UNIFIED GENERALIZED BAYESIAL ACCRUAL OF EVIDENCE FOR DATA FUSION
Small Business Information
500 WEST CUMMINGS PARK, Woburn, MA, 01801
AbstractThe lack of a solid foundation for filtering and fusing servations (e.g. features, natural-language statements, rules) has produced such a proliferation of techniques that current-generation evidence accrual suffers from the YARC ( syndrome. The proposed project addresses this problem by developing a unified, theoretically defensible, and computationally tractable paradigm that will ultimately permit the simultaneous estimation of target numbers, identities, and kinematics based on the systematic accrual of all types of evidence, including: random point data, ambiguous obseravations, and image data. Our approach, which we call Bayesian Multisource, Multi-target, Multi-evidence Filtering (or Bayesian multi-filtering (BMF) for short), is a direct generalization of recursive Bayes-Markov nonlinear filtering (NLF) theory. BMF generalized NLF to: (1) multiple targets of unknown number, identity, and kinematics, observed by multiple sensors, including (2) sensors which collect image data; and (3) sources which collect ambiguous observations. Specific Phase I tasks are: (1) Develop or acquire simulated multi-source data for multi sensor sources. (2) Develop multi-source, multi-target, multi-evidene unified evidence accrual algortihms. (3) Test and evaluate the novel evidence accrual algorithms. (4) Statistically characterize the performance error of the novel algorithms. (5) Final report and Phase II recommendations. Phase II will further develop the chosen evidence accrual paradigm and evaluate it using appropriate metrics. The project team includes Dr. Ronald Mahler of Lockheed Martin and Dr. Anuj Srivastava, both originators of BMF algorithms. Lockheed Martin will provde technical and commercialization support in the application of BMF.
* information listed above is at the time of submission.