Refinement of extracted visual attention areas in video sequences | IEEE Conference Publication | IEEE Xplore