熵编解码方法和装置
    1.
    发明申请

    公开(公告)号:WO2023020320A1

    公开(公告)日:2023-02-23

    申请号:PCT/CN2022/110827

    申请日:2022-08-08

    发明人: 郭天生 王晶

    IPC分类号: H04N19/13 H04N19/42

    摘要: 本申请提供一种熵编解码方法和装置。本申请熵编码方法,包括:获取待编码数据,所述待编码数据是当前数据流包含的多个数据中非首位编码的数据;获取参照信息,所述参照信息至少包括第一上下文信息和第一边信息中的至少一项,所述第一上下文信息是将至少一个已编码数据输入自注意力解码网络得到的,所述第一边信息是将所述多个数据输入自注意力编码网络得到的;根据所述参照信息估计得到第一估计概率分布;根据所述第一估计概率分布对所述待编码数据进行熵编码,以得到第一码流。本申请可以提高当前数据流包含的多个数据的传输效率。

    영상 부호화/복호화 방법, 비트스트림을 전송하는 방법 및 비트스트림을 저장한 기록 매체

    公开(公告)号:WO2023014076A1

    公开(公告)日:2023-02-09

    申请号:PCT/KR2022/011468

    申请日:2022-08-03

    摘要: 영상 부호화/복호화 방법, 비트스트림 전송 방법 및 비트스트림을 저장하는 컴퓨터 판독 가능한 기록 매체가 제공된다. 본 개시에 따른 영상 복호화 방법은, 영상 복호화 장치에 의해 수행되는 영상 복호화 방법으로서, 디블록킹(deblocking) 필터링을 복원 샘플들에 적용하는 단계; 및 SAO(sample adaptive offset) 필터링 또는 CC-SAO(cross component sample adaptive offset) 필터링 중에서 하나 이상을 상기 디블록킹 필터링이 적용된 복원 샘플들에 적용하여, 수정된 루마 샘플 및 수정된 크로마 샘플들을 유도하는 단계를 포함하고, 상기 수정된 루마 샘플은, 상기 SAO 필터링을 상기 디블록킹 필터링이 적용된 루마 샘플에 적용하여 유도되며, 상기 수정된 크로마 샘플들은, 상기 SAO 필터링을 상기 디블록킹 필터링이 적용된 크로마 샘플들에 적용하고, 상기 CC-SAO 필터링을 상기 디블록킹 필터링이 적용된 크로마 샘플들 및 제1인루프 필터링이 적용된 루마 샘플에 적용하여 유도되는 영상 복호화 방법일 수 있다.

    数据处理的方法和系统
    3.
    发明申请

    公开(公告)号:WO2022247735A1

    公开(公告)日:2022-12-01

    申请号:PCT/CN2022/094004

    申请日:2022-05-20

    申请人: 于江鸿

    摘要: 本说明书提供的数据处理的方法和系统,可以将初始帧划分为多个单元,并使用不同的边界调节系数对每个单元的中频至高频区域的幅值进行调节。若当前单元中的中频至高频区域的幅值较大时,使用小于1大于0的边界调节系数以降低中频至高频区域的幅值,从而减少数据信息量,提高数据压缩的效率。若当前单元中的中频至高频区域的幅值较小时,则使用大于1的边界调节系数以增强中频至高频区域的幅值,避免细节损失。在数据解压时,所述方法和系统能够使用与边界调节系数对应的边界补偿系数对每个单元的中频至高频区域的幅值进行边界补偿,使经过所述边界调节的压缩数据恢复至初始帧的清晰度甚至高于所述初始帧的清晰度,得到解压帧。

    SECURE DECODER AND SECURE DECODING METHODS
    4.
    发明申请

    公开(公告)号:WO2022243672A1

    公开(公告)日:2022-11-24

    申请号:PCT/GB2022/051238

    申请日:2022-05-17

    发明人: CLUCAS, Richard

    摘要: A module having access to a secure memory and an unsecure memory, the module configured to: read, from the secure memory via a secure channel, a decrypted decoded rendition of a base layer of a video signal, wherein the decrypted decoded rendition of the base layer has a first resolution; read, from the unsecure memory, a decoded rendition of an enhancement layer of the video signal, wherein decoded rendition of the enhancement layer has a second resolution, wherein the second resolution is higher than the first resolution; generate an upsampled decrypted decoded rendition of the base layer of the video signal by upsampling the decrypted decoded rendition of the base layer such that the upsampled decrypted decoded rendition of the base layer has the second resolution; apply the decoded rendition of the enhancement layer to the upsampled decrypted decoded rendition of the base layer to generate an output plane; output the output plane, via the secured channel, to the secure memory for storage in the secure memory, wherein the output plane is for use in outputting by an output module to a display.

