Abstract: An image prediction method, apparatus, and system, a device, and a storage medium are provided. The method includes: (401) obtaining a split mode of a current node, where the current node is an image block in a coding tree unit in a current image; (402) determining, based on the split mode of the current node and a size of the current node, whether the current node satisfies a first condition; and (403) when it is determined that the current node satisfies the first condition, performing intra prediction on all coding blocks belonging to the current node, to obtain predictors of all the coding blocks belonging to the current node.

Abstract: This application relates to the field of artificial intelligence technologies, and provides a feature map processing method and a related device. The method is implemented by invoking a neural network model that includes a plurality of input adaptation branches and a post-processing part. An output of each of the plurality of input adaptation branches is an input of the post-processing part, and each of the plurality of input adaptation branches can downsample a feature map at a different ratio.

Abstract: A picture encoding/decoding method and a related apparatus are provided. The picture decoding method includes obtaining a current picture from a video, the current picture being segmented using random access points into multiple video segments supporting a random access function; selecting, from a knowledge base, K reference pictures of the current picture, wherein at least one reference picture in the knowledge base does not belong to a random access segment in which the current picture is located, the random access segment comprising a picture sequence arranged in a decoding order from a closest random access point before the current picture to a closest random access point after the current picture; and decoding the current picture according to the K reference pictures.

Abstract: This application provides a motion vector obtaining method and apparatus. The method includes: determining a target offset vector of a block and identifier information of a target picture, wherein the block comprises at least one sub-block; determining a location of the sub-block; determining, as a target location coordinate value of a collocated sub-block, a location coordinate value obtained by performing a clipping operation on an initial location coordinate value in a range, wherein the initial location coordinate value is based on the location of the sub-block and the target offset vector; and obtaining a motion vector of the sub-block based on a motion vector corresponding to the target location coordinate value. Thus, a range of the target offset vector is limited, so that a quantity of memory read times can be reduced in a process of obtaining the motion vector of the collocated sub-block.

Abstract: A deblocking method, for deblocking a transform block boundary within a coding block in an image encoding and/or an image decoding is provided, wherein the coding block comprising transform blocks is coded in inter prediction mode and the transform blocks comprises a first transform block and a second transform block. The method comprises: determining, when the boundary between the first transform block and the second transform block is a transform block boundary and at least one of the first transform block and the second transform block has one or more non-zero transform coefficients, a value of a BS parameter for the boundary between the first transform block and the second transform block to be a first value; and performing de-blocking filtering process to the boundary between the first transform block and the second transform block at least based on the value of the boundary strength parameter.

Abstract: A video decoding method, including: parsing a received bitstream to obtain prediction information of a coding unit (CU); obtaining a target transform mode of a residual transform unit (TU); parsing the received bitstream to obtain transform coefficients of the residual TU; applying an inverse quantization to the transform coefficients of the residual TU to obtain dequantized coefficients; applying, based on the target transform mode, an inverse transform to the dequantized coefficients to obtain a residual block of the residual TU; obtaining a prediction block of the CU based on the prediction information; obtaining a video block based on the residual block and the prediction block; and outputting a video sequence including a video frame that includes the video block.

Abstract: An intra prediction method by using cross component liner prediction mode (CCLM), includes: determining a luma block corresponding to a current chroma block; obtaining luma reference samples of the luma block based on determining L available chroma reference samples of the current chroma block, wherein the obtained luma reference samples of the luma block are down-sampled luma reference samples; calculating linear model coefficients based on the luma reference samples and chroma reference samples that correspond to the luma reference samples; and obtaining a prediction for the current chroma block based on the linear model coefficients and values of a down-sampled luma block of the luma block.

Abstract: This application provides picture or audio encoding and decoding methods and apparatuses, and relates to the field of artificial intelligence (AI)—based picture or audio encoding and decoding technologies, and specifically, to the field of neural network-based picture feature map or audio feature variable encoding and decoding technologies. The encoding method includes: obtaining a to-be-encoded target, where the to-be-encoded target includes a plurality of feature elements, and the plurality of feature elements include a first feature element. The method further includes: obtaining a probability estimation result of the first feature element; determining, based on the probability estimation result of the first feature element, whether to perform entropy encoding on the first feature element; and performing entropy encoding on the first feature element only when it is determined that entropy encoding needs to be performed on the first feature element.

Abstract: This application provides a chroma block prediction method and apparatus. The method includes: obtaining chroma values of chroma samples; obtaining, luma values of luma samples corresponding to the chroma samples; obtaining, from the luma values, a first luma set and a second luma set; grouping the chroma values into a first chroma set and a second chroma set; determining a scaling coefficient in a linear model based on an average value of luma values in the first luma set, an average value of luma values in the second luma set, an average value of chroma values in the first chroma set, and an average value of chroma values in the second chroma set; determining, based on the scaling coefficient, an offset factor in the linear model; and determining prediction information of the chroma block based on the scaling coefficient and the offset factor.

Abstract: Embodiments include a picture reconstruction method and apparatus for a video picture. The picture reconstruction method includes: obtaining a prediction mode of a current coding unit, and/or obtaining a prediction partition mode of the current coding unit, where the current coding unit includes a luma coding block and a chroma coding block, and the prediction partition mode is a mode of splitting the current coding unit into prediction blocks or prediction units; obtaining a transform block of the current coding unit based on the prediction partition mode and/or the prediction mode; and generating a reconstructed picture block of the current coding unit based on the transform block. According to the method in the embodiments, video coding efficiency can be improved.

