WO2021110170A1 - Updating of hmvp tables - Google Patents

Updating of hmvp tables Download PDF

Info

Publication number
WO2021110170A1
WO2021110170A1 PCT/CN2020/134287 CN2020134287W WO2021110170A1 WO 2021110170 A1 WO2021110170 A1 WO 2021110170A1 CN 2020134287 W CN2020134287 W CN 2020134287W WO 2021110170 A1 WO2021110170 A1 WO 2021110170A1
Authority
WO
WIPO (PCT)
Prior art keywords
candidate
video
hmvp
motion
conversion
Prior art date
Application number
PCT/CN2020/134287
Other languages
French (fr)
Inventor
Li Zhang
Kai Zhang
Hongbin Liu
Yue Wang
Siwei Ma
Original Assignee
Beijing Bytedance Network Technology Co., Ltd.
Bytedance Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Bytedance Network Technology Co., Ltd., Bytedance Inc. filed Critical Beijing Bytedance Network Technology Co., Ltd.
Priority to CN202080084219.8A priority Critical patent/CN114747218A/en
Publication of WO2021110170A1 publication Critical patent/WO2021110170A1/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/169Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
    • H04N19/17Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
    • H04N19/176Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/10Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
    • H04N19/102Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
    • H04N19/103Selection of coding mode or of prediction mode
    • H04N19/105Selection of the reference unit for prediction within a chosen coding or prediction mode, e.g. adaptive choice of position and number of pixels used for prediction
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51Motion estimation or motion compensation
    • H04N19/513Processing of motion vectors
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51Motion estimation or motion compensation
    • H04N19/513Processing of motion vectors
    • H04N19/517Processing of motion vectors by encoding
    • H04N19/52Processing of motion vectors by encoding by predictive encoding
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51Motion estimation or motion compensation
    • H04N19/577Motion compensation with bidirectional frame interpolation, i.e. using B-pictures

Definitions

  • This patent document relates to image and video coding and decoding.
  • Digital video accounts for the largest bandwidth use on the internet and other digital communication networks. As the number of connected user devices capable of receiving and displaying video increases, it is expected that the bandwidth demand for digital video usage will continue to grow.
  • the present document discloses techniques that can be used by video encoders and decoders to perform cross-component adaptive loop filtering during video encoding or decoding.
  • a method of video processing is disclosed.
  • a method of video processing includes performing a conversion between a video block of a video and a bitstream representation of the video; and selectively updating a table of motion candidates used for another video block by removing a Kth entry of the table, wherein K is a positive integer based on a rule.
  • a method of video processing includes performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate by removing an existing candidate in the table, wherein the rule depends on a characteristic of the existing candidate.
  • a method of video processing includes performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate based on the conversion at a Kth position from 0th position in the table, wherein K is dependent on a rule.
  • a method of video processing includes performing a conversion between a video block of a video processing unit of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block, wherein the rule depends on the video processing unit.
  • a method of video processing includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.
  • HMVP history-based motion vector prediction
  • a method of video processing includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table with a motion candidate based on position of a redundant candidate in the HMVP table, wherein the motion candidate is derived from motion information derived during the conversion and the redundant candidate is identical to or similar to the motion candidate.
  • HMVP history-based motion vector prediction
  • a method of video processing includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table by adding a motion candidate to a Kth entry in the HMVP table, wherein the motion candidate added into the HMVP table is derived from motion information derived during the conversion and K is a positive integer and is determined based on a predetermined rule.
  • HMVP history-based motion vector prediction
  • a method of video processing includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between video processing units of video blocks of a video and a bitstream representation of the video; and updating the HMVP table based on one or more rules associated with the video processing units, wherein the rules change from one video processing unit to another video processing unit.
  • HMVP history-based motion vector prediction
  • a video encoder apparatus comprising a processor configured to implement above-described methods.
  • a video encoder apparatus comprising a processor configured to implement above-described methods.
  • a computer readable medium having code stored thereon is disclose.
  • the code embodies one of the methods described herein in the form of processor-executable code.
  • FIG. 1 shows a typical HEVC video encoder (with decoder modeling elements shaded in light gray) .
  • FIG. 2 shows examples of video block partitions.
  • FIG. 3 shows examples of allowed prediction blocks for an MxM coding unit or video block.
  • FIGS. 4A-4B shows a splitting pattern
  • FIGS. 5A-5B shows a splitting pattern example using quadtree binary tree.
  • FIG. 6 (a) to (e) show various block partitioning patterns.
  • FIG. 7 shows a derivation process for merge candidates list construction.
  • FIG. 8 shows example positions of spatial merge candidates.
  • FIG. 9 shows candidate pairs considered for redundancy check of spatial merge candidates.
  • FIG. 10A-10B show positions for the second PU of N ⁇ 2N and 2N ⁇ N partitions.
  • FIG. 11 is an illustration of motion vector scaling for temporal merge candidate.
  • FIG. 12 shows candidate positions for temporal merge candidate, C0 and C1, in the col-located picture.
  • FIG. 13 shows an example of combined bi-predictive merge candidate.
  • FIG. 14 shows derivation process for motion vector prediction candidates.
  • FIG. 15 is an illustration of motion vector scaling for spatial motion vector candidate.
  • FIG. 16 shows an example coding flow of history based motion vector prediction (HMVP) .
  • HMVP history based motion vector prediction
  • FIG. 17 shows a modified merge list construction process.
  • FIGS. 18A-18B show redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
  • FIGS. 19A-19B show examples of redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
  • FIG. 20 shows CU and PU partitioning example.
  • FIG. 21 shows an example of a current block and its neighboring blocks used in AVS3.
  • FIGS. 22A-22D depict examples of redundancy-removal based HMVP updating method (with table size set to L) .
  • FIGS. 23A-23F depict examples if HMVP table updating processes.
  • FIG. 24 is a block diagram of an example video processing system in which disclosed techniques may be implemented.
  • FIG. 25 is a block diagram of an example hardware platform used for video processing.
  • FIG. 26 is a flowchart for an example method of video processing.
  • FIG. 27 is a flowchart for an example method of video processing.
  • FIG. 28 is a flowchart for an example method of video processing.
  • FIG. 29 is a flowchart for an example method of video processing.
  • FIG. 30 is a flowchart for an example method of video processing.
  • This patent document is related to video coding technologies. Specifically, it is related to motion vector coding in image/video coding. It may be applied to the existing video coding standard like HEVC, or the standard (Versatile Video Coding) , third generation of the Chinese Audio and Video coding Standard (AVS3) to be finalized. It may be also applicable to future video coding standards or video codec.
  • Video coding standards have evolved primarily through the development of the well-known ITU-T and ISO/IEC standards.
  • the ITU-T produced H. 261 and H. 263, ISO/IEC produced MPEG-1 and MPEG-4 Visual, and the two organizations jointly produced the H. 262/MPEG-2 Video and H. 264/MPEG-4 Advanced Video Coding (AVC) and H. 265/HEVC standards.
  • AVC H. 264/MPEG-4 Advanced Video Coding
  • H. 265/HEVC High Efficiency Video Coding
  • the video coding standards are based on the hybrid video coding structure wherein temporal prediction plus transform coding are utilized.
  • An example of a typical HEVC encoder framework is depicted in FIG. 1.
  • the core of the coding layer in previous standards was the macroblock, containing a 16 ⁇ 16 block of luma samples and, in the usual case of 4: 2: 0 color sampling, two corresponding 8 ⁇ 8 blocks of chroma samples.
  • An intra-coded block uses spatial prediction to exploit spatial correlation among pixels.
  • Two partitions are defined: 16x16 and 4x4.
  • An inter-coded block uses temporal prediction, instead of spatial prediction, by estimating motion among pictures.
