Information theoretic feature extraction for ATR

Utilizing principles of information theory, nonparametric statistics and machine learning we describe a task-driven feature extraction approach. Specifically, the features preserve information related to the specific estimation problem. Mutual information, motivated by Fano´s inequality, is the criterion used for feature extraction. The novelty of our approach is that we optimize mutual information in the feature space (thereby avoiding the curse of dimensionality) and we do so without explicit estimation or modeling of the underlying density. We present experimental results for pose estimation of high-resolution SAR imagery.