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