  • Motion can be estimated independently for either 16x16 macroblock or any of its sub-macroblock partitions: 16x8, 8x16, 8x8, 8x4, 4x8, 4x4 (see Fig. 2) . Only one motion vector (MV) per sub-macroblock partition is allowed.
  • a CTU is split into CUs by using a quadtree structure denoted as coding tree to adapt to various local characteristics.
  • the decision whether to code a picture area using inter-picture (temporal) or intra-picture (spatial) prediction is made at the CU level.
  • Each CU can be further split into one, two or four PUs according to the PU splitting type. Inside one PU, the same prediction process is applied and the relevant information is transmitted to the decoder on a PU basis.
  • a CU can be partitioned into transform units (TUs) according to another quadtree structure similar to the coding tree for the CU.
  • TUs transform units
  • Coding tree units and coding tree block (CTB) structure The analogous structure in HEVC is the coding tree unit (CTU) , which has a size selected by the encoder and can be larger than a traditional macroblock.
  • the CTU consists of a luma CTB and the corresponding chroma CTBs and syntax elements.
  • HEVC then supports a partitioning of the CTBs into smaller blocks using a tree structure and quadtree-like signaling.
  • Coding units and coding blocks (CBs)
  • the quadtree syntax of the CTU specifies the size and positions of its luma and chroma CBs. The root of the quadtree is associated with the CTU. Hence, the size of the luma CTB is the largest supported size for a luma CB.
  • the splitting of a CTU into luma and chroma CBs is signaled jointly.
  • a CTB may contain only one CU or may be split to form multiple CUs, and each CU has an associated partitioning into prediction units (PUs) and a tree of transform units (TUs) .
  • PUs prediction units
  • TUs tree of transform units
  • Prediction units and prediction blocks The decision whether to code a picture area using inter picture or intra picture prediction is made at the CU level.
  • a PU partitioning structure has its root at the CU level.
  • the luma and chroma CBs can then be further split in size and predicted from luma and chroma prediction blocks (PBs) .
  • HEVC supports variable PB sizes from 64 ⁇ 64 down to 4 ⁇ 4 samples.
  • Fig. 3 shows examples of allowed PBs for a MxM CU.
  • the prediction residual is coded using block transforms.
  • a TU tree structure has its root at the CU level.
  • the luma CB residual may be identical to the luma transform block (TB) or may be further split into smaller luma TBs. The same applies to the chroma TBs.
  • Integer basis functions similar to those of a discrete cosine transform (DCT) are defined for the square TB sizes 4 ⁇ 4, 8 ⁇ 8, 16 ⁇ 16, and 32 ⁇ 32.
  • DCT discrete cosine transform
  • an integer transform derived from a form of discrete sine transform (DST) is alternatively specified.
  • a CB can be recursively partitioned into transform blocks (TBs) .
  • the partitioning is signaled by a residual quadtree. Only square CB and TB partitioning is specified, where a block can be recursively split into quadrants, as illustrated in FIGS. 4A-4B.
  • a flag signals whether it is split into four blocks of size M/2 ⁇ M/2. If further splitting is possible, as signaled by a maximum depth of the residual quadtree indicated in the SPS, each quadrant is assigned a flag that indicates whether it is split into four quadrants.
  • the leaf node blocks resulting from the residual quadtree are the transform blocks that are further processed by transform coding.
  • the encoder indicates the maximum and minimum luma TB sizes that it will use. Splitting is implicit when the CB size is larger than the maximum TB size. Not splitting is implicit when splitting would result in aluma TB size smaller than the indicated minimum.
  • the chroma TB size is half the luma TB size in each dimension, except when the luma TB size is 4 ⁇ 4, in which case a single 4 ⁇ 4 chroma TB is used for the region covered by four 4 ⁇ 4 luma TBs.
  • intra-picture-predicted CUs the decoded samples of the nearest-neighboring TBs (within or outside the CB) are used as reference data for intra picture prediction.
  • the HEVC design allows a TB to span across multiple PBs for inter-picture predicted CUs to maximize the potential coding efficiency benefits of the quadtree-structured TB partitioning.
  • a CTB is divided according to a quad-tree structure, the nodes of which are coding units.
  • the plurality of nodes in a quad-tree structure includes leaf nodes and non-leaf nodes.
  • the leaf nodes have no child nodes in the tree structure (i.e., the leaf nodes are not further split) .
  • The, non-leaf nodes include a root node of the tree structure.
  • the root node corresponds to an initial video block of the video data (e.g., a CTB) .
  • the respective non-root node corresponds to a video block that is a sub-block of a video block corresponding to a parent node in the tree structure of the respective non-root node.
  • Each respective non-leaf node of the plurality of non-leaf nodes has one or more child nodes in the tree structure.
  • JVET Joint Video Exploration Team
  • the QTBT structure removes the concepts of multiple partition types, i.e. it removes the separation of the CU, PU and TU concepts, and supports more flexibility for CU partition shapes.
  • a CU can have either a square or rectangular shape.
  • a coding tree unit (CTU) is first partitioned by a quadtree structure.
  • the quadtree leaf nodes are further partitioned by a binary tree structure.
  • the binary tree leaf nodes are called coding units (CUs) , and that segmentation is used for prediction and transform processing without any further partitioning.
  • a CU sometimes consists of coding blocks (CBs) of different colour components, e.g. one CU contains one luma CB and two chroma CBs in the case of P and B slices of the 4: 2: 0 chroma format and sometimes consists of a CB of a single component, e.g., one CU contains only one luma CB or just two chroma CBs in the case of I slices.
  • CBs coding blocks
  • - CTU size the root node size of a quadtree, the same concept as in HEVC.
  • MaxBTSize the maximally allowed binary tree root node size.
  • the CTU size is set as 128 ⁇ 128 luma samples with two corresponding 64 ⁇ 64 blocks of chroma samples
  • the MinQTSize is set as 16 ⁇ 16
  • the MaxBTSize is set as 64 ⁇ 64
  • the MinBTSize (for both width and height) is set as 4 ⁇ 4
  • the MaxBTDepth is set as 4.
  • the quadtree partitioning is applied to the CTU first to generate quadtree leaf nodes.
  • the quadtree leaf nodes may have a size from 16 ⁇ 16 (i.e., the MinQTSize) to 128 ⁇ 128 (i.e., the CTU size) .
  • the quadtree leaf node is also the root node for the binary tree and it has the binary tree depth as 0.
  • MaxBTDepth i.e., 4
  • no further splitting is considered.
  • MinBTSize i.e., 4
  • no further horizontal splitting is considered.
  • the binary tree node has height equal to MinBTSize
  • no further vertical splitting is considered.
  • the leaf nodes of the binary tree are further processed by prediction and transform processing without any further partitioning. In the JEM, the maximum CTU size is 256 ⁇ 256 luma samples.
  • FIG. 5A illustrates an example of block partitioning by using QTBT
  • FIG. 5B illustrates the corresponding tree representation.
  • the solid lines indicate quadtree splitting and dotted lines indicate binary tree splitting.
  • each splitting (i.e., non-leaf) node of the binary tree one flag is signalled to indicate which splitting type (i.e., horizontal or vertical) is used, where 0 indicates horizontal splitting and 1 indicates vertical splitting.
  • the quadtree splitting there is no need to indicate the splitting type since quadtree splitting always splits a block both horizontally and vertically to produce 4 sub-blocks with an equal size.
  • the QTBT scheme supports the ability for the luma and chroma to have a separate QTBT structure.
  • the luma and chroma CTBs in one CTU share the same QTBT structure.
  • the luma CTB is partitioned into CUs by a QTBT structure
  • the chroma CTBs are partitioned into chroma CUs by another QTBT structure. This means that a CU in an I slice consists of a coding block of the luma component or coding blocks of two chroma components, and a CU in a P or B slice consists of coding blocks of all three colour components.
