Abstract: A video decoding method incudes: obtaining a bitstream including a plurality of coded frames of a video signal; decoding each of the plurality of coded frames into a plurality of coding tree units (CTUs) and each of the plurality of CTUs into a plurality of residual blocks; recovering a coding block for each of the plurality of residual blocks based on multiple reference line intra prediction (MRLP) flags and reference samples included in each coded frame, where S2 is an allowed intra prediction mode set (AIPM) corresponding to a reference line that is immediately adjacent to the coding block, S1 is another AIPM corresponding to another reference line that is not immediately adjacent to the coding block, and S1 is a subset of S2, and S1 is determined by including angular intra prediction modes with nominal angles and either a SMOOTH_V mode or a SMOOTH_H mode.

Abstract: Methods, apparatus, and computer readable storage medium for offset refinement for intra prediction and multiple reference line intra prediction in video decoding. The method includes receiving, by a device, a coded video bitstream for a block. The device includes a memory storing instructions and a processor in communication with the memory. The method further includes determining, by the device, whether offset refinement for intra prediction applies to the block, based on mode information of the block, the mode information of the block comprising at least one of the following: a reference line index of the block, an intra prediction mode of the block, and a size of the block; and in response to determining that the offset refinement for intra prediction applies to the block, performing, by the device, the offset refinement to generate an intra predictor for intra prediction of the block.

Abstract: Various implementations described herein include methods and systems for coding video. In one aspect, a video bitstream includes a current coding block of an image frame, a first syntax element for a cross-component sample offset mode, and a second syntax element for a filter type and a filter shape index of a loop filter. Based on the loop filter, an electronic device identifies one or more neighboring luma samples of a first luma sample of the current coding block. The electronic device further determines and quantizes one or more difference values between the one or more neighboring luma samples and the first luma sample. The electronic device classifies a first color sample co-located with the first luma sample based on the quantized difference values to determine a first sample offset. The current coding block is reconstructed at least by adjusting the first color sample based on the first sample offset.

Abstract: This disclosure generally describes a set of advanced video coding technologies, and is specifically related to cross-component sample offset (CCSO) filtering. For example, a CCSO filter having fractional CCSO filtering tap positions may be used to generate sample offsets from one color component of a reconstructed video. The sample offsets may then be applied to a second color component of the reconstructed video. The reconstructed samples of the first color component may first be interpolated by an interpolation filter in order to generate interpolated samples of the first color component at the fractional CCSO tap positions. The interpolated samples may be used as input to the CCSO filter for the generation of the sample offsets.

Abstract: The various implementations described herein include methods and systems for coding video. In one aspect, a video bitstream includes a current coding block of an image frame and signals a cross-component intra prediction mode and a hypothesis tap index. An computing system identifies a first luma sample and a first chroma sample co-located with the first luma sample in the current coding block, and selects one of a plurality of hypothesis tap combinations based on the hypothesis tap index. The computing system identifies neighboring luma samples of the first luma sample based on the selected hypothesis tap combinations, and generates hypothesis values based on the identified neighboring luma samples of the first luma sample. The computing system further generates the first chroma sample based on at least the first luma sample and the hypothesis values, and reconstructs the current coding block including the first chroma sample.

Abstract: Aspects of the disclose provide methods and an apparatus for video and/or image coding. The apparatus includes processing circuitry that receive a bitstream including a frame. Coding information for a block in the frame indicates that the block is coded with a joint motion vector difference (JMVD) coding mode and a compound weighted prediction mode. The coding information includes scaling factor information of the JMVD coding mode. If the scaling factor information indicates that each of scaling factors of components of at least one MVD associated with at least one respective reference frame of the block is a pre-defined scaling factor, a weighting factor of the compound weighted prediction mode is determined based on a list of weighting factors. Otherwise, the weighting factor of the compound weighted prediction mode is determined based on a subset of the list of weighting factors.

