Research Plan Presentation

1
Research Plan Presentation

• Cornelius Toole, Jr.
• Louis Stokes Mississippi Alliance for Minority
  Participation Bridge to the Doctoral Program
• Department of Computer Science
• Jackson State University
• September 16, 2004
• Program Director Dr. James Perkins
• Research Advisor Dr. Loretta A. Moore

2
Outline

• Advisor
• Areas of Interest
• Career Objectives
• Academic Preparation
• Research Experience
• Topic Areas Overview
• Research Plan
• Methodology
• Activities / Coursework
• Needed Preparation
• Prospective Doctoral Programs

3
Advisor - Dr. Loretta Moore

• Educational Background
• PhD., Computer Science, Illinois Institute of
  Technology, 1991
• M.S., Computer Science, Illinois Institute of
  Technology, 1986
• B.S., Computer Science, Jackson State
  University, 1985
• Current Appointment
• Associate Professor and Chair of Computer
  Science Department ( 1998 present )

4
Advisor - Continued

• Academic and Industrial Experiences
• Auburn University, Lawrence Livermore National
  Laboratory, NASA Marshall Space Flight Center,
  NASA Kennedy Space Center, ATT Bell Labs
• Research Interests
• Intelligent systems, data fusion, graphics and
  image processing, information assurance, computer
  security

5
Academic Preparation

• Began Computer Science Program at JSU Fall 1999
• Began Art Program at JSU mid-semester Fall 1999
• Courses Overview
• Java, C, C programming
• Data Structures, Discrete Structures, Programming
  Languages, Computer Architecture, Algorithms,
  Operating Systems, Data Mining, Computer Ethics,
  Senior Project
• Biology, Chemistry, Physics I II
• Trigonometry, Calculus I II III, Probability
  and Statistics I II
• Graphic Design, Drawing, Computer Art
• BS Computer Science May 2003

6
Areas of Interest

• Computer Graphics and Visualization
• Interface Design
• High Performance Computing

7
Career Objectives

• Complete Masters Degree in Computer Science at
  Jackson State University
• Enter complete PhD program in Computer Science
• Attain academic/research appointment
• Conduct research and development in the areas of
  information/scientific visualization and new
  media tools
• Form RD lab
• Creation of coalitions between academia and
  industry in visual computing

8
Research Experience

• Student Research Assistant in JSU Computer
  Science Department, spring 2001 fall 2002
• Topic Machine Vision, Image Processing, Smart
  Camera
• Advisor Dr. Loretta Moore
• Student Research Intern at the Computations/Engine
  ering Directorates at LLNL, summer 2001
• Topic Smart Camera
• Project Manager Doug Coffland
• Advisor Virgil Kohlhepp
• Student Research at the Engineering Directorate
  at LLNL, summer 2002
• Topic Smart Camera
• Project Manager Doug Coffland
• Advisor Virgil Kohlhepp

9
Research Experience cont'd

• Student Research Intern at Geo-Structures
  Laboratory at the U.S. Army Corps of Engineers
  ERDC, summer 2003
• Topic Protective Structures
• Division Chief Frank Dallriva
• Advisor Matt Hosseley
• Student Research Guest at Center for Science and
  Engineering Education / Computing Sciences
  Division at LBNL in the Visualization Group,
  summer 2004
• Topic Visualization of Discrete Ordinates Method
  Radiation Transport Data
• Mentor John Shalf

10
Undergraduate Research Experience

• Advisor -Dr. Loretta Moore
• Began research, which was funded by LLNL, on
  image processing, machine vision, and artificial
  intelligence at JSU Spring 2001
• Preparation for research experience at LLNL to
  work on collaborative project
• Wrote a report on machine vision techniques and
  applications
• Continued this work until the Spring of 2002

