Abstract: A picture partitioning method and apparatus are described. The method includes determining a split mode of a current node, where the current node includes a luma block and a chroma block. The method further includes determining, based on the split mode of the current node and a size of the current node, that the chroma block of the current node is not further split. In accordance with the chroma block of the current node not being further split, the method includes splitting the luma block of the current node based on the split mode of the current node. In accordance with the chroma block of the current node not being further split, the method may be used to split only the luma block of the current node, thereby improving encoding and decoding efficiency, reducing a maximum throughput of a codec, and facilitating implementation of the codec.

Abstract: A method for encoding the region enhancement layer includes obtaining reconstructed pixels of a base layer of a target region; inputting the reconstructed pixels into a correction network to obtain correction information of the target region; inputting the correction information and original pixels of the target region into an encoding network to obtain a residual feature map of an enhancement layer of the target region; and encoding the residual feature map to obtain an enhancement layer bitstream of the target region.

Abstract: Provided is a fringe line detection and phase unwrapping method and system based on an FL-Net convolutional neural network. The method includes: constructing the FL-Net convolutional neural network; detecting fringe lines of an input image by using the FL-Net convolutional neural network to obtain an image with detected fringe lines; performing circulation integral on the detected fringe lines to repair the detected fringe lines to thereby obtain an image with repaired fringe lines; performing path integral on the repaired fringe lines to unwrap the repaired fringe lines to thereby obtain an image with unwrapped fringe lines; and identifying error points of the unwrapped fringe lines by using the FL-Net convolutional neural network; and processing the error points. Specifically, an HDC method is used to replace a down-sampling method to avoid the loss of resolution, and residual connection is used to prevent the network from being too deep.

Abstract: Technologies for inter prediction for a block in a video are provided. In one example, a method includes: when a size of a current block satisfies a preset condition, parsing a bitstream to obtain a syntax element, where the syntax element includes at least an index of optimal candidate motion information of the current block; determining, based on the syntax element, to construct a first list or a second list for the current block, where the first list is a block-based candidate motion information list, and the second list is a subblock-based candidate motion information list; determining optimal motion information from the first list or the second list based on the index; and predicting the current block based on the optimal motion information. In this method, the block-based candidate motion information list and the subblock-based candidate motion information list are distinguished from each other. This effectively reduces transmission costs of the optimal candidate motion information.

Abstract: A video picture prediction method and apparatus are provided, to provide a manner of determining a maximum length of a candidate motion vector list corresponding to a subblock merge mode. The method comprises: parsing a first indicator from a bitstream; if the first indicator indicates that a candidate mode used to inter predict the to-be-processed block comprises an affine mode, parsing a second indicator from the bitstream, where the second indicator is used to indicate a maximum length of a first candidate motion vector list, and the first candidate motion vector list is constructed for the to-be-processed block, a subblock merge prediction mode is used for the to-be-processed block; and determining the maximum length of the first candidate motion vector list based on the second indicator.

Abstract: Inter prediction method and apparatus are disclosed to resolve problems in the conventional technology that prediction samples obtained in an inter prediction mode are spatially discontinuous, prediction efficiency is affected, and prediction residual energy is relatively high. The method includes: parsing a bitstream to obtain motion information of a to-be-processed picture block; performing motion compensation on the to-be-processed picture block based on the obtained motion information, to obtain a prediction block of the to-be-processed picture block, where the prediction block of the to-be-processed picture block includes a prediction value of a target sample; and performing weighting calculation on one or more reconstructed values of one or more reference samples and the prediction value of the target sample, to update the prediction value of the target sample, where the reference sample(s) have a preset spatial position relationship with the target sample.

Abstract: A picture prediction method is provided, which includes: obtaining initial motion information of a current picture block; determining, based on a matching cost criterion, that positions of a pair of reference blocks are a position of a target forward reference block of the current picture block and a position of a target backward reference block of the current picture block, where positions of each pair of reference blocks include a position of a forward reference block and a position of a backward reference block; and for the positions of each pair of reference blocks, a first position offset and a second position offset are in a mirror relationship; and obtaining a predicted value of a pixel value of the current picture block based on a pixel value of the target forward reference block and a pixel value of the target backward reference block.

Abstract: Disclosed are an inter prediction method based on an affine prediction mode and a related apparatus. The method includes: obtaining a predicted pixel value of a current affine block; obtaining an LIC flag of the current affine block, where the LIC flag is used to indicate whether local illumination compensation LIC is applied to the current affine block; and when the LIC flag indicates that LIC is applied to the current affine block, updating the predicted pixel value of the current affine block based on an update coefficient for LIC, to obtain an updated predicted pixel value of the current affine block.

Abstract: An image prediction method and device are disclosed. The method includes: obtaining a first reference unit of an image unit, where respective predicted images are obtained for the image unit and the first reference unit by using a same affine model; obtaining motion information of basic motion compensation units at two or more preset positions of the first reference unit; and obtaining motion information of a basic motion compensation unit of the image unit according to the motion information. In this way, motion information of the first reference unit using a same affine motion prediction model is reused, and a more accurate motion vector of a current image unit is obtained, improving prediction accuracy and maintaining encoding and decoding complexity, thereby improving encoding and decoding performance.

