History The Image Processing and Neural Networks Laboratory (IPNNL) has existed, in the Department of Electrical Engineering at the University of Texas at Arlington since 1985. The research and consulting activities of the laboratory have been funded by many sources in industry and government.

 

Current Work Current work focuses on (1) development of advanced neural network training and pruning algorithms, (2) simultaneous training and validation of multiple networks, (3) feature selection for estimation and classification, (4) reduction of input and output redundancies, (5) network sizing, (6) applications including fingerprint verification, face recognition, speech recognition, automatic target identification, forecasting,  prognostics, and (7) the production of user-friendly software for implementing these systems.

 

Lab Facilities The Image Processing and Neural Networks Laboratory has facilities in room 136 of Nedderman Hall. Equipment reserved for image processing and neural networks includes four Pentium 4 computers, extensive libraries of software for signal processing, image processing, and neural networks. Personnel in the IPNNL also have access to Sun workstations, a model 6400 nCUBE 2 parallel processing computer, a Meiko Computing Surface at the Automation and Robotics Research Institute (ARRI) in Fort Worth, and a Cray Y-MP 8/864 at the UT System Center for High Performance Computing in Austin.

 

Recent Sources of Support NSF, NASA, E-Systems, Garland and Greenville divisions, Loral- Vought, Texas Advanced Technology Program, Teledyne Controls, EPRI, Abbot Labs, Williams Pyro, and Lockheed Martin.

 

Output Software produced by consulting activities, and related to lab projects, can be found in the Neural Network Software area. Publications, dissertations, and theses can be found in the Neural Network Publications section.