competing technologies, the development of software procedures for stitching, measurement, and analysis of models, and the challenges of international technology deployment. We will also highlight some of the outgrowth applications of this technology, including scanning of objects for forensics, cultural heritage, and digital art.
Top of page
Parallel Sessions
|