Abstract: Techniques are described of a linear model prediction mode. A set of reconstructed neighboring luma samples are determined. Two pairs of luma value and chroma value are determined according to N reconstructed neighboring luma samples, N reconstructed neighboring chroma samples, M reconstructed neighboring luma samples, and M reconstructed neighboring chroma samples. The minimum value of the N reconstructed neighboring luma samples is not less than the maximum luma value of the remaining reconstructed neighboring luma samples of the set of reconstructed neighboring luma samples. The maximum value of the M reconstructed neighboring luma samples is not larger than the minimum luma value of the remaining reconstructed neighboring luma samples of the set of reconstructed neighboring luma samples. M>1 and N>1. Linear model parameters are determined based on the two pairs of luma value and chroma value, and a predictive block is determined based on the linear model parameters.

Abstract: A prediction method includes: performing downsampling on a luma component of a to-be-processed image block to obtain an initial signal of the luma component; obtaining a residual signal of the luma component based on the obtained initial signal of the luma component; obtaining a residual signal of a chroma component of the to-be-processed image block based on the chroma component of the to-be-processed image block without downsampling. The method further includes: respectively performing a transform on the residual signal of the luma component and the residual signal of the chroma component, to obtain transform coefficients, performing quantization on the obtained transform coefficients, to obtain quantization coefficients, and performing entropy encoding on the obtained quantization coefficients to obtain a bitstream.

Abstract: Embodiments of this application disclose a decoding method includes: obtaining a bitstream including picture data; parsing the bitstream to obtain node split mode information of a first-level coding tree and node split mode information of a second-level coding tree, if the split mode corresponding to the first node is no further splitting, parsing the bitstream to obtain encoding information of the first node; and decoding and reconstructing, based on the encoding information of the first node, a coding unit corresponding to the first node, to obtain a picture corresponding to the picture data.

Abstract: The invention provide a rendering method and an electronic device. The method includes: first, obtaining channel feedback information of an encoded transmission channel, where the encoded transmission channel is used to transmit an encoded bitstream corresponding to a rendered image; subsequently, determining, from a plurality of groups of preset rendering input information, rendering input information corresponding to the channel feedback information, where complexity of the plurality of groups of preset rendering input information is different; then, performing rendering based on the rendering input information corresponding to the channel feedback information, to obtain the rendered image, where content complexity of the obtained rendered image is associated with complexity of the rendering input information corresponding to the channel feedback information.

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: Image encoding and decoding methods, apparatuses and computer-readable media are described. An example image encoding method includes: determining N encoded units from adjacent encoded units of a to-be-encoded unit according to a first preset rule, where a motion prediction mode of the N encoded units is the same as that of the to-be-encoded unit; generating an nth motion vector group by using a first preset algorithm and based on a motion vector of an nth encoded unit; determining, from obtained N motion vector groups according to a second preset rule, one motion vector group as an optimal motion vector group; determining a prediction sample value of each sample unit in the to-be-encoded unit by using a second preset algorithm and based on the optimal motion vector group; and encoding a prediction residual of each sample unit and an index identifier of the optimal motion vector group, to obtain a bitstream corresponding to the to-be-encoded unit.

Abstract: Embodiments of this application disclose an interpolation filter training method and apparatus, a video picture encoding and decoding method, an encoder, and a decoder. According to the training method, a first sub-pixel picture obtained through interpolation by using a conventional interpolation filter is used as label data, to train a second interpolation filter, so that the second interpolation filter obtained through training can be directly used for a pixel value, obtained through interpolation, of a first fractional pixel position. Therefore, the label data is more accurate, and coding performance of a video picture is improved. According to the encoding method, during inter prediction, a target interpolation filter used for a current encoding picture block is determined from a set of candidate interpolation filters, and the encoder selects, according to content of the current encoding picture block, an appropriate interpolation filter to perform an interpolation operation.

Abstract: The present disclosure provides an image decoding method based on a wavelet transform. The method includes: performing entropy decoding on the compressed bitstream to obtain a reconstructed wavelet coefficient; dequantizing the reconstructed wavelet coefficient to obtain a dequantized wavelet coefficient; and performing N inverse wavelet transforms on the dequantized wavelet coefficient based on the wavelet transform model to obtain a reconstructed image, where the wavelet transform model is implemented based on a convolutional neural network (CNN).

Abstract: The present application discloses example encoding methods and apparatuses, and decoding methods and apparatuses, and relate to the field of media technologies. One example method includes obtaining a bitstream, and obtaining a plurality of data packets based on the bitstream. Based on identifiers of the plurality of data packets, the plurality of data packets is sent to a plurality of entropy decoders for decoding, to obtain a plurality of syntax elements. Media content is restored based on the plurality of syntax elements.

Abstract: A video processing method includes: obtaining prediction information of a CU; obtaining, when the CU comprises only one residual TU and a size of the residual TU is less than a size of the CU, a TU partitioning mode of the CU and a residual position of the residual TU, wherein the TU partitioning mode and the residual position are used to determine a horizontal transform type and a vertical transform type; obtaining transform coefficients of the residual TU based on the horizontal transform type and the vertical transform type; and generating a bitstream that is for storing or transmitting and that includes the prediction information, a first flag that indicates the TU partitioning mode, a second flag that indicates the residual position, and the transform coefficients.

Abstract: A video encoder, a video decoder, and a corresponding method are provided. The method includes: parsing a bitstream to obtain an index, where the index indicates a target candidate motion vector group of a current coding block; determining the target candidate motion vector group in an affine candidate motion vector list based on the index, where the affine candidate motion vector list includes at least a first candidate motion vector group, the first candidate motion vector group is obtained based on a first group of control points of a first neighboring affine coding block, and the first group of control points is determined based on a CTU located relative to the current coding block, wherein the first neighboring affine coding block is located in the CTU; and predicting a predicted sample value of the current coding block based on the target candidate motion vector group.