Abstract: A method for video decoding in a decoder is provided. Coding information of a block to be reconstructed is decoded from a coded video bitstream. The coding information indicates intra prediction information for the block. Responsive to the block being coded with a directional mode, the directional mode is determined based on a nominal mode and an angular offset, the coding information indicating the nominal mode and the angular offset, a first non-separable transform set of one or more non-separable transforms for the block is determined based on the nominal mode, a non-separable transform in the first non-separable transform set is determined based on a non-separable transform index indicated by the coding information, and the block is reconstructed based on the directional mode and the non-separable transform.

Abstract: A process-in-memory architecture based on a resistive random access memory and a matrix decomposition acceleration algorithm, which is configured for transformer neural network acceleration. The present disclosure first optimizes a self-attention computing process, decomposes a weight matrix, and reduces computing and writing operands; and further reduces whole power consumption using a softmax computing array of a selection and comparison logic structure based on the resistive random access memory. The present disclosure proposes an optimized matrix multiplication computing based on Re-Transformer, and further eliminates data dependency and reduces computing delay in scaled dot-product attention by using matrix decomposition. Meanwhile, the present disclosure reduces power consumption by using hybrid softmax based on the resistive random access memory.

Abstract: This disclosure relates generally to video coding and particularly to methods and systems for deriving context adaptive interpolation filter used for generating prediction samples in inter prediction involving a fractional motion vector or block vector. For example, an encoder and a decoder may derive interpolation filters adaptively using reconstructed samples. The reconstructed samples may belong to templates near a current block and a reference block related by the motion vector or block vector. The derivation of the adaptive interpolation filters may be based on minimizing distortions between interpolated templates associated with the reference block and the templates associated with the current block.

Abstract: This disclosure relates generally to video coding/decoding and particularly for a transform dictionary. A method includes receiving the coded video bitstream; determining, based on the coded video bitstream, a transform pile for the current block, the transform pile being one transform pile in a transform dictionary, the transform dictionary comprising a plurality of transform piles, and the transform pile comprising a plurality of transform sets; determining a transform set in the transform pile; determining a transform kernel in the transform set; performing an inverse transform for the current block based on the transform kernel for obtaining a residual block for the current block; and reconstructing the current block based on the residual block.

Abstract: This disclosure relates generally to video coding/decoding and particularly for enhancing block adaptive weighted prediction (BAWP) with a block vector. One method includes receiving a coded video bitstream; identifying, from the coded video bitstream, a block vector corresponding to a reference block associated with the current block of the current frame; determining a scaling factor based on a syntax explicitly signaled in the coded video bitstream; generating a predicted block based on the reference block according to a linear equation associated with the scaling factor; and reconstructing, by the device, the current block based on the predicted block.

Abstract: This disclosure relates generally to video coding/decoding and particularly for enhancing block adaptive weighted prediction (BAWP). One method includes receiving a coded video bitstream; identifying, from the coded video bitstream, a motion vector corresponding to a reference block associated with the current block of the current frame; obtaining a scaling factor based on a syntax explicitly signed in the coded video bitstream; determining a template used to derive an offset value; deriving the offset value based on the template; generating a predicted block based on the reference block according to a linear equation, the linear equation being associated with the scaling factor and the offset value; and reconstructing, by the device, the current block based on the predicted block.

Abstract: Dual-precision analog memory cells and arrays are provided. In some embodiments, a memory cell, comprises a non-volatile memory element having an input terminal and at least one output terminal; and a volatile memory element having a plurality of input terminals and an output terminal, wherein the output terminal of the volatile memory element is coupled to the input terminal of the non-volatile memory element, and wherein the volatile memory element comprises: a first transistor coupled between a first supply and a common node, and a second transistor coupled between a second supply and the common node; wherein the common node is coupled to the output terminal of the volatile memory element; and wherein gates of the first and second transistors are coupled to respective ones of the plurality of input terminals of the volatile memory element.

Abstract: A method of and an apparatus for controlling intra prediction for decoding of a video sequence are provided. The method includes setting a table including a plurality of intra prediction modes respectively corresponding to a plurality of intra prediction angles, the plurality of intra prediction angles including diagonal intra prediction angles respectively corresponding to diagonal directions of all shapes of coding units, and selecting, for decoding the video sequence, one of the plurality of intra prediction modes, from the set table.