  • inter prediction for small blocks is restricted to reduce the memory access of motion compensation, such that bi-prediction is not supported for 4 ⁇ 8 and 8 ⁇ 4 blocks, and inter prediction is not supported for 4 ⁇ 4 blocks.
  • these restrictions are removed.
  • TT ternary tree
  • FIG. 6 (a) quad-tree partitioning 6 (b) vertical binary-tree partitioning 6 (c) horizontal binary-tree partitioning 6 (d) vertical center-side ternary-tree partitioning 6 (e) horizontal center-side ternary-tree partitioning.
  • a CTU is firstly partitioned by region tree (RT) .
  • a RT leaf may be further split with prediction tree (PT) .
  • a PT leaf may also be further split with PT until max PT depth is reached.
  • a PT leaf is the basic coding unit. It is still called CU for convenience.
  • a CU cannot be further split.
  • Prediction and transform are both applied on CU in the same way as JEM.
  • the whole partition structure is named ‘multiple-type-tree’ .
  • Each inter-predicted PU has motion parameters for one or two reference picture lists.
  • Motion parameters include a motion vector and a reference picture index. Usage of one of the two reference picture lists may also be signalled using inter_pred_idc.
  • Motion vectors may be explicitly coded as deltas relative to predictors, such a coding mode is called AMVP mode.
  • a merge mode is specified whereby the motion parameters for the current PU are obtained from neighbouring PUs, including spatial and temporal candidates.
  • the merge mode can be applied to any inter-predicted PU, not only for skip mode.
  • the alternative to merge mode is the explicit transmission of motion parameters, where motion vector, corresponding reference picture index for each reference picture list and reference picture list usage are signalled explicitly per each PU.
  • the PU When signalling indicates that one of the two reference picture lists is to be used, the PU is produced from one block of samples. This is referred to as ‘uni-prediction’ . Uni-prediction is available both for P-slices and B-slices.
  • Bi-prediction When signalling indicates that both of the reference picture lists are to be used, the PU is produced from two blocks of samples. This is referred to as ‘bi-prediction’ . Bi-prediction is available for B-slices only.
  • Step 1.2 Redundancy check for spatial candidates
  • a maximum of four merge candidates are selected among candidates that are located in five different positions.
  • a maximum of one merge candidate is selected among two candidates. Since constant number of candidates for each PU is assumed at decoder, additional candidates are generated when the number of candidates does not reach to maximum number of merge candidate (MaxNumMergeCand) which is signalled in slice header. Since the number of candidates is constant, index of best merge candidate is encoded using truncated unary binarization (TU) . If the size of CU is equal to 8, all the PUs of the current CU share a single merge candidate list, which is identical to the merge candidate list of the 2N ⁇ 2N prediction unit.
  • TU truncated unary binarization
  • a maximum of four merge candidates are selected among candidates located in the positions depicted in FIG. 8.
  • the order of derivation is A 1 , B 1 , B 0 , A 0 and B 2 .
  • Position B 2 is considered only when any PU of position A 1 , B 1 , B 0 , A 0 is not available (e.g. because it belongs to another slice or tile) or is intra coded.
  • candidate at position A 1 is added, the addition of the remaining candidates is subject to a redundancy check which ensures that candidates with same motion information are excluded from the list so that coding efficiency is improved.
  • a redundancy check To reduce computational complexity, not all possible candidate pairs are considered in the mentioned redundancy check. Instead only the pairs linked with an arrow in FIG.
  • FIGS. 10A-10B depict the second PU for the case of N ⁇ 2N and 2N ⁇ N, respectively.
  • candidate at position A1 is not considered for list construction. In fact, by adding this candidate will lead to two prediction units having the same motion information, which is redundant to just have one PU in a coding unit.
  • position B 1 is not considered when the current PU is partitioned as 2N ⁇ N.
  • a scaled motion vector is derived based on co-located PU belonging to the picture which has the smallest POC difference with current picture within the given reference picture list.
  • the reference picture list to be used for derivation of the co-located PU is explicitly signalled in the slice header.
  • the scaled motion vector for temporal merge candidate is obtained as illustrated by the dotted line in 13, which is scaled from the motion vector of the co-located PU using the POC distances, tb and rd, where tb is defined to be the POC difference between the reference picture of the current picture and the current picture and td is defined to be the POC difference between the reference picture of the co-located picture and the co-located picture.
  • the reference picture index of temporal merge candidate is set equal to zero.
  • a practical realization of the scaling process is described in the HEVC specification. For a B-slice, two motion vectors, one is for reference picture list 0 and the other is for reference picture list 1, are obtained and combined to make the bi-predictive merge candidate.
  • FIG. 11 is an illustration of motion vector scaling for temporal merge candidate.
  • the position for the temporal candidate is selected between candidates C 0 and C 1 , as depicted in FIG. 12. If PU at position C 0 is not available, is intra coded, or is outside of the current CTU, position C 1 is used. Otherwise, position C 0 is used in the derivation of the temporal merge candidate.
  • merge candidates there are two additional types of merge candidates: combined bi-predictive merge candidate and zero merge candidate.
  • Combined bi-predictive merge candidates are generated by utilizing spatio-temporal merge candidates.
  • Combined bi-predictive merge candidate is used for B-Slice only.
  • the combined bi-predictive candidates are generated by combining the first reference picture list motion parameters of an initial candidate with the second reference picture list motion parameters of another. If these two tuples provide different motion hypotheses, they will form a new bi-predictive candidate. As an example, FIG.
  • Zero motion candidates are inserted to fill the remaining entries in the merge candidates list and therefore hit the MaxNumMergeCand capacity. These candidates have zero spatial displacement and a reference picture index which starts from zero and increases every time a new zero motion candidate is added to the list. The number of reference frames used by these candidates is one and two for uni and bi-directional prediction, respectively. Finally, no redundancy check is performed on these candidates.
  • HEVC defines the motion estimation region (MER) whose size is signalled in the picture parameter set using the “log2_parallel_merge_level_minus2” syntax element. When a MER is defined, merge candidates falling in the same region are marked as unavailable and therefore not considered in the list construction.
  • log2_parallel_merge_level_minus2 plus 2 specifies the value of the variable Log2ParMrgLevel, which is used in the derivation process for luma motion vectors for merge mode as specified in clause 8.5.3.2.2 and the derivation process for spatial merging candidates as specified in clause 8.5.3.2.3.
  • the value of log2_parallel_merge_level_minus2 shall be in the range of 0 to CtbLog2SizeY-2, inclusive.
  • Log2ParMrgLevel is derived as follows:
  • Log2ParMrgLevel log2_parallel_merge_level_minus2 + 2
  • Log2ParMrgLevel indicates the built-in capability of parallel derivation of the merging candidate lists. For example, when Log2ParMrgLevel is equal to 6, the merging candidate lists for all the prediction units (PUs) and coding units (CUs) contained in a 64x64 block can be derived in parallel.
  • PUs prediction units
  • CUs coding units
  • Motion vector prediction exploits spatio-temporal correlation of motion vector with neighbouring PUs, which is used for explicit transmission of motion parameters. It constructs a motion vector candidate list by firstly checking availability of left, above temporally neighbouring PU positions, removing redundant candidates and adding zero vector to make the candidate list to be constant length. Then, the encoder can select the best predictor from the candidate list and transmit the corresponding index indicating the chosen candidate. Similarly with merge index signalling, the index of the best motion vector candidate is encoded using truncated unary. The maximum value to be encoded in this case is 2 (see FIG. 14) . In the following sections, details about derivation process of motion vector prediction candidate are provided.
  • FIG. 14 summarizes derivation process for motion vector prediction candidate.
  • motion vector candidate two types are considered: spatial motion vector candidate and temporal motion vector candidate.
  • spatial motion vector candidate derivation two motion vector candidates are eventually derived based on motion vectors of each PU located in five different positions as depicted in FIG. 8.
