Intelligent Fusion of Multi-modal Biometric Inputs for Decision Making
While techniques provide some beneficial results, they do not effectively utilize all of the information available to a score-level fusion system, including score values, quality estimates, and score distribution statistics.
APL's Process for Multi-Biometric Fusion provides expandable, Bayesian belief network-based architecture for multimodal biometric fusion. Studies have been carried out using the architecture to fuse fingerprints and voice data. This invention is unique because: 1) the method fuses information in a manner that optimizes matching of poor quality biometric data and 2) the algorithm uses easily characterized parametric statistical distributions for match and non-match distributions, allowing more rapid training on smaller training sets.
Patent Status: U.S. patent(s) 7747044 issued.
*Reduced to practice; available for licensingCONTACT: