Gradient-based prediction refinement for video coding

    公开(公告)号:US11962796B2

    公开(公告)日:2024-04-16

    申请号:US16836013

    申请日:2020-03-31

    IPC分类号: H04N11/02 H04N19/513

    CPC分类号: H04N19/513

    摘要: This disclosure describes gradient-based prediction refinement. A video coder (e.g., video encoder or video decoder) determines one or more prediction blocks for inter-predicting a current block (e.g., based on one or more motion vectors for the current block). In gradient-based prediction refinement, the video coder modifies one or more samples of the prediction block based on various factors such as displacement in a horizontal direction, the horizontal gradient, a displacement in the vertical direction, and a vertical gradient. This disclosure provides for gradient-based prediction refinement where a precision level of the displacement (e.g., at least one of the horizontal or vertical displacement) is unified (e.g., the same) for different prediction modes.

    Early termination of motion vector refinement process in video coding

    公开(公告)号:US11792423B2

    公开(公告)日:2023-10-17

    申请号:US16952736

    申请日:2020-11-19

    摘要: A video coder is configured to determine bi-directional motion vectors of a current block of the video data and determine that a condition is satisfied with respect to the current block based on each component of the bi-directional motion vectors of the current block being less than a threshold value. The video coder is further configured to, based on the condition being satisfied with respect to the current block, early terminate application of a motion vector refinement process to the bi-directional motion vectors of the current block. The video coder is further configured to determine a prediction block for the current block based on the bi-directional motion vectors of the current block and reconstruct the current block based on the prediction block for the current block.

    SIGNALING OF TRIANGLE MERGE MODE INDEXES IN VIDEO CODING

    公开(公告)号:US20230262265A1

    公开(公告)日:2023-08-17

    申请号:US18308773

    申请日:2023-04-28

    摘要: A video decoder obtains a first triangle merging index syntax element specifying a first triangle merging candidate index. The first triangle merging candidate index indicates a first triangle merging candidate of a triangular shape-based motion compensation candidate list. The video decoder may determine whether the maximum number of triangle merging candidates is greater than 2. Based on the maximum number of triangle merging candidates not being greater than 2, the video decoder may infer that a second triangle merging candidate index indicates a second triangle merging candidate of the triangular shape-based motion compensation candidate list without obtaining any syntax element specifying the second triangle merging candidate index from the bitstream, the second triangle merging candidate being different from the first triangle merging candidate.