  • one motion vector candidate is selected from two candidates, which are derived based on two different co-located positions. After the first list of spatio-temporal candidates is made, duplicated motion vector candidates in the list are removed. If the number of potential candidates is larger than two, motion vector candidates whose reference picture index within the associated reference picture list is larger than 1 are removed from the list. If the number of spatio-temporal motion vector candidates is smaller than two, additional zero motion vector candidates is added to the list.
  • a maximum of two candidates are considered among five potential candidates, which are derived from PUs located in positions as depicted in FIG. 8, those positions being the same as those of motion merge.
  • the order of derivation for the left side of the current PU is defined as A 0 , A 1 , and scaled A 0 , scaled A 1 .
  • the order of derivation for the above side of the current PU is defined as B 0 , B 1 , B 2 , scaled B 0 , scaled B 1 , scaled B 2 .
  • the no-spatial-scaling cases are checked first followed by the spatial scaling. Spatial scaling is considered when the POC is different between the reference picture of the neighbouring PU and that of the current PU regardless of reference picture list. If all PUs of left candidates are not available or are intra coded, scaling for the above motion vector is allowed to help parallel derivation of left and above MV candidates. Otherwise, spatial scaling is not allowed for the above motion vector.
  • the motion vector of the neighbouring PU is scaled in a similar manner as for temporal scaling, as depicted as FIG. 15.
  • the main difference is that the reference picture list and index of current PU is given as input; the actual scaling process is the same as that of temporal scaling.
  • AMVP mode For the AMVP mode, four parts may be signalled in the bitstream, i.e., prediction direction, reference index, MVD and mv predictor candidate index (highlighted in the syntax table below) . While for the merge mode, only a merge index may need to be signalled.
  • five_minus_max_num_merge_cand specifies the maximum number of merging MVP candidates supported in the slice subtracted from 5.
  • the maximum number of merging MVP candidates, MaxNumMergeCand is derived as follows:
  • MaxNumMergeCand 5-five_minus_max_num_merge_cand (2-1)
  • MaxNumMergeCand shall be in the range of 1 to 5, inclusive.
  • merge_flag [x0] [y0] specifies whether the inter prediction parameters for the current prediction unit are inferred from a neighbouring inter-predicted partition.
  • the array indices x0, y0 specify the location (x0, y0) of the top-left luma sample of the considered prediction block relative to the top-left luma sample of the picture.
  • merge_flag [x0] [y0] is inferred to be equal to 0.
  • merge_idx [x0] [y0] specifies the merging candidate index of the merging candidate list where x0, y0 specify the location (x0, y0) of the top-left luma sample of the considered prediction block relative to the top-left luma sample of the picture.
  • VVC the translational motion and affine motion are handled in different ways.
  • the enhanced AMVP and Merge design are employed.
  • HMVP history-based motion vector prediction
  • LUT-based motion vector prediction techniques using one or more tables (e.g., look up tables) with at least one motion candidate stored to predict motion information of a block can be implemented in various embodiments to provide video coding with higher coding efficiencies.
  • a look up table is an example of a table which can be used to include motion candidates to predict motion information of a block and other implementations are also possible.
  • Each LUT can include one or more motion candidates, each associated with corresponding motion information.
  • Motion information of a motion candidate can include partial or all of the prediction direction, reference indices/pictures, motion vectors, LIC flags, affine flags, Motion Vector Derivation (MVD) precisions, and/or MVD values.
  • Motion information may further include block position information to indicate from which the motion information is coming.
  • the LUT-based motion vector prediction based on the disclosed technology which may enhance both existing and future video coding standards, is elucidated in the following examples described for various implementations. Because the LUTs allow the encoding/decoding process to be performed based on historical data (e.g., the blocks that have been processed) , the LUT-based motion vector prediction can also be referred to as History-based Motion Vector Prediction (HMVP) method. In the LUT-based motion vector prediction method, one or multiple tables with motion information from previously coded blocks are maintained during the encoding/decoding process. These motion candidates stored in the LUTs are named HMVP candidates.
  • HMVP candidates are named HMVP candidates.
  • the associated motion information in LUTs may be added to the motion candidate lists (e.g., merge/AMVP candidate lists) , and after encoding/decoding one block, LUTs may be updated.
  • the updated LUTs are then used to code the subsequent blocks. That is, the updating of motion candidates in the LUTs are based on the encoding/decoding order of blocks.
  • HMVP the previously coded motion information is stored.
  • the motion information of a previously coded block is defined as an HMVP candidate.
  • Multiple HMVP candidates are stored in a table, named as the HMVP table, and this table is maintained during the encoding/decoding process on-the-fly.
  • the HMVP table is emptied when starting coding/decoding a new tile/LCU row/a slice.
  • non-TPM non-triangular prediction mode
  • Step 1 Derivation of temporal merge candidates (TMVP, temporal motion vector prediction)
  • HMVP candidates could be used in both AMVP and merge candidate list construction processes.
  • FIG. 17 depicts the modified merge candidate list construction process (highlighted in blue) .
  • HMVP candidates stored in the HMVP table could be utilized to fill in the merge candidate list.
  • the HMVP candidates in the table are inserted in a descending order of indices. The last entry in the table is firstly added to the list, while the first entry is added in the end. Similarly, redundancy removal is applied on the HMVP candidates. Once the total number of available merge candidates reaches the maximal number of merge candidates allowed to be signaled, the merge candidate list construction process is terminated.
  • HMVP table contains up to 5 motion candidates and each of them is unique.
  • a candidate is only added to the list if the corresponding candidate used for redundancy check has not the same motion information.
  • Such comparison process is called pruning process.
  • the pruning process among the spatial candidates is dependent on the usage of TPM for current block.
  • the HEVC pruning process i.e., five pruning for the spatial merge candidates is utilized.
  • the HMVP table is updated. Motion information of current block may be firstly pruned to all available HMVP candidates in the HMVP table.
  • a counter is assigned to the HMVP table which records number of available HMVP candidates in the table. It is initialized to be 0 if a HMVP table is reset.
  • the first case and last two cases are further depicted in the FIGS. 18A-18B and FIGS. 19A-19B, respectively.
  • FIG. 18A shows the case when redundant candidate found, and the LUT is full before adding a new motion candidate.
  • FIG. 18B shows the case when redundant candidate found, and the LUT is not full before adding a new motion candidate.
  • FIGS. 18A-18B show redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
  • FIG. 19A Redundant candidate NOT found, and the LUT is full before adding a new motion candidate
  • FIG. 19B Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
  • the MVP candidate hMvpCand consists of the luma motion vectors mvL0 and mvL1, the reference indices refldxL0 and refldxL1, the prediction list utilization flags predFlagL0 and predFlagL1, the bi-prediction weight index bcwldx, and the half sample interpolation filter index hpellfldx.
  • the candidate list HmvpCandList is modified using the candidate hMvpCand by the following ordered steps:
  • variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
  • the candidate list HmvpCandList is updated as follows:
  • HmvpCandList [i -1] is set equal to HmvpCandList [i] .
  • HmvpCandList [NumHmvpCand -1] is set equal to hMvpCand.
  • HmvpCandList [NumHmvpCand++] is set equal to hMvpCand.
  • HMVP is introduced to the AMVP list without pruning.
  • the AMVP and Merge list is shared, that is, the same construction process is utilized.
  • HMVP candidates in the IBC HMVP table and zero block vectors (BVs) are added to the candidate list.
  • two spatial IBC candidates, HMVP candidates, and zero BVs are added in order with partial pruning.
  • the IBC HMVP table is updated accordingly, in a similar way for the updating process of HMVP table used for inter mode.
  • the candidate list HmvplbcCandList is modified by the following ordered steps:
  • variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
  • the candidate list HmvplbcCandList is updated as follows:
  • HmvplbcCandList [i -1] is set equal to HmvplbcCandList [i] .
