Context initialization based on decoder picture buffer
View Patent ↗A method for decoding a video frame includes receiving a slice header of at least one of a plurality of slices, identifying a type of a first slice and a type of a second slice, and initializing a first context associated with the first slice and a second context associated with the second slice using one of a first initialization method and a second initialization method. The first initialization method or the second initialization method is selected according to the type of the first slice and the type of the second slice.
1. A method for decoding a video frame of a video sequence, the method comprising:
(a) receiving a slice;
(b) receiving a flag for determining one of three initialization methods; and
(c) initializing a Context-Adaptive Binary Arithmetic Coding (CABAC) context associated with a slice type using an initialization method; wherein
in a case that the slice type indicates a B slice type and a value of the flag is equal to a first value, a CABAC context associated with the B slice type is initialized by a first initialization method of the three initialization methods; and
in a case that the slice type indicates the B slice type and the value of the flag is equal to a second value, the CABAC context associated with the B slice type is initialized by a second initialization method of the three initialization methods, the second initialization method being used to initialize a CABAC context associated with a P slice type in a case that the slice type indicates the P slice type and the value of the flag is equal to the first value.
2. The method of claim 1 , wherein the third initialization method differs from the first initialization method and the second initialization method.
3. The method of claim 1 , the method further comprising:
receiving information including at least (1) first information indicating a number of first regions in a slice and (ii) second information indicating offset of an i-th first region from an (i−1)th first region in a bitstream, wherein the video frame includes one or more of the slice(s) and one or more of the first region(s).