Inter-prediction method and device based on DMVR and BDOF

    公开(公告)号:US11936878B2

    公开(公告)日:2024-03-19

    申请号:US17959001

    申请日:2022-10-03

    Abstract: An image decoding method includes: deriving a refined L0 motion vector and a refined L1 motion vector by applying decoder-side motion vector refinement (DMVR) based on an L0 motion vector and an L1 motion vector of a current block, deriving prediction samples for the current block based on L0 prediction samples derived based on the refined L0 motion vector and L1 prediction samples derived based on the refined L1 motion vector, and generating reconstructed samples for the current block based on the prediction samples. The L0 prediction samples are derived based on an L0 reference picture and the refined L0 motion vector, and the L1 prediction samples are derived based on an L1 reference picture and the refined L1 motion vector, and deriving the prediction samples comprises determining whether to apply bi-directional optical flow (BDOF) process based on a first POC difference and a second POC difference are the same.

    Method and apparatus for processing video data

    公开(公告)号:US11876957B2

    公开(公告)日:2024-01-16

    申请号:US17413419

    申请日:2019-12-18

    Abstract: A video processing method according to an embodiment of the present disclosure includes obtaining motion information and bi-directional prediction weight information for inter prediction of a current block, deriving a prediction sample of the current block based on the motion information and the bi-directional prediction weight information, and updating an HMVP candidate list based on a history-based motion vector predictor (HMVP) candidate including the motion information and the bi-directional prediction weight information. The motion information includes a first reference index and a first motion vector related to a first reference picture and a second reference index and a second motion vector related to a second reference picture. The bi-directional prediction weight information may be used to derive a first weight value applied to a first prediction sample from the first reference picture and a second weight value applied to a second prediction sample from the second reference picture.

    SbTMVP-based inter prediction method and apparatus

    公开(公告)号:US11800089B2

    公开(公告)日:2023-10-24

    申请号:US17414893

    申请日:2020-01-02

    Abstract: A method by which a decoding device performs image decoding, according to the present document, comprises the steps of: deriving a reference block on the basis of a motion shift; determining whether a subblock temporal motion vector predictor (sbTMVP) candidate is available, on the basis of whether the reference block is a current picture referencing (CPR) block; deriving motion information about a current block on the basis of the sbTMVP candidate when the sbTMVP candidate is available; generating prediction samples on the basis of the motion information; and generating a restored picture on the basis of the prediction samples, wherein the sbTMVP candidate is determined to be unavailable with respect to the reference block which is the CPR block.

Patent Agency Ranking