Stochastic Systems Group
		Home Group Members Publications Research Alumni Mailing List Demos Mission Statement Programs Seminars Library

• Probabilistic Modeling and Statistical Inference for Random Fields and Space-Time Processes, Office of Naval Research (N00014-91-J-1004)
  
  The focus of this project is on basic theory for spatially-distributed and space-time random phenomena. Areas of current interest focus on fusion and assimilation of large, multi-sensor data sets in remote sensing using multiresolution stochastic models. The theory of such models, especially for handling multiresolution data and nonstationary phenomena is of particular importance. While there are no formal relationships tied to this grant, this research involves substantial collaboration with two M.I.T. groups: one headed by Prof. Carl Wunsch (Dep't. of Earth and Planetary Sciences) for applications in oceanography and one headed by Prof. Dennis McLaughlin (Civil Engineering) for applications in hydrology. An addition to this work also involves interaction with Dr. David Rossi of Schlumberger Studies and Production (Clamart, France) in the area of geophysical exploration.
  
  
• Multiresolution Methods in Systems, Signals, and Images, Air Force Office of Scientific Research(F49620-95-1-0083)
  
  The focus of this project is on the basic theory of multiresolution stochastic models with applications to a variety of signal and image processing problems including the modeling and analysis of fractals, motion estimation from image sequences, image segmentation, and efficient algorithms for optimal detection in multiple access communications. There are no formal relationships tied to this grant, but the research involves both collaboration and transitions of our research to Alphatech, Inc. (point of contact: Dr. Robert Washburn, Vice-President) and Lincoln Laboratory (point of contact: Dr. Donald Boroson).
  
  
• Defense Advanced Research Project Agency (through AFOSR and Alphatech) (F49620-93-1-0604)
  
  These are an interconnected set of research projects aimed at developing a fundamental, unified theory for automatic target recognition (ATR), including basic sensor physics, image and signal analysis, and object recognition and performance analysis. These efforts involve substantive interactions with two other groups at M.I.T., one headed by Prof. Jeffrey Shapiro and the other by Prof. Eric Grimson and Paul Viola. Early work on one of these projects (61477) involved research on several sensing modalities including laser radar and synthetic aperture radar (SAR), but current and planned activities have focused and will focus on SAR-based ATR. These projects also involve considerable collaboration with Lincoln Laboratory (Drs. Leslie Novak and Daniel Dudgeon) and Alphatech (Drs. Robert Washburn, William Irving, Ronald Chaney, and Alan Chao), and one of these projects (63645) is in fact a joint project with Alphatech and Yale University.
  
  
• Air Force Office of Scientific Research (through Boston University)
  
  This project represents a large, multi-university effort led by Boston University (Drs. David Castenon and W. Clem Karl) and also involving Northeastern University, Courant Institute, University of Michigan, University of Minnesota, and Stanford University (See MURI web-site). The overall theme of this research is the development of basic analytical methods for reduced-signature target recognition. Other M.I.T. participants in this program include Prof. Jeffrey Shapiro, Prof. Gilbert Strang, Prof. Eric Grimson, and Prof. Paul Viola. Current work on this project includes nonlinear evolution for image segmentation and image analysis for moving targets.
  
  
• Army Research Office (through Washington University)
  
  This project is part of the Center for Imaging Science, based at Washington University (Prof. Michael Miller) and also involving the University of Texas. The basic objective of this research center is the introduction and study of innovative stochastic models for imagery. The intellectual aim is the development of a fundamental understanding of problems of extracting information from imagery, and the application context is primarily automatic target recognition. Other M.I.T. participants include Prof. Jeffrey Shapiro and Sanjoy Mitter and Dr. Thomas Green of Lincoln Laboratory. Sensors of interest here include electro-optical, lasar radar, as well as radar. Current projects include statistical methods for multiresolution image segmentation and for image formation in the presence of lasar speckle.
  
  

Problems with this site should be emailed to jonesb@mit.edu