  • HmvplbcCandList [NumHmvplbcCand -1] is set equal to bvL.
  • HmvplbcCandList [NumHmvplbcCand ++] is set equal to bvL.
  • AVS2 Similar to HEVC, AVS2 also adopts the concept of CU, PU and TU-based coding/prediction/transform structure.
  • LCUs largest coding units
  • One LCU can be a single CU or can be split into four smaller CUs with a quad-tree partition structure; a CU can be recursively split until it reaches the smallest CU size limit, as shown in FIG. 20.
  • the leaf node CUs can be further split into PUs.
  • PU is the basic unit for intra-and inter prediction and allows multiple different shapes to encoder irregular image patterns, as shown in FIG. 20.
  • a P frame is a forward-predicted frame using a single reference picture
  • a B frame is a bipredicted frame that consists of forward, backward, bi-prediction, and symmetric prediction, using two reference frames.
  • symmetric prediction is defined as a special bi-prediction mode, wherein only one forward motion vector (MV) is coded and the backward MV is derived from the forward MV based on the picture-order-counter (POC) distances.
  • the symmetry mode could efficiently represent the linear motion model of an object.
  • the SKIP mode s motion information of the current block is derived from previously decoded blocks and no residual information is encoded. Similar to SKIP mode, DIRECT mode has no motion information to transmit while prediction residuals and mode information are transmitted.
  • a priority-based motion information derivation method is designed which takes block’s motion model (prediction direction) into consideration. A higher priority is assigned to the motion information of neighbor blocks with the same motion model as current block.
  • the motion information derivation process can be divided into three steps performed in order.
  • Motion model-matched search As shown in FIG. 21, an initial process of finding neighbor blocks with the same motion model as current block at positions F, G, C, A, B, D is conducted in that order. Once the first block sharing the same motion model with current block is found, the motion information of that block is assigned to current block.
  • N fw For bi-direction DIRECT/SKIP mode, count how many spatial neighbors are coded with forward-direction, denoted by N fw and how many spatial neighbors are coded with backward-direction, denoted by N bw . If N fw and N bw are both equal to or larger than 1, the combination of the first forward and the first backward prediction blocks’ motion information is assigned to current block.
  • the searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
  • N bi - count how many spatial neighbors are coded with bi-direction
  • the motion information of neighbor bi-direction predicted blocks’ motion information of the last block in scanning order of F, G, C, A, B and D (which is equal to the first block in the scanning order of D, B, A, C, G and F) is assigned to current block.
  • N bi if N bi is smaller than 2) if N bw is equal to or larger than 1 wherein N bw denotes how many spatial neighbors are coded with backward-direction, the backward motion vector (denoted by MvE1) of the first backward prediction blocks’ motion information is assigned to current block and the forward motion vector MvE0 is set equal to Clip3 (-32768, 32767, -MvE1) .
  • the searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
  • N fw denotes how many spatial neighbors are coded with forward-direction
  • the forward motion vector (denoted by MvE0) of the first forward prediction blocks’ motion information is assigned to current block and the backward motion vector MvE1 is set equal to Clip3 (-32768, 32767, -MvE0) .
  • the searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
  • N bi For backward DIRECT/SKIP mode, count how many spatial neighbors are coded with bi-direction, denoted by N bi . the backward motion information of the last neighbor bi-direction predicted block is assigned to current block.
  • the searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
  • N bi For forward DIRECT/SKIP mode, count how many spatial neighbors are coded with bi-direction, denoted by N bi . the forward motion information of the last neighbor bi-direction predicted block is assigned to current block.
  • the searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
  • step 3 Default MVs construction: this step is invoked only when both step 1) and step 2) fail to find available motion vectors.
  • step 3 Default MVs construction:
  • the backward motion vector is set to be a zero MV, i.e., (0, 0)
  • the forward motion vector is set to be a zero MV, i.e., (0, 0) .
  • the HMVP candidate index in a merge/amvp list is dynamically changed due to pruning among spatial/temporal candidates, in AVS3, the HMVP candidates are fixed to be from the index equal to 4 to 11 for B slices, and 2 to 9 for P slices.
  • the updating process is same as that in VVC.
  • the HMVP table is a table to store candidates wherein a candidate is not limited to a motion candidate (e.g., with the prediction direction, reference indices/pictures, motion vectors, LIC flag, affine flag, MVD precision, MVD values, block position information, block vectors (BVs) for IBC coded blocks) , it could also be a intra mode candidate or others.
  • a motion candidate e.g., with the prediction direction, reference indices/pictures, motion vectors, LIC flag, affine flag, MVD precision, MVD values, block position information, block vectors (BVs) for IBC coded blocks
  • BVs block vectors
  • HMVP table size is denoted by L, and each of candidate in the table is associated with an index, in the range of [0, L-1] .
  • HMVP table updating processing instead of removing the firstly (the earliest) added candidate (e.g. maybe with candidate index set to 0) in a HMVP table, it is proposed to remove the K-th entry in the HMVP table (i.e., the HMVP candidate with index equal to K) wherein K is a non-zero integer value.
  • K is pre-defined, e.g., it is set to 1.
  • K is dependent on the HMVP table size, i.e., maximum number of candidates to be stored in the HMVP table.
  • K is dependent on how the HMVP table is utilized.
  • K when updating the HMVP tables used for inter-coded blocks, K may be set to K0, and when updating the HMVP tables used for IBC-coded blocks, K may be set to K1 wherein K0 is unequal to K1.
  • K when updating the HMVP tables used for intra-coded blocks, K may be set equal to K2.
  • K2 maybe unequal to K1 or/and K0.
  • K is dependent on the coded information, such as block dimension, block position, how many times the table has been updated, the parity of times the table has been updated.
  • K may be signaled in a bitstream, such as in sequence/video/picture/slice/tile/subpicture/brick/other video processing unit-level.
  • K may be signaled in SPS/PPS/Picture header/slice header/VPS/DPS.
  • a variable may be further assigned, such as, to indicate the frequency of the candidate.
  • K may be adaptively changed on-the-fly.
  • K may be set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  • the first or the second value may be equal to 0.
  • the above method may be applied under certain conditions.
  • HMVP table when updating a HMVP table with a new candidate, if the table is full, and there is no redundant candidate in the HMVP table (i.e., a candidate in the HMVP table is identical to or similar to the new candidate) , then the K-th entry in the HMVP table is removed.
  • a redundant candidate in the HMVP table i.e., a candidate in the HMVP table is identical to or similar to the new candidate
  • the redundant one is associated with index equal to 0
  • How to update an HMVP table with a new candidate may depend on where the redundant (or similar) candidate is in the table (e.g., the index of the redundant candidate) .
  • the redundant candidate may be firstly modified and stored in the HMVP table.
  • the motion vectors/BVs associated with the redundant candidate may be modified by adding one or multiple offsets or being scaled.
  • the modified redundant candidate may be associated with the same index, i.e., N.
  • K is smaller or equal to T, wherein T is the number of candidates already in the table.
  • How to add a new candidate may depend on the index of the redundant HMVP candidate in the HMVP table.
  • K is pre-defined, e.g., 1.
  • K is dependent on the HMVP table size, i.e., maximum number of candidates to be stored in the HMVP table.
  • K is dependent on how the HMVP table is utilized.
  • K when updating the HMVP tables used for inter-coded blocks, K may be set to K0, and when updating the HMVP tables used for IBC-coded blocks, K may be set to K1 wherein K0 is unequal to K1.
  • K when updating the HMVP tables used for intra-coded blocks, K may be set equal to K2.
  • K2 maybe unequal to K1 or/and K0.
  • K is dependent on the coded information, such as block dimension, block position, how many times the table has been updated, the parity of times the table has been updated.
