NPAC REU 94 - Available Projects
-
INTERACTIVE TV
Project Mentor: Joan Deppa (jadeppa@mailbox)
This project will evaluate technologies required to deliver
video segments from a central server as requested interactively
by remote viewers.
-
INTERACTIVE MULTIMEDIA INFORMATION SYSTEMS
Project Mentor: Joan Deppa (jadeppa@mailbox)
This project will involve the development of an interactive
multimedia information system for Rome Laboratory. A system
developed previously for Syracuse University (OrangeSource)
will provide the basis for this project.
-
OPTIMIZATION USING SIMULATED ANNEALING AND TEMPERING
Project Mentor: Paul Coddington (paulc@npac)
Simulated annealing is a very powerful and general optimization
technique. We are working on improving simulated annealing by using
ideas from statistical mechanics, including a new technique known
as simulated tempering.
-
RANDOM NUMBER GENERATORS/MONTE CARLO SIMULATION
Project Mentor: Paul Coddington (paulc@npac)
Random number generators are widely used in computer simulation,
but high quality (i.e. good randomness properties) generators are
hard to find, as are good tests for randomness. This is especially
true for parallel computers. This project will develop and test
(using Monte Carlo simulations) a variety of parallel algorithms for
generating random numbers, including an intrinsic function for
High Performance Fortran.
-
DR LINK NATURAL LANGUAGE TEXT RETRIEVAL
Project Mentor: Nancy McCracken (nancy@npac)
The DR Link system provides intelligent text retrieval by using natural
language processing to find linguistic characteristics of the query and
use them in retrieving text. This approach requires more processing power
than traditional methods based on word matching and high performance
computing may help achieve good performance. A student working in this
area would help port the text retrieval system to the IBM SP1 from its
current workstation base or participate in ongoing research in detecting
meaningful linguistic characteristics through natural language processing
(or both). The student would participate with the DR Link research group
in the Information Studies department.
- MULTIMEDIA SCRAPBOOK USING MOSAIC
Project Mentor: Paul Coddington (paulc@npac)
Create an on-line multimedia scrapbook using Mosaic, including articles
(text and images) from magazines and newspapers, and digitized videos.
This involves learning how to create, edit and display digitized images
and video, use optical character recognition (OCR) software to create
electronic text from hardcopies, and use Mosaic to format and display the
multimedia documents. The project also requires the creation of an
interface to the scrapbook to allow it to be easily searched and browsed.
-
THE LIVING TEXTBOOK
Project Mentors: Nancy McCracken (nancy@npac) and
E.A. Bogucz (bogucz@npac)
The Living Textbook Project is a joint project between NPAC, the
School of Education, NYNEX, and Columbia University to define how new
communication technologies will enhance the quality of teaching and learning.
Several pilot projects design multimedia educational applications and tools
appropriate for high-speed comuting and communications networks.
Possible student projects include designing the interface for a
meteorological education module that would allow students in the classroom
to explore tornados from the classroom by running a simulation on a parallel
computer over a high-speed network, or redesigning the computational
organization of the MapPower Geographic Information System system by sending
some of the computation from a Macintosh to a Unix host.
-
REAL-TIME TERRAIN RENDERING FOR A GIS
Project Mentor: Paul Coddington (paulc@npac)
Use Geographic Information System (GIS) terrain elevation and Landsat
data to construct a realistic real-time 3D rendering to simulate an
interactive flight over New York state. This will involve investigation
of parallel rendering algorithms.
-
VIRTUAL REALITY
Project mentor: Wojtek Furmanski (furm@npac)
Install and investigate available public domain VR packages and
develop demonstrations using these systems.
-
PARALLEL METHODS FOR COMPUTATIONAL FLUID DYNAMICS
Project Mentors: Thong Q. Dang (tqdang@mailbox) and
E.A. Bogucz (bogucz@npac)
This project involves the evaluation of different solution
schemes for computational fluid dynamics. The REU participant will
modify an existing simulation program for supersonic mixing
layers to evaluate effects of alternate approximation schemes.
-
VISUALIZATION OF UNSTEADY FLOWS
Project Mentor: Ammar Degani (degani@npac)
The purpose of this project is to develop an environment for the
visualization of unsteady fluid flow simulations. Input data will
be obtained from existing codes. The REU participant will learn to use
visualization software and prepare modules for visualizing instantaneous
flow fields and animating sequences of images.
-
VIDEO DATABASE BROWSING/INDEXING
Project Mentors: Marek Podgorny (podgorny@npac) and
Roman Markowski (roman@npac)
Use speech recognition software for automatic indexing of a video
database based on the audio content, with either text or audio
keyword input from the user.
-
CREATING AND PORTING APPLICATION KERNELS
Project Mentor: Ken Hawick (hawick@npac)
The student would work on a subset of a large project to create exemplar
program kernels for a wide range of applications, in sequential and
parallel (message-passing and High Performance Fortran) versions.
Many of the targeted applications concern Computational Fluid Dynamics but
the application areas can be contingent on the student's background and
interests. This project would provide good experience in writing parallel
programs.
- APPLICATION FOR THE VIRTUAL COMPUTING ENVIRONMENT
Project Mentor: Salim Hariri (hariri@cat)
A network of supercomputers and high-performance workstations appears to be
a cost-effective approach to provide adequate computing resources for the
Grand National Challenge problems. Such a collection of computers and
supporting software environment is called a Virtual Computing Environment
(VCE). This project will involve development of an application that can
utilize dynamically the computing resources of different parallel
computers available at NPAC.