🚀 go-pugleaf

See: http://amp.ece.cmu.edu/ECESeminar/
Prof. Tsuhan Chen, Host, 2000 ECE Seminar Series, email: tsuhan@ece.cmu.edu

                        SPRING 2000 ECE SEMINAR SERIES
                      ____________________________________

                           THURSDAY, APRIL 27, 2000
                                    4:00 PM
                            Refreshments:  3:30 PM
                        Scaife Hall Auditorium, Room 125
                      ____________________________________

                              Dr. Thomas H. Drayer
                          U.S. Department of Defense

         ``GOVERNMENT RESEARCH OPPORTUNITIES IN MULTIMEDIA PROCESSING''
                
                                  ABSTRACT

This talk will discuss opportunities for working with Government research
laboratories in the fields of image, video and multimedia processing by
illustrating current areas of interest and research programs in both
hardware and software system development. Recent declines in the cost
of transmitting and storing digital information have made it economically
viable to create large databases of heterogeneous multimedia data. It is
desirable to allow users to search and browse the documents of these
databases based upon multimedia content. However, manual annotation is
infeasible in all but a few specialized cases, requiring the automated
extraction of the multimedia content. This automated extraction of content
from large heterogeneous databases requires both robust algorithms and
high computational throughput beyond today's capabilities. Several possible
solutions are expected to have a significant impact on the challenges
associated with automated content analysis. These solutions include
algorithms that combine results from multiple modalities of the same signal
(image, audio, and metadata) or algorithms that combine knowledge from
existing external sources to provide increased accuracy. New computing
paradigms such as adaptive computing systems that can be reconfigured
for processing large volumes of streamed audio, video or metadata also
show promise. This talk will provide a roadmap for near-term and long-term
research projects in automated content extraction from multimedia data.

                                   BIO

Dr. Drayer is currently employed by the United States Department of
Defense and by Virginia Polytechnic Institute and State University. He joined
the Department of Defense in 1997 as senior engineer to perform image
processing research. Dr. Drayer also became an adjunct faculty member
with Virginia Tech at their graduate campus in Northern Virginia in the same
year. At Virginia Tech, he teaches courses in Hardware Descriptive
Languages, Image and Video Compression, and Fundamentals of Computer
Science. His other interests include pattern recognition, machine vision, and
reconfigurable or adaptive computing. Previously, Dr. Drayer had been
employed at the Spatial Data Analysis Laboratory of Virginia Tech. His
research work at the SDA lab included the design, implementation, and
testing of real-time reconfigurable computer architectures for image
processing applications. This included the development of machine vision
systems for hardwood lumber defect detection, wood panel color sorting,
and airport luggage explosives detection. Tom Drayer had a long and
distinguished career as an undergraduate and graduate student at Virginia
Tech, completing his PhD work as a Bradley Fellow in January of 1997. His
research was on a development system for real-time image processing
hardware. He is the author of several papers on hardware system design
and image processing development systems and has been awarded a
patent for the hardwood lumber defect detection system.