Loading patent…
AssigneeHuawei Technologies Co., Ltd.
InventorsSriram Sethuraman, Sagar Kotecha, Jeeva Raj A, Semih Esenlik
Estimating the term…
It is provided a method of decoding a video bitstream implemented by a decoding device, the video bitstream including coded data for a plurality of pictures. The decoding method comprises: parsing an enabling flag for decoder motion vector refinement (DMVR) from a sequence parameter set of the video bitstream; parsing a disabling flag for DMVR from a syntax structure of a lower hierarchic syntax structure level than the sequence parameter set (SPS) level of the video bitstream; and performing DMVR based inter bi-prediction for a current block within at least one region of the current picture to obtain predicted sample values of the current block when at least one pre-defined condition is fulfilled, and wherein the at least one pre-defined condition comprises that the parsed value of the disabling flag specifies that DMVR based inter bi-prediction is enabled for at least one region of the current picture.
A method of coding a video bitstream comprising coded data for a plurality of pictures implemented in a coding device, the method comprising:
A device for video coding, comprising:
A non-transitory storage medium which includes an encoded bitstream decoded by an image decoding device, the encoded bitstream being generated by determining whether bi-directional optical flow (BDOF) inter prediction is enabled at a sequence parameter set (SPS) level;
Dependencies are inferred from the claim text.