Abstract: Methods and apparatuses for performing chroma from luma (CfL) intra prediction, including: receiving a current block from a coded video bitstream; obtaining, from the coded video bitstream, a syntax element indicating which of two or more downsampling filters is used for predicting the current block in a CfL intra prediction mode; in response to the syntax element indicating that a first downsampling filter is used for the current block: determining a plurality of filter coefficients according to the first downsampling filter; and downsampling the current block based on the determined plurality of coefficients using a first number of sampling positions; in response to the syntax element indicating that a second downsampling filter is used for the current block: determining the plurality of filter coefficients according to the second downsampling filter; downsampling the current block based on the determined plurality of coefficients using a second number of sampling positions, wherein the second number of sam

Abstract: An example method of video decoding includes receiving a video bitstream comprising a plurality of frames, including a current frame having a chroma component and a luma component. When a semi-decoupled partitioning (SDP) mode is enabled for the current frame, the method includes parsing one or more parameters of a first constrained directional enhancement filter (CDEF) process from the video bitstream, applying the first CDEF process to the chroma component of the current frame according to the one or more parameters, parsing one or more second parameters of a second CDEF process from the video bitstream, applying the second CDEF process to the luma component of the current frame according to the one or more second parameters, and reconstructing the current frame in accordance with the first CDEF process and the second CDEF process. In this example, the second CDEF process is separate from the first CDEF process.

Abstract: In a method for video decoding in a video decoder, at least one syntax element is decoded from a coded video sequence. The at least one syntax element is indicative of an intra prediction mode. An intra prediction angle that corresponds to the intra prediction mode is determined based on a predetermined relationship between a plurality of intra prediction modes and a plurality of intra prediction angles. At least one sample of a block is reconstructed using the intra prediction angle that is determined to correspond to the intra prediction mode. The plurality of intra prediction modes in the predetermined relationship includes at least one of a first plurality of wide angle prediction modes beyond a bottom left direction diagonal mode and a second plurality of wide angle prediction modes beyond a top right direction diagonal mode. The block is a non-square block.

Abstract: A method of signaling an intra prediction mode used to encode a current block in an encoded video bitstream using at least one processor includes generating a first most probable mode (MPM) list corresponding to a zero reference line of the current block, wherein the first MPM list includes a plurality of angular intra prediction modes; generating a second MPM list corresponding to one or more non-zero reference lines of the current block, wherein the second MPM list includes the plurality of angular intra prediction modes; signaling a reference line index indicating a reference line used to encode the current block from among the zero reference line and the one or more non-zero reference lines; and signaling an intra mode index indicating the intra prediction mode within the first MPM list or the second MPM list.

Abstract: Disclosed are methods, apparatus, and computer-readable storage medium for entropy/decoding contexts for intra prediction modes in video encoding/decoding. An example method includes receiving a video bitstream that includes a current block and deriving a first context used to entropy encode a set index based on coded information. The coded information comprises at least one of: information of the current block, and mode information of at least one neighboring block of the current block. The method further includes deriving a second context used to entropy encode a mode index for the current block based on the coded information, entropy decoding the set index according to the first context, and entropy decoding the mode index according to the second context. The method also includes determining the intra prediction mode for the current block based on the set index and the mode index.

Abstract: A high-rotational speed cup-shaped grinding wheel includes an annular base, several blades and a flow splitting structure. The blades are fixed on a side of the base at an interval in a circumferential direction to form a blade ring. The side of the blade ring away from the base forms an annular working surface, and two adjacent blades are spaced apart from each other to form a water passage channel for delivering cooling water to the working surface. The flow splitting structure is fixed on the blade ring and splits the cooling water into two branches, where a first branch delivers the cooling water to an outer area of the working surface, and a second branch delivers the cooling water to an inner area of the working surface, and then delivers the cooling water from an inner area of the working surface to an outer side area thereof.