Learning-Aided Tracking  
     
 

AIT has pioneered on-the-fly learning methods to fingerprint targets from multi-modality sensor signatures. Utilizing both radar and imaging sensors, salient features are discovered that serve to fingerprint targets in the context of their surroundings. This new technology will enable long-duration tracking of multiple targets through complex urban environments, stitching together broken tracks, and reacquisition of targets emerging from obscuration and hide.