11
Undergraduate Research Experience Smart Camera
Camera/Sensor
12
Undergraduate Research Experience Smart Camera

• Summer 2001
• Project Manager Doug Coffland
• Advisor Virgil Kohlhepp
• Research experience at Lawrence Livermore
  National Laboratory to work on Smart Camera
  collaborative project
• Activities
• System configuration, systems support, video
  capture, machine vision software debugging and
  testing, developed Windows drivers
• Presented work at ACM Mid-Southeast Region
  Conference and Science and Engineering Alliance
  Symposium

13
Undergraduate Research Experience Smart Camera _at_
JSU

• Fall 2001 Spring 2002
• Continued work from Summer 2001
• Prepared paper on this work
• Presented at the Science and Engineering Alliance
  Symposium, October 2001, in Gaithersburg, MD
• Presented paper at ACM Mid-Southeast Region
  Conference, Nov. 2001, in Gatlinburg, TN
• Continued Smart Camera research in preparation
  for another experience at LLNL for Summer 2002

14
Undergraduate Research Experience Smart Camera

• Summer 2002
• Project Manager Doug Coffland
• Advisor Virgil Kohlhepp
• Continued work at Lawrence Livermore National
  Laboratory on Smart Camera collaborative project
• Activities
• Ported video control software to Linux for video
  capture
• Presented work in Student Poster Symposium at LLNL

15
Undergraduate Research Experience Smart Camera

• Fall 2002
• Presented work from Summer 2002 at ACM
  Mid-Southeast Region Conference, November 2002,
  Gatlinburg, TN
• Won 3rd place in Best Student Paper Category

16
Undergraduate Research Experience Protective
Structures

• Advisor Matt Hosseley
• Division Chief Frank Dallriva
• Research experience Summer 2003 at
  Geo-Sciences/Geo-Structures Laboratory at the
  U.S. Army Corps of Engineers Engineering and
  Development Research Center to work on Protective
  Structures Project
• Wrote software for analysis of explosives test
  data
• Such a tool would be used in the determination of
  blast area safe zones

17
Graduate Research Experience Radiation Transport

• Part of a scientific team interested in
  discovering the cause of supernovae
• Radiation transport involved in key processes
  taking place in supernovae
• Radiation transport describes processes that
  occur when the properties of protons are altered
  by scattering, absorption, and reemission of the
  matter through which the protons are passing

18
Graduate Research Experience contd
Integro-differential form of Boltzmann Radiative
Transfer Equation

• Equation extremely difficult to solve because of
  the integration on an angle
• Scientist uses Discrete Ordinates Methods to
  circumvent this
• DOM divides the domain into regions called
  ordinates through discretization

Discretization of integration over an angle by
quadrature sums times appropriate weights at
finite points in interval
19
Radiation Density Visualization
20
Radiation Density Relief
21
Time Derivative of Radiation Density
22
Summer Research Summary

• Provide scientist with tools to explore radiation
  transport data
• Interest in remote distributed visualization
• Chose this work in order to gain experience in
  scientific visualization and to better understand
  needs and challenges of visualization first

23
Scientific Visualization Definition

• Visualization is a method of computing. It
  transforms the symbolic into the geometric,
  enabling researchers to (more effectively)
  observe their simulations and computations.
  Visualization offers a method for seeing the
  unseen. It enriches the process of scientific
  discovery and fosters profound and unexpected
  insights...
• -B. McCormick, T. DeFanti, and M. Brown

24
Scientific Visualization Rationale

• Has promise as method of interpretation and
  navigation through huge amounts of data
• It is estimated that 50 of the brain's neurons
  are associated with vision. "The purpose of
  scientific computing is insight, not numbers."
• - Richard Hamming,1982.