  • K may be signaled in a bitstream, such as in sequence/video/picture/slice/tile/subpicture/brick/other video processing unit-level.
  • K may be signaled in SPS/PPS/Picture header/slice header/VPS/DPS.
  • a variable may be further assigned, such as, to indicate the frequency of the candidate.
  • K may be adaptively changed on-the-fly.
  • K may be set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  • the first or the second value may be equal to 0.
  • K may be dependent on the pruning results, e.g., whether there is one identical or similar one in the HMVP table before being updated.
  • HMVP tables may be changed from one video processing unit to another video processing unit.
  • a for a first rule, pruning is applied, and for a second rule, non-pruning is applied.
  • partial pruning is applied (i.e., compare the new candidate to be added with partial of existing ones in the table)
  • full pruning is applied (i.e., compare the new candidate to be added with all of existing ones in the table) .
  • HMVP table it may depend on how the HMVP table is going to be used, e.g., for blocks with which prediction mode the HMVP table is utilized.
  • the coded information may include the coded mode information.
  • the above method or the existing method may be applied, and when updating the HMVP tables used for IBC-coded blocks, the FIFO rule may be applied.
  • the coded information may include the coded mode information.
  • the above method when updating the HMVP tables used for inter-coded blocks, the above method may be applied, and when updating the HMVP tables used for IBC-coded blocks, the existing method may be applied.
  • the HMVP table updating process is modified for the case when the table is full, and no redundant candidate is found among the existing candidates in the HMVP table.
  • FIG. 22A shows the case of: Redundant candidate found, and the LUT is full before adding a new motion candidate.
  • FIG. 22B shows the case of: Redundant candidate found, and the LUT is not full before adding a new motion candidate.
  • FIG. 22C shows the case of: Redundant candidate NOT found, and the LUT is full before adding a new motion candidate.
  • FIG. 22D shows the case of: Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
  • pls refer to FIG. 23E.
  • pls refer to FIG. 23C.
  • pls refer to FIG. 23D.
  • FIG. 23A shows the case of: Redundant candidate (with index unequal to 0) found, and the LUT is full before adding a new motion candidate.
  • FIG. 23B shows the case of: Redundant candidate (with index unequal to 0) found, and the LUT is not full before adding a new motion candidate.
  • FIG. 23C shows the case of: Redundant candidate (with index equal to 0) found, and the LUT is full before adding a new motion candidate.
  • FIG. 23D shows the case of: Redundant candidate (with index equal to 0) found, and the LUT is not full before adding a new motion candidate.
  • FIG. 23E shows the case of: Redundant candidate NOT found, and the LUT is full before adding a new motion candidate.
  • FIG. 23F shows the case of: Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
  • the deleted text is marked with and the newly added is highlighted.
  • the candidate list HmvplbcCandList is modified by the following ordered steps:
  • variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
  • the candidate list HmvplbcCandList is updated as follows:
  • HmvplbcCandList [i - 1] is set equal to HmvplbcCandList [i] .
  • HmvplbcCandList [NumHmvplbcCand - 1] is set equal to bvL.
  • HmvplbcCandList [NumHmvplbcCand ++] is set equal to bvL.
  • the MVP candidate hMvpCand consists of the luma motion vectors mvL0 and mvL1, the reference indices refIdxL0 and refIdxL1, the prediction list utilization flags predFlagL0 and predFlagL1, the bi-prediction weight index bcwIdx, and the half sample interpolation filter index hpelIfIdx.
  • the candidate list HmvpCandList is modified using the candidate hMvpCand by the following ordered steps:
  • variable identicalCandExist is set equal to FALSE and the variable removeIdx is set equal to 0.
  • the candidate list HmvpCandList is updated as follows:
  • HmvpCandList [i-1] is set equal to HmvpCandList [i] .
  • HmvpCandList [NumHmvpCand -1] is set equal to hMvpCand.
  • HmvpCandList [NumHmvpCand++] is set equal to hMvpCand.
  • FIG. 24 is a block diagram showing an example video processing system 1900 in which various techniques disclosed herein may be implemented.
  • the system 1900 may include input 1902 for receiving video content.
  • the video content may be received in a raw or uncompressed format, e.g., 8 or 10 bit multi-component pixel values, or may be in a compressed or encoded format.
  • the input 1902 may represent a network interface, a peripheral bus interface, or a storage interface. Examples of network interface include wired interfaces such as Ethernet, passive optical network (PON) , etc. and wireless interfaces such as Wi-Fi or cellular interfaces.
  • PON passive optical network
  • the system 1900 may include a coding component 1904 that may implement the various coding or encoding methods described in the present document.
  • the coding component 1904 may reduce the average bitrate of video from the input 1902 to the output of the coding component 1904 to produce a coded representation of the video.
  • the coding techniques are therefore sometimes called video compression or video transcoding techniques.
  • the output of the coding component 1904 may be either stored, or transmitted via a communication connected, as represented by the component 1906.
  • the stored or communicated bitstream (or coded) representation of the video received at the input 1902 may be used by the component 1908 for generating pixel values or displayable video that is sent to a display interface 1910.
  • the process of generating user-viewable video from the bitstream representation is sometimes called video decompression.
  • certain video processing operations are referred to as “coding” operations or tools, it will be appreciated that the coding tools or operations are used at an encoder and corresponding decoding tools or operations that reverse the results of the coding will be performed by
  • peripheral bus interface or a display interface may include universal serial bus (USB) or high definition multimedia interface (HDMI) or Displayport, and so on.
  • storage interfaces include SATA (serial advanced technology attachment) , PCI, IDE interface, and the like.
  • FIG. 25 is a block diagram of a video processing apparatus 3600.
  • the apparatus 3600 may be used to implement one or more of the methods described herein.
  • the apparatus 3600 may be embodied in a smartphone, tablet, computer, Internet of Things (IoT) receiver, and so on.
  • the apparatus 3600 may include one or more processors 3602, one or more memories 3604 and video processing hardware 3606.
  • the processor (s) 3602 may be configured to implement one or more methods described in the present document.
  • the memory (memories) 3604 may be used for storing data and code used for implementing the methods and techniques described herein.
  • the video processing hardware 3606 may be used to implement, in hardware circuitry, some techniques described in the present document.
  • a method of video processing (e.g., method 2600 depicted in FIG. 26) , comprising: performing (2602) a conversion between a video block of a video and a bitstream representation of the video; and selectively updating (2604) a table of motion candidates used for another video block by removing a Kth entry of the table, wherein K is a positive integer based on a rule.
  • the rule specifies conditions under which the updating is done and conditions under which the updating is not performed (e.g., when there is no new candidate to be added) .
  • a method of video processing comprising: performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate by removing an existing candidate in the table, wherein the rule depends on a characteristic of the existing candidate.
  • a method of video processing comprising: performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate based on the conversion at a Kth position from 0th position in the table, wherein K is dependent on a rule.
  • a method of video processing comprising: performing a conversion between a video block of a video processing unit of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block, wherein the rule depends on the video processing unit.
  • a video decoding apparatus comprising a processor configured to implement a method recited in one or more of solutions 1 to 18.
  • a video encoding apparatus comprising a processor configured to implement a method recited in one or more of solutions 1 to 18.
  • a computer program product having computer code stored thereon, the code, when executed by a processor, causes the processor to implement a method recited in any of solutions 1 to 18.
  • FIG. 27 shows a flowchart of an example method for video processing.
  • the method includes maintaining (2702) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2704) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2706) the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.
  • HMVP history-based motion vector prediction
  • the performing a conversion is at least in part based on the HMVP table.
  • the predetermined rule specifies that K is pre-defined.
  • K is set to 1.
  • the predetermined rule specifies that K is dependent on a size of the HMVP table.
  • the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
  • K when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
  • K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
  • the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
  • K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
  • a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
  • the predetermined rule specifies that K is adaptively changed on-the-fly.
