ATR Using Multi-view Morphing
Mubarak A. Shah · 2005
Abstract : In 2005, reviewed current research literature in view morphing and matching for single views, and discussed a feature based approach to sparse view morphing that also shows promise for compression of the model data corpus. Worked on a new Bayesian approach for the problem from a mid-level segmentation problem. We obtained good results for feature-based morphing, as well as preliminary demonstration of the novel object representation. In 2006, we completed work in automatic target recognition using video. Introduced a novel approach for adaptive video.