Abstract: A mechanism for position dependent spatial varying transform (SVT) for video coding. A prediction block and a corresponding transformed residual block are received at a decoder. A type of spatial varying transform (SVT) employed to generate the transformed residual block is determined. A position of the SVT relative to the transformed residual block is also determined. An inverse of the SVT is applied to the transformed residual block to reconstruct a reconstructed residual block. The reconstructed residual block is then combined with the prediction block to reconstruct an image block.

Abstract: A method includes parsing a transform coefficient of a transform block in a current coding unit to obtain a first transform coefficient matrix. The method also includes obtaining a quantity K of non-zero transform coefficients in a top-left preset region of the first transform coefficient matrix. The method also includes parsing an index value based on the quantity K being greater than a threshold. The method also includes determining a transform matrix based on the index value. The method also includes multiplying N transform coefficients of the first transform coefficient matrix by the transform matrix to obtain M transform coefficients. The method also includes updating the first transform coefficient matrix by using the M transform coefficients to obtain a second transform coefficient matrix. The method also includes performing an inverse transform on the second transform coefficient matrix to obtain residual samples of the current coding unit.

Abstract: This application discloses a processing method applied to video encoding or video decoding. The method includes: obtaining size information of a current block; determining a context model index ctxInc of a flag of the current block based on the obtained size information of the current block and a size-based constraint; and performing subsequent processing based on the context model index ctxInc of the flag of the current block. The method can be used to improve coding efficiency.

Abstract: A prediction mode determining method includes determining a split mode for a current picture block, determining whether a first picture subblock that meets a preset condition is obtained after the current picture block is split in the split mode, splitting the current picture block in the split mode to obtain a plurality of picture subblocks in response to determining that the first picture subblock is obtained after the current picture block is split, and determining that a same prediction mode is used for the plurality of picture subblocks. The plurality of picture subblocks comprise the first picture subblock. The prediction mode for the plurality of picture subblocks is an intra prediction mode or an inter prediction mode.

Abstract: A method of decoding a coded video bitstream implemented by a video decoder. The method includes determining a split mode for a coding tree unit (CTU) in an encoded bitstream, where the split mode is one of triple tree (TT), vertical binary tree (BT), and horizontal BT; marking a coding unit (CU) from the CTU as not to be predicted by intra prediction when one of three conditions is met; parsing the encoded bitstream to obtain a prediction unit and residual information for the CU from the CTU; and reconstructing the CU based on the prediction unit and the residual information when the CU has been marked.

Abstract: A picture encoding/decoding method and a related apparatus are provided. The picture decoding method includes obtaining a current picture from a video, the current picture being segmented using random access points into multiple video segments supporting a random access function; selecting, from a knowledge base, K reference pictures of the current picture, wherein at least one reference picture in the knowledge base does not belong to a random access segment in which the current picture is located, the random access segment comprising a picture sequence arranged in a decoding order from a closest random access point before the current picture to a closest random access point after the current picture; and decoding the current picture according to the K reference pictures.

Abstract: An image processing method includes: obtains a to-be-processed image, performs feature extraction on the to-be-processed image by using a feature extraction layer included in a convolutional neural network, to obtain a first feature map, and compresses the first feature map by using a feature compression layer included in the convolutional neural network, to obtain a second feature map, where a channel quantity of the second feature map is less than a channel quantity of the first feature map; and the sending node sends the second feature map to a receiving node. The first feature map of the to-be-processed image is obtained through extraction by using the feature extraction layer, and the first feature map is compressed by using the feature compression layer, to obtain the second feature map.

Abstract: An image prediction method, apparatus, and system, a device, and a storage medium are provided. The method includes: (401) obtaining a split mode of a current node, where the current node is an image block in a coding tree unit in a current image; (402) determining, based on the split mode of the current node and a size of the current node, whether the current node satisfies a first condition; and (403) when it is determined that the current node satisfies the first condition, performing intra prediction on all coding blocks belonging to the current node, to obtain predictors of all the coding blocks belonging to the current node.

Abstract: A deblocking method, for deblocking a transform block boundary within a coding block in an image encoding and/or an image decoding is provided, wherein the coding block comprising transform blocks is coded in inter prediction mode and the transform blocks comprises a first transform block and a second transform block; wherein the method comprises: determining, when the boundary between the first transform block and the second transform block is a transform block boundary and at least one of the first transform block and the second transform block has one or more non-zero transform coefficients, a value of a BS parameter for the boundary between the first transform block and the second transform block to be a first value; and performing de-blocking filtering process to the boundary between the first transform block and the second transform block at least based on the value of the boundary strength parameter.

Abstract: Apparatuses and methods for Intra Sub-Partitions (ISP) are provided. The method includes: obtaining information of ISP, and determining a size of a chroma transform block (TB) of the coding unit based on SubWidthC and SubHeightC when at least a first condition is fulfilled. The first condition includes the information of ISP indicates that ISP is used for splitting the luma coding block. SubWidthC and SubHeightC are variables depending on chroma format information. The chroma format information indicates a chroma format of a picture the coding unit belongs to. The method can apply to all chroma format. The chroma format includes at least one of: 4:2:0, or 4:2:2, or 4:4:4. An accurate and versatile way to determine the size of the chroma transform block for the ISP is achieved.