  • K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  • the first value or the second value is equal to 0.
  • the updating is applied under certain conditions.
  • the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
  • the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
  • FIG. 28 shows a flowchart of an example method for video processing.
  • the method includes maintaining (2802) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2804) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2806) the table with a motion candidate based on position of a redundant candidate in the HMVP table, wherein the motion candidate is derived from motion information derived during the conversion and the redundant candidate is identical to or similar to the motion candidate.
  • HMVP history-based motion vector prediction
  • the performing a conversion is at least in part based on the HMVP table.
  • the redundant candidate is associated with an index equal to N in the HMVP table, the redundant candidate is firstly modified and stored in the table, N being an integer.
  • N 0 or L-1, wherein L is a size of the HMVP table.
  • motion vectors or block vectors associated with the redundant candidate are modified by adding one or multiple offsets or being scaled.
  • the modified redundant candidate in the updated HMVP table, is associated with the same index N.
  • FIG. 29 shows a flowchart of an example method for video processing.
  • the method includes maintaining (2902) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2904) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2906) the table by adding a motion candidate to a Kth entry in the HMVP table, wherein the motion candidate added into the HMVP table is derived from motion information derived during the conversion and K is a positive integer and is determined based on a predetermined rule.
  • HMVP history-based motion vector prediction
  • the performing a conversion is at least in part based on the HMVP table.
  • the predetermined rule specifies that K is smaller or equal to T, wherein T is the number of candidates already in the HMVP table.
  • how to add the motion candidate depends on an index of a redundant candidate in the HVMP table, wherein the redundant candidate is identical to or similar to the motion candidate.
  • the predetermined rule specifies that K is pre-defined.
  • K is set to 1.
  • the motion candidate is added as a candidate with index set to (L-2) , wherein L is a size of the HMVP table.
  • the predetermined rule specifies that K is dependent on a size of the HMVP table.
  • the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
  • K when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
  • K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
  • the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
  • K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
  • a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
  • the predetermined rule specifies that K is adaptively changed on-the-fly.
  • K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  • the first value or the second value is equal to 0.
  • the predetermined rule specifies that K is dependent on pruning results indicating whether there is one identical or similar candidate in the HMVP table before being updated.
  • FIG. 30 shows a flowchart of an example method for video processing.
  • the method includes maintaining (3002) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (3004) a conversion between a current video block of a video and a bitstream representation of the video; and updating (3006) the HMVP table based on one or more rules associated with the video processing units, wherein the rules change from one video processing unit to another video processing unit.
  • HMVP history-based motion vector prediction
  • the performing a conversion is at least in part based on the HMVP table.
  • pruning is applied, and for a second rule, non-pruning is applied.
  • partial pruning is applied, and for a second rule, full pruning is applied.
  • the rules are determined dependent on one or more characteristics of the conversion of the video processing unit.
  • the one or more characteristics include a prediction mode of the processing unit.
  • the rules are determined based on coded mode information of the video processing unit.
  • the rules change from one video processing unit to another video processing unit or the rules specify adaptive order according to pruning results.
  • the rules are FIFO rule.
  • the updating is dependent of coded information associated with the video blocks.
  • the coded information includes coded mode information.
  • the conversion generates the video blocks of video from the bitstream representation.
  • the conversion generates the bitstream representation from the video blocks of video.
  • video processing may refer to video encoding, video decoding, video compression or video decompression.
  • video compression algorithms may be applied during conversion from pixel representation of a video to a corresponding bitstream representation or vice versa.
  • the bitstream representation of a current video block may, for example, correspond to bits that are either co-located or spread in different places within the bitstream, as is defined by the syntax.
  • a macroblock may be encoded in terms of transformed and coded error residual values and also using bits in headers and other fields in the bitstream.
  • the disclosed and other solutions, examples, embodiments, modules and the functional operations described in this document can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this document and their structural equivalents, or in combinations of one or more of them.
  • the disclosed and other embodiments can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer readable medium for execution by, or to control the operation of, data processing apparatus.
  • the computer readable medium can be a machine-readable storage device, a machine-readable storage substrate, a memory device, a composition of matter effecting a machine-readable propagated signal, or a combination of one or more them.
  • data processing apparatus encompasses all apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers.
  • the apparatus can include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
  • a propagated signal is an artificially generated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus.
  • a computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
  • a computer program does not necessarily correspond to a file in a file system.
  • a program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document) , in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code) .
  • a computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.
  • the processes and logic flows described in this document can be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output.
  • the processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit) .
  • processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer.
  • a processor will receive instructions and data from a read only memory or a random-access memory or both.
  • the essential elements of a computer are a processor for performing instructions and one or more memory devices for storing instructions and data.
  • a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks.
  • mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks.
  • a computer need not have such devices.
  • Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto optical disks; and CD ROM and DVD-ROM disks.
  • semiconductor memory devices e.g., EPROM, EEPROM, and flash memory devices
  • magnetic disks e.g., internal hard disks or removable disks
  • magneto optical disks e.g., CD ROM and DVD-ROM disks.
  • the processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

Updating of history based motion vector prediction tables is disclosed. One example method of video processing includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.

Description

UPDATING OF HMVP TABLES
CROSS-REFERENCE TO RELATED APPLICATION
Under the applicable patent law and/or rules pursuant to the Paris Convention, this application is made to timely claim the priority to and benefits of International Patent Application No. PCT/CN2019/123670, filed on December 6, 2019. The entire disclosures of International Patent Application No. PCT/CN2019/123670 are incorporated by reference as part of the disclosure of this application.
TECHNICAL FIELD
This patent document relates to image and video coding and decoding.
BACKGROUND
Digital video accounts for the largest bandwidth use on the internet and other digital communication networks. As the number of connected user devices capable of receiving and displaying video increases, it is expected that the bandwidth demand for digital video usage will continue to grow.
SUMMARY
The present document discloses techniques that can be used by video encoders and decoders to perform cross-component adaptive loop filtering during video encoding or decoding. In one example aspect, a method of video processing is disclosed.
In one example aspect, a method of video processing is disclosed. The method includes performing a conversion between a video block of a video and a bitstream representation of the video; and selectively updating a table of motion candidates used for another video block by removing a Kth entry of the table, wherein K is a positive integer based on a rule.
In one example aspect, a method of video processing is disclosed. The method includes performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate by removing an existing candidate in the table, wherein the rule depends on a characteristic of the existing candidate.
In one example aspect, a method of video processing is disclosed. The method includes performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate based on the conversion at a Kth position from 0th position in the table, wherein K is dependent on a rule.
In one example aspect, a method of video processing is disclosed. The method includes performing a conversion between a video block of a video processing unit of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block, wherein the rule depends on the video processing unit.
In one example aspect, a method of video processing is disclosed. The method includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.
In one example aspect, a method of video processing is disclosed. The method includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table with a motion candidate based on position of a redundant candidate in the HMVP table, wherein the motion candidate is derived from motion information derived during the conversion and the redundant candidate is identical to or similar to the motion candidate.
In one example aspect, a method of video processing is disclosed. The method includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between a current video block of a video and a bitstream representation of the video; and updating the table by adding a motion candidate to a Kth entry in the HMVP table, wherein the motion candidate added into the HMVP table is derived from motion information derived during the conversion and K is a positive integer and is determined based on a predetermined rule.
In one example aspect, a method of video processing is disclosed. The method includes maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing a conversion between video processing units of video blocks of a video and a bitstream representation of the video; and updating the HMVP table based on one  or more rules associated with the video processing units, wherein the rules change from one video processing unit to another video processing unit.
In yet another example aspect, a video encoder apparatus is disclosed. The video encoder comprises a processor configured to implement above-described methods.
In yet another example aspect, a video encoder apparatus is disclosed. The video encoder comprises a processor configured to implement above-described methods.