25
Principles of Scientific Visualization
The scientific investigation model as it relates
to scientific visualization
26
Distributed Systems Computational Grids

• Can support the execution of large-scale,
  resource-intensive, distributed applications

• Distributed systems
• Linking resources in a geographically remote
  manner

27
Distributed Systems Computational Grids

• Example of an e-ChemLab enabled by grid technology

• Spring 2004 special topics course on grid
  technology
• Instructed by Drs. Geoffrey Fox, Bryan Carpenter,
  and Marlon Pierce from Indiana University

28
Distributed Visualization Architecture DiVA
29
Distributed Visualization Motivation

• Resources and capabilities for visualization are
  dispersed.
• Large data store may be at one site
• Special display system at another
• High graphics performance processing capabilities
  may exist at a rendering farm in another site
• What if we could create a system to take
  advantage of all these capabilities in an
  application that works seamlessly as a single
  machine?
• Remote distributed visualization (RDV) can enable
  this

30
Distributed Visualization

• Applications
• Particle Physics
• Astronomy
• Genomics
• Bioinformatics
• Computation Biology
• Any scientific domain that generates huge amounts
  of data and can take advantage of high
  performance computing

31
Distributed Visualization Architecture Remote
Distributed Visualization

• Key Problem Areas in Remote Distributed
  Visualization
• New Network Transport and Communications
  Protocols
• Job Control
• Brokering of Grid Services for Visualization
• Common / Unified Data Formats
• New Visualization Algorithms

32
RDV Need for a Broker

• Possible to manually configure a RDV application
  on limited number of machines in a distributed
  environment
• Infeasible with many ( hundreds ) of machines and
  components not to mention that one would have to
  know all the info about each component
• This needs to be automated could be done by an
  agent called a broker

33
Visualization Services Broker

• A Broker provides services for
• Service/Component Discovery Location
• Performance Prediction / Modelling
• Optimal Configuration
• Application Invocation
• Permissions Granting / Acquisition
• Recording Performance History / Configuration
  Refinements

34
Research Plan

• Tentative Thesis Topic Area
• DiVA Distributed Visualization Architecture
• Brokering for Visualization Grid Services
• Research Goal
• To develop and implement components that provide
  brokering services for distributed visualization
  applications
• This work's main focuses are studies in
  visualization, and distributed computing.

35
Research Plan cont'd

• Currently in literature review
• DiVA research is being conducted at
• Lawrence Berkeley National Laboratory
• Carnegie Mellon University
• Louisiana State University
• Becoming familiar with grid computing frameworks
• Studies of various visualization techniques

36
Research Plan Methodology

• Implement visualization components as grid
  services
• Implement performance models and profiles for
  each component
• Implement components for service location
• Implement components for optimal configuration
  selection
• Implement agent to invoke visualization pipeline
  application
• Implement methods to monitor and refine
  application performance

37
Activities / Coursework
38
Publication Plans / Professional Development

• ACM Mid Southeast Region Conference in
  Gatlinburg, TN, Nov 2004 present paper on work
  at LBNL
• Overset Symposium in Huntington Beach, CA, Oct 5
  -7, 2004
• Winter 2004
• Publications on visualization brokering grid
  services

39
Preparation Needed

• Mathematics
• Reinforcement of calculus, linear algebra,
  differential equations, set theory and logic, and
  probability statistics
• Numerical analysis
• Computer Graphics
• Computer Animation

40
Preparation Needed

• Neural Networks
• Algorithm Design and Analysis( reinforcement )
• Parallel Programming
• Distributed Computing
• Web\Grid Services

41
Proposed Doctoral Programs

• University of California Davis Computer Science
  Program
• Kwa Liu Ma, PhD
• Scientific Visualization / Information
  Visualization
• University of Illinois Chicago Computer Program
• The Electronic Visualization Laboratory
• University of Florida Computer Science Program
• Paul Fishwick
• Aesthetic Computing

42
Acknowledgements

• My parents
• My wife
• Jackson State University Computer Science
  Department Faculty and Staff
• Dr. Loretta Moore
• Dr. James Perkins
• The National Science Foundation and the Louis
  Stokes Alliance for Minority Participation
• The Office of Science of the Department of Energy
  and Lawrence Berkeley National Laboratory