Automatic target detection using entropy optimized shared-weight neural networks

Standard shared-weight neural networks previously demonstrated inferior per formance to that of morphological shared-weight neural networks for automat ic target detection. Empirical analysis showed that entropy measures of the features generated by the standard shared-weight neural networks were cons istently lower than those generated by the morphological shared-weight neur al networks. Based on this observation, an entropy maximization term was ad ded to the standard shared-weight network objective function. In this paper , we present automatic target detection results for standard shared-weight neural network trained with and without the added entropy term.