Abstract: A method and apparatus for inter prediction in video coding system are disclosed. According to the method, input data associated with a current block comprising at least one colour block are received. A blending predictor is determined according to a weighted sum of at least two candidate predictions generated based on one or more first hypotheses of prediction, one or more second hypotheses of prediction, or both. The first hypotheses of prediction are generated based on one or more intra prediction modes comprising a DC mode, a planar mode or at least one angular modes. The second hypotheses of prediction are generated based on one or more cross-component modes and a collocated block of said at least one colour block. The input data associated with the colour block is encoded or decoded using the blending predictor.

Abstract: A video coding system that uses multiple models to predict chroma samples is provided. The video coding system receives data for a block of pixels to be encoded or decoded as a current block of a current picture of a video. The system constructs two or more chroma prediction models based on luma and chroma samples neighboring the current block. The system applies the two or more chroma prediction models to incoming or reconstructed luma samples of the current block to produce two or more model predictions. The system computes predicted chroma samples by combining the two or more model predictions. The system uses the predicted chroma samples to reconstruct chroma samples of the current block or to encode the current block.

Abstract: Interconnect structures for front-to-front stacked chips/dies and methods of fabrication thereof are disclosed herein. An exemplary system on integrated circuit (SoIC) includes a first die that is front-to-front bonded with a second die, for example, by bonding a first topmost metallization layer of a first frontside multilayer interconnect of the first die to a second topmost metallization layer of a second frontside multilayer interconnect of the second die. A through via extends partially through the first frontside multilayer interconnect of the first die, through a device layer of the first die, through a backside power rail of the first die, and through a carrier substrate. The backside power rail is between the carrier substrate and the device layer, and the backside power rail may be a portion of a backside multilayer interconnect of the first die. The through via may be connected to a redistribution layer (RDL) structure.

Abstract: A video coding system that uses multiple models to predict chroma samples is provided. The video coding system receives data for a block of pixels to be encoded or decoded as a current block of a current picture of a video. The video coding system derives multiple prediction linear models based on luma and chroma samples neighboring the current block. The video coding system constructs a composite linear model based on the multiple prediction linear models. The video coding system applies the composite linear model to incoming or reconstructed luma samples of the current block to generate a chroma predictor of the current block. The video coding system uses the chroma predictor to reconstruct chroma samples of the current block or to encode the current block.

Abstract: A method includes forming a first multilayer interconnect structure over a first side of a device layer, forming a first portion of a second multilayer interconnect structure under a second side of the device layer, forming a trench that extends through the second dielectric layer, the device layer, and the first dielectric layer, forming a conductive structure in the trench, and forming a second portion of the second multilayer interconnect structure under the first portion of the second multilayer interconnect structure. The second portion of the second multilayer interconnect structure includes patterned metal layers disposed in a third dielectric layer, and wherein one or more of the patterned metal layers are in electrical connection with the conductive structure.

Abstract: A method and apparatus for video coding system that uses intra prediction based on cross-colour linear model are disclosed. According to the method, model parameters for a first-colour predictor model are determined and the first-colour predictor model provides a predicted first-colour pixel value according to a combination of at least two corresponding reconstructed second-colour pixel values. According to another method, the first-colour predictor model provides a predicted first-colour pixel value based on a second degree model or higher of one or more corresponding reconstructed second-colour pixel values. First-colour predictors for the current first-colour block are determined according to the first-colour prediction model. The input data are then encoded at the encoder side or decoded at the decoder side using the first-colour predictors.

Abstract: A video coding system that uses chroma prediction is provided. The system receives data for a block of pixels to be encoded or decoded as a current block of a current picture of a video. The system constructs a chroma prediction model based on luma and chroma samples neighboring the current block. The system signals a set of chroma prediction related syntax element and a refinement to the chroma prediction model. The system performs chroma prediction by applying the chroma prediction model to reconstructed luma samples of the current block to obtain predicted chroma samples of the current block. The system uses the predicted chroma samples to reconstruct 10 chroma samples of the current block or to encode the current block.

Abstract: A method for specifying layout of subpictures in video pictures is provided. A video decoder receives data from a bitstream to be decoded as a current picture of a video. For a current subpicture of a set of subpictures of the current picture, the video decoder determines a position of the current subpicture based on a width and a height of the current picture and a previously determined width and height of a particular subpicture in the set of subpictures. The video decoder reconstructs the current picture and the current subpicture based on the determined position.

Abstract: Video processing methods comprise receiving input data of a current block in a current picture having a plurality of Virtual Pipeline Data Units (VPDUs), wherein each VDPU is a fixed size, performing a decoding process to decode Coding Units (CUs) in the current VPDU and generate decoded samples of the current VPDU, wherein each CU coded in a Current Picture Referencing (CPR) mode is decoded by fetching decoded samples in a reference block stored in a CPR referring buffer, and updating the CPR referring buffer by storing the decoded samples of the current VPDU when finishing the decoding process of the current VPDU, wherein the CPR referring buffer is updated one VPDU by one VPDU.

Abstract: Video processing methods and apparatuses for coding a current block include receiving input data of a current block, splitting the current block into sub-blocks, deriving motion information for the sub-blocks, performing motion compensation for the sub-blocks using the motion information to derive a final predictor, deriving and storing a representative motion vector for each grid in the current block, and encoding or decoding the current block using the final predictor. The representative motion vector for grids in a weighted area is determined by combining the motion vectors of the sub-blocks if the motion vectors of the sub-blocks are in different lists. The representative motion vector for grids in the weighted area is determined as one of the motion vectors of the sub-blocks if the motion vectors of the sub-blocks are in the same list.

