The conventional multiview video coding (MVC) schemes, utilizing both neighboring temporal fr a mes and view fr a mes as possible references, have only shown a slight gain over those using temporal fr a mes alone in terms of coding efficiency.