    METHODS AND SYSTEMS FOR IMAGE COMPENSATION
    5.
    发明申请

    公开(公告)号:WO2022174685A1

    公开(公告)日:2022-08-25

    申请号:PCT/CN2021/143872

    申请日:2021-12-31

    摘要: The present disclosure relates to systems and methods for image compensation. The systems may obtain a reconstructed image frame. The systems may determine a primary classification manner of pixels in the reconstructed image frame. The systems may determine a primary compensation value of each category of pixels in the reconstructed image frame that are classified based on the primary classification manner. The systems may obtain a compensated image frame by compensating the reconstructed image frame based on the primary compensation value of each category of pixels in the reconstructed image frame.

    解码方法、解码设备及可读存储介质

    公开(公告)号:WO2022156113A1

    公开(公告)日:2022-07-28

    申请号:PCT/CN2021/097135

    申请日:2021-05-31

    发明人: 刘金风

    IPC分类号: H04N19/44 H04N19/42 H04L1/00

    摘要: 本申请公开了一种解码方法、解码设备及可读存储介质,该解码方法通过对第一比特流中对应的指定比特位进行简单的逻辑运算,据此生成对应的第四比特流,可以得到编码前的信息,该解码方式的逻辑设计简单,既可以降低逻辑电路设计的复杂度,又可以提高解码的可靠性。

    视频编解码以及模型训练方法与装置

    公开(公告)号:WO2022155974A1

    公开(公告)日:2022-07-28

    申请号:PCT/CN2021/073669

    申请日:2021-01-25

    发明人: 马展 夏琪 刘浩杰

    IPC分类号: H04N19/42

    摘要: 本申请提供一种视频编解码以及模型训练方法与装置,通过将图像中目标物区域和背景区域的特征图分离开,分别编码成不同码率的码流;直接对目标物的特征图进行智能分析,省去解码过程,实现高效的面向人机混合视觉的任务,提高了编码效率,降低了解码复杂性。

    CHROMA FROM LUMA PREDICTION FOR VIDEO CODING

    公开(公告)号:WO2022125180A1

    公开(公告)日:2022-06-16

    申请号:PCT/US2021/054223

    申请日:2021-10-08

    摘要: A video coder may be configured to code a chroma block of video data as a function of reconstructed luma samples. The video coder may store, during the prediction process for a luma block, a sum of the reconstructed luma sample values for each sub-block of the luma block in a first buffer, and store, during the prediction process for the luma block, an average of the reconstructed luma sample values in a second buffer. The video coder may perform chroma from luma prediction to reconstruct a chroma block of the video data corresponding to the luma block using the sums and the average.

    点云媒体的数据处理方法、装置、设备及存储介质

    公开(公告)号:WO2022111348A1

    公开(公告)日:2022-06-02

    申请号:PCT/CN2021/131101

    申请日:2021-11-17

    发明人: 胡颖

    IPC分类号: H04N19/37 H04N19/42

    摘要: 本申请实施例提供了一种点云媒体的数据处理方法、装置、设备及存储介质,该方法包括:获取点云媒体的第i个属性成分的指示信息,该点云媒体包括N个属性成分,该第i个属性成分为该N个属性成分中的任一个,该第i个属性成分的指示信息用于指示该第i个属性成分的必要性和优先级中的至少一项,N、i均为正整数且i∈[1,N];基于该第i个属性成分的指示信息解析该第i个属性成分。该方法涉及点云媒体技术领域,该方法通过对点云媒体的属性成分的必要性和优先级进行指示,能够在一定程度上提升点云媒体的解析处理效率。

    HIGH LEVEL SYNTAX IN PICTURE HEADER
    10.
    发明申请

    公开(公告)号:WO2021202391A1

    公开(公告)日:2021-10-07

    申请号:PCT/US2021/024672

    申请日:2021-03-29

    申请人: BYTEDANCE INC.

    IPC分类号: H04N7/12 H04N19/42

    摘要: Methods and apparatus for video processing are described. The video processing may include video encoding, video decoding or video transcoding. One example video processing method includes performing a conversion between a video comprising a picture comprising multiple slices and a bitstream of the video. The bitstream conforms to a format rule that specifies that a flag in a picture header controls applicability of adaptive loop filtering of all slices in the picture.