Michael D. DeVore
Assistant Professor
Systems and Information Engineering
University of Virginia
151 Engineer's Way
P.O. Box 400747
Charlottesville, VA 22904

Office: 102G Olsson Hall
e-mail: mdevore@virginia.edu
Phone: (434) 924-4488
Fax: (434) 982-2972

Laboratories	Courses	Students	Research	Publications
MATLAB parallel-processing	MATLAB Tutorial

LABORATORIES

	Computer Vision Information Systems Laboratory The Computer Vision Information Systems (CVIS) Laboratory is the research laboratory of Prof. Michael D. DeVore. Research in the laboratory is at the intersection of computer vision systems (image and sensor modeling, object shape and reflectance models, object recognition, accuracy and complexity) and of information systems (distributed hardware and software systems, throughput and performance models, data movement and storage). See Research, below.

	Wireless Internet Center for Advanced Technology The Wireless Internet Center for Advanced Technology (WICAT) is a multi- university R&D center sponsored by the National Science Foundation under its program of Industry/University Cooperative Research Centers (I/UCRC), of which there are about 40. Polytechnic University is the lead institution in WICAT, which also includes Columbia University and the University of Virginia

COURSES

Fall, 2006: SYS682/ECE781 Statistical Pattern Recognition; SYS453 Senior Design I
Spring, 2006: SYS202 Data and Information Engineering; SYS454 Senior Design II

STUDENTS

Current Graduate Students

Xin Zhou, Ph.D.: Object Recognition from Three-Dimensional Point-Cloud Data; Model Fitting for THz Spectral Measurements of Biological Macro-Molecules
Kuei-Yung Teng, Ph.D.: Human-Computer Collaborative Systems for Target Detection and Recognition
Shankha Basu, M.S.: A High-Level Modeling Framework for Total System Performance in Optical Recognition Systems
Megan White, M.S.: Infrastructure for Persistent Surveillance of a Metropolitan Region

Undergraduate Research Assistants

Adebolanle Badejo, 3^rd Year: Web-Enabled Database of Registered 2D Electro-Optical and 3D Laser-Scanned Imagery

Former Graduate Students

Joshua Stafford, M.S. May 2005: Performance Analysis of a Three-Dimensional Point Measurement System with Six Degrees of Freedom

RESEARCH

My research is best summarized by the term "Computer Vision Information Systems." This involves two key elements that are a natural fit, but are seldom seen together:

Computer Vision: Incorporating imaging and sensor modeling; object shape and reflectance models; automated object recognition; and accuracy and complexity tradeoffs
Information Systems: Including computational hardware; throughput and performance models; and data representation, movement, and storage

In short, my interests lie in computer vision problems in a network-centric domain. Pattern recognition systems are generally reported in the literature as stand-alone systems, but most applications actually involve a complicated environment. Future systems will consist of distributed collections of sensors, processors, databases, and information consumers into a collaborative system for analyzing some theatre of operation. While individual elements from this list have been looked at extensively, there is a dearth of research into several unifying subjects.

Specifically, current research involves the following:

Automatic object recognition and object pose estimation from radar, electro-optical, and ladar data
Statistical inference from incomplete models
Statistical model assessment, representation, and storage
Statistical shape theory and shape representations
Performance subject to constraints on data complexity and system resource availability
Distributed architectures and systems with multiple, simultaneous objectives
Architectures for high-throughput recognition systems

Research Projects:

Active Computations for Networked ATR Systems - Office of Naval Research
Prototype System for TeraHertz Spectroscopy of Biological Molecules - W. M. Keck Foundation
Infrastructure for Persistent Surveillance of a Metropolitan Region - Accenture (WICAT)

PUBLICATIONS

Journal Papers

Shape Recognition From Three-Dimensional Point Measurements with Range and Direction Uncertainty: Optical Engineering, December 2005.
Estimates of Error Probability for Complex Gaussian Channels with Generalized Likelihood Ratio Detection: Transactions on Pattern Analysis and Machine Intelligence, October 2005.
Quantitative Statistical Assessment of Conditional Models for Synthetic Aperture Radar: IEEE Transactions on Image Processing, February 2004
Target-Centered Models and Information-Theoretic Segmentation for Automatic Target Recognition: Multidimensional Systems and Signal Processing, January 2003
Performance Complexity Study of Several Approaches to Automatic Target Recognition from SAR Images: IEEE Transactions on Aerospace and Electronic Systems, April 2002.
SAR ATR Performance Using a Conditionally Gaussian Model: IEEE Transactions on Aerospace and Electronic Systems, January 2001.

Conference Papers

Minimum Probability of Error Recognition of Three-Dimensional Laser-Scanned Targets: Proc. SPIE Automatic Target Recognition, Vol. 6234, 2006
Engineering Trade Study: Extract, Transform, Load Tools for Data Migration: Proc. IEEE Systems and Information Engineering Design Symposium, 2005.
Analytical Performance Evaluation of SAR ATR with Inaccurate or Estimated Models: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 5427, 2004.
Simulation of a Distributed Target Recognition System with Variable Operating Conditions: Proc. IEEE Systems and Information Engineering Design Symposium, 2004.
Azimuth Correlation Models for Radar Imaging: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 5495, 2003.
Conditionally Gamma- and K-Distribution Models for SAR ATR: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 4727, 2002.
Tradeoffs Between Quality of Results and Resource Consumption in a Recognition System: Proc. IEEE International Conference on Application-Specific Systems, Architectures, and Processors, 2002.
Relationships Between Computational System Performance and Recognition System Performance: Proc. SPIE Automatic Target Recognition, Vol. 4379, 2001
Statistical Assessment of Model Fit for Synthetic Aperture Radar Data: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 4382, 2001.
Probabilistic Approach to Model Extraction from Training Data: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 4382, 2001.
Performance Complexity Tradeoffs for Several Approaches to ATR from SAR Images: Proc. SPIE Algorithms for Synthetic Aperture Radar, Vol. 4053, 2000.
ATR Performance of a Rician Model for SAR Images: Proc. SPIE Automatic Target Recognition, Vol. 4050, 2000
Analytical and Experimental Performance-Complexity Tradeoffs in ATR: Proc. IEEE Asilomar Conference on Signals, Systems, and Computers, 2000.

Workshop Presentations

Dependence of Recognition Accuracy on Available Network Bandwidth: Proc. High Performance Embedded Computing Workshop, Lincoln Laboratories, September 2001.
Model-Based Approaches to Rejecting Confuser Targets: Proc. Automatic Target Recognition Theory Workshop, Wright State University, August 2001.
Analysis of Computational System Performance in Automatic Target Recognition: Proc. High Performance Embedded Computing Workshop, Lincoln Laboratories, September 2000.

Doctoral Dissertation

Recognition Performance from Synthetic Aperture Radar Imagery Subject to System Resource Constraints: Washington University in St. Louis, August 2001.; Advisor: Joseph A. O'Sullivan