Abstract: Video processing methods comprise receiving input data of a current block in a current picture, constructing a candidate list for the current block by including one or more History-based Motion Vector Predictor (HMVP) candidates, selecting one candidate from the candidate list, locating a reference block according to motion information of the selected candidate, and encoding or decoding the current block by predicting the current block using the reference block. The one or more HMVP candidates are fetched from a normal HMVP table if the current block is to be processed by a normal inter prediction mode, and the one or more HMVP candidates are fetched form a Current Picture Referencing (CPR) HMVP table if the current block is to be processed by a CPR mode. The two HMVP tables are separately maintained and updated.

Abstract: A video encoder receives raw pixel data to be encoded as a current block of a current picture of a video into a bitstream. The video encoder identifies multiple candidate bi-prediction positions for the current block, including a center position, a first set of offset positions, and a second set of offset positions. The first set of offset positions and the second set of offset positions interleave each other. The encoder computes distortion values for each of the candidate bi-prediction positions based on several possible weighting parameter values. The distortion values for the center position are based on each of the several possible weighting parameter values. The distortion values for the first set of offset positions are based on a first subset of the possible weighting parameter values. The distortion values for the second set of offset positions are based on a second subset of the possible weighting parameter values.

Abstract: Various schemes for managing search memory are described, which are beneficial in achieving enhanced coding gain, low latency, and/or reduced hardware for a video encoder or decoder. In processing a current block of a current picture, an apparatus determines a quantity of a plurality of reference pictures of the current picture. The apparatus subsequently determines, for at least one of the reference pictures, a corresponding search range size based on the quantity. The apparatus then determines, based on the search range size and a location of the current block, a search range of the reference picture, based on which the apparatus encodes or decodes the current block.

Abstract: A method and apparatus for video coding are disclosed. According to this method, a current block is received at an encoder side or compressed data comprising the current block is received at a decoder side, where the current block is partitioned into two geometric prediction units. Motion information for the two geometric prediction units is determined. Weighting information for the two geometric prediction units is determined. A motion storage type variable based on the weighting information is determined, where the motion information associated with the current block is stored according to a value of the motion storage type variable. A geometric predictor for the current block is generated by blending two geometric predictors according to the weighting information, where the two geometric predictors are derived for the two geometric prediction units using the motion information associated with the current block.

Abstract: A method for signaling adaptive loop filter (ALF) settings is provided. A video decoder receives data from a bitstream for a block of pixels to be decoded as a current block of a current picture of a video. The video decoder parses whether to use adaptive parameter set (APS) for filtering the current block without determining whether to select a particular APS from a plurality of APSs. When the APS is used to filter the current block, the video decoder parses a first filter selection index for selecting a filter from a set of filters in the APS. When the APS is not used to filter the current block, the video decoder signals a second filter selection index for selecting a filter from a set of default fixed filters. The video decoder filters the current block based on the selected filter. The video decoder outputs the filtered current block.

Abstract: A method and apparatus for video coding are disclosed. According to this method, a current block is received at an encoder side or compressed data comprising the current block is received at a decoder side, where the current block is partitioned into two geometric prediction units. Motion information for the two geometric prediction units is determined. Weighting information for the two geometric prediction units is determined. A motion storage type variable based on the weighting information is determined, where the motion information associated with the current block is stored according to a value of the motion storage type variable. A geometric predictor for the current block is generated by blending two geometric predictors according to the weighting information, where the two geometric predictors are derived for the two geometric prediction units using the motion information associated with the current block.

Abstract: A method and apparatus of video encoding video coding for a video encoder or decoder using Neural Network (NN) are disclosed. According to this method, the multiple frames in a video sequence comprises multiple segments, where each of the multiple segments comprises a set of frames. The NN (Neural Network) processing is applied to a target signal in one or more encoded frames of a target segment in the encoder side or to the target signal in one or more decoded frames of the target segment in the decoder side using one NN parameter set for the target segment. The target signal may correspond to reconstructed residual, reconstructed output, de-blocked output, SAO (sample adaptive offset) output, ALF (adaptive loop filter) output, or a combination thereof. In another embodiment, the NN processing is applied to a target signal only in one or more specific encoded or decoded frames.

Abstract: Method and apparatus for constrained mode signaling are disclosed. According to one method, a plurality of mode syntaxes are used to signal a current mode used for the current block. If one or more of the plurality of candidate modes are unavailable according to one or more conditions related to the current block, signaling a mode syntax is skipped at the video encoder side or parsing the mode syntax is skipped at the video decoder side. In one embodiment, a regular flag indicating on or off of a regular Merge mode or Skip mode is not signaled if Merge mode with MVD (MMVD) mode, Affine mode, subblock Temporal Motion Vector Prediction (SbTMVP) mode, Combined Inter and Intra Prediction (CIIP) Merge mode, and Triangle Prediction Mode (TPM) are disabled as indicated by an SPS flag.

Abstract: A video system that partitions slices of video pictures into slice chunks is provided. A video decoder receives data from a bitstream to be decoded as a current picture of a video. When a first syntax element in a picture parameter set (PPS) of the current picture indicates that one or more slices of the current picture are allowed to be divided into multiple slice chunks and when a second syntax element in a slice header of a current slice of the current picture indicates that data of the current slice is encoded and delivered in two or more slice chunks, the video decoder parses and specifies sizes and positions of the slice chunks of the current slice and reconstructs the current picture based on the slice chunks of the current slice.

Abstract: Video processing methods and apparatuses for coding a current block by constructing a candidate set including an average candidate generated from two or more original motion candidates. At least one MV of the average candidate is derived by directly averaging MVs of the original motion candidates in one of list 0 and list 1 without scaling regardless whether the MVs are pointing to same or different reference pictures. A selected candidate is determined from the candidate set as a MVP for a current MV of the current block for encoding or decoding the current block.