I. Introduction
An object audio coding that independently processes audio objects has been receiving the increasing interest in users' audio systems and intelligent multimedia. Desired object can be controlled by users in music, news and other multi-media service. Users are authorized to design the listening sound of multimedia service. Currently, spatial audio object coding (SAOC), which is based on audio down-mixing technology, has been standardizing in moving picture experts group (MPEG) [1]. SAOC has a high compression ratio due to the audio down-mixing technology, but an up-mixed object receives a little interference from other objects. The object that is selected to turn off at an object audio system cannot be perfectly removed in music by the interference.