Contextual Weighting of Patches for Local Matching in Still-to-Video Face Recognition | IEEE Conference Publication | IEEE Xplore