Automated Entity Classification in Video Using Soft Biometrics

With the growing concerns surrounding security and terrorism around the world, biometrics has become one of the premier solutions to combat these problems. Today, even the most sophisticated identification algorithms can take up to 10 seconds to accurately identify a single subject due to the massive size of realistic databases (~1 million subjects). In this proposal, we will discuss a detailed plan to develop a prototype application capable of automatically extracting “soft biometric” feature information from video imagery for identification and subject indexing. To attain the highest performance at the lowest risk, we plan to partner with renowned domain experts at the University of Notre Dame and West Virginia University to leverage cutting edge biometric algorithms and techniques. To ensure success in developing a prototype application with a modular biometric feature framework, we will leverage a partnership with fellow Phase I award winner intuVision, that will allow us to base many of our GUI designs on their successful commercial products. At the conclusion of the Phase II and Phase II Option research and development tasks, we plan to deliver a fully functional prototype that will illuminate a path for successful Phase III integration.