In yet another example aspect, a computer readable medium having code stored thereon is disclose. The code embodies one of the methods described herein in the form of processor-executable code.
These, and other, features are described throughout the present document.
BRIEF DESCRIPTION OF DRAWINGS
FIG. 1 shows a typical HEVC video encoder (with decoder modeling elements shaded in light gray) .
FIG. 2 shows examples of video block partitions.
FIG. 3 shows examples of allowed prediction blocks for an MxM coding unit or video block.
FIGS. 4A-4B shows a splitting pattern.
FIGS. 5A-5B shows a splitting pattern example using quadtree binary tree.
FIG. 6 (a) to (e) show various block partitioning patterns.
FIG. 7 shows a derivation process for merge candidates list construction.
FIG. 8 shows example positions of spatial merge candidates.
FIG. 9 shows candidate pairs considered for redundancy check of spatial merge candidates.
FIG. 10A-10B show positions for the second PU of N×2N and 2N×N partitions.
FIG. 11 is an illustration of motion vector scaling for temporal merge candidate.
FIG. 12 shows candidate positions for temporal merge candidate, C0 and C1, in the col-located picture.
FIG. 13 shows an example of combined bi-predictive merge candidate.
FIG. 14 shows derivation process for motion vector prediction candidates.
FIG. 15 is an illustration of motion vector scaling for spatial motion vector candidate.
FIG. 16 shows an example coding flow of history based motion vector prediction (HMVP) .
FIG. 17 shows a modified merge list construction process.
FIGS. 18A-18B show redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
FIGS. 19A-19B show examples of redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
FIG. 20 shows CU and PU partitioning example.
FIG. 21 shows an example of a current block and its neighboring blocks used in AVS3.
FIGS. 22A-22D depict examples of redundancy-removal based HMVP updating method (with table size set to L) .
FIGS. 23A-23F depict examples if HMVP table updating processes.
FIG. 24 is a block diagram of an example video processing system in which disclosed techniques may be implemented.
FIG. 25 is a block diagram of an example hardware platform used for video processing.
FIG. 26 is a flowchart for an example method of video processing.
FIG. 27 is a flowchart for an example method of video processing.
FIG. 28 is a flowchart for an example method of video processing.
FIG. 29 is a flowchart for an example method of video processing.
FIG. 30 is a flowchart for an example method of video processing.
DETAILED DESCRIPTION
Section headings are used in the present document for ease of understanding and do not limit the applicability of techniques and embodiments disclosed in each section only to that section. Furthermore, H. 266 terminology is used in some description only for ease of understanding and not for limiting scope of the disclosed techniques. As such, the techniques described herein are applicable to other video codec designs also.
1. Introduction
This patent document is related to video coding technologies. Specifically, it is related to motion vector coding in image/video coding. It may be applied to the existing video coding standard like HEVC, or the standard (Versatile Video Coding) , third generation of the Chinese Audio and Video coding Standard (AVS3) to be finalized. It may be also applicable to future video coding standards or video codec.
2. Brief Description
Video coding standards have evolved primarily through the development of the well-known ITU-T and ISO/IEC standards. The ITU-T produced H. 261 and H. 263, ISO/IEC produced MPEG-1 and MPEG-4 Visual, and the two organizations jointly produced the H. 262/MPEG-2 Video and H. 264/MPEG-4 Advanced Video Coding (AVC) and H. 265/HEVC standards. Since H. 262, the  video coding standards are based on the hybrid video coding structure wherein temporal prediction plus transform coding are utilized. An example of a typical HEVC encoder framework is depicted in FIG. 1.
2.1. Partition Structure
2.1.1. Partition tree structure in H. 264/AVC
The core of the coding layer in previous standards was the macroblock, containing a 16× 16 block of luma samples and, in the usual case of 4: 2: 0 color sampling, two corresponding 8×8 blocks of chroma samples.
An intra-coded block uses spatial prediction to exploit spatial correlation among pixels. Two partitions are defined: 16x16 and 4x4.
An inter-coded block uses temporal prediction, instead of spatial prediction, by estimating motion among pictures. Motion can be estimated independently for either 16x16 macroblock or any of its sub-macroblock partitions: 16x8, 8x16, 8x8, 8x4, 4x8, 4x4 (see Fig. 2) . Only one motion vector (MV) per sub-macroblock partition is allowed.
2.1.2. Partition tree structure in HEVC
In HEVC, a CTU is split into CUs by using a quadtree structure denoted as coding tree to adapt to various local characteristics. The decision whether to code a picture area using inter-picture (temporal) or intra-picture (spatial) prediction is made at the CU level. Each CU can be further split into one, two or four PUs according to the PU splitting type. Inside one PU, the same prediction process is applied and the relevant information is transmitted to the decoder on a PU basis. After obtaining the residual block by applying the prediction process based on the PU splitting type, a CU can be partitioned into transform units (TUs) according to another quadtree structure similar to the coding tree for the CU. One of key feature of the HEVC structure is that it has the multiple partition conceptions including CU, PU, and TU.
In the following, the various features involved in hybrid video coding using HEVC are highlighted as follows.
1) Coding tree units and coding tree block (CTB) structure: The analogous structure in HEVC is the coding tree unit (CTU) , which has a size selected by the encoder and can be larger than a traditional macroblock. The CTU consists of a luma CTB and the corresponding chroma CTBs and syntax elements. The size L×L of a luma CTB can be chosen as L = 16, 32, or 64 samples, with the larger sizes typically enabling better compression. HEVC then supports a partitioning of the CTBs into smaller blocks using a tree structure and quadtree-like signaling.
2) Coding units (CUs) and coding blocks (CBs) : The quadtree syntax of the CTU specifies the size and positions of its luma and chroma CBs. The root of the quadtree is associated with the CTU. Hence, the size of the luma CTB is the largest supported size for a luma CB. The splitting of a CTU into luma and chroma CBs is signaled jointly. One luma CB and ordinarily two chroma CBs, together with associated syntax, form a coding unit (CU) . A CTB may contain only one CU or may be split to form multiple CUs, and each CU has an associated partitioning into prediction units (PUs) and a tree of transform units (TUs) .
3) Prediction units and prediction blocks (PBs) : The decision whether to code a picture area using inter picture or intra picture prediction is made at the CU level. A PU partitioning structure has its root at the CU level. Depending on the basic prediction-type decision, the luma and chroma CBs can then be further split in size and predicted from luma and chroma prediction blocks (PBs) . HEVC supports variable PB sizes from 64×64 down to 4×4 samples. Fig. 3 shows examples of allowed PBs for a MxM CU.
4) TUs and transform blocks: The prediction residual is coded using block transforms. A TU tree structure has its root at the CU level. The luma CB residual may be identical to the luma transform block (TB) or may be further split into smaller luma TBs. The same applies to the chroma TBs. Integer basis functions similar to those of a discrete cosine transform (DCT) are defined for the square TB sizes 4×4, 8×8, 16×16, and 32×32. For the 4×4 transform of luma intra picture prediction residuals, an integer transform derived from a form of discrete sine transform (DST) is alternatively specified.
2.1.2.1. Tree-Structured Partitioning into Transform Blocks and Units
For residual coding, a CB can be recursively partitioned into transform blocks (TBs) . The partitioning is signaled by a residual quadtree. Only square CB and TB partitioning is specified, where a block can be recursively split into quadrants, as illustrated in FIGS. 4A-4B. For a given luma CB of size M×M, a flag signals whether it is split into four blocks of size M/2×M/2. If further splitting is possible, as signaled by a maximum depth of the residual quadtree indicated in the SPS, each quadrant is assigned a flag that indicates whether it is split into four quadrants. The leaf node blocks resulting from the residual quadtree are the transform blocks that are further processed by transform coding. The encoder indicates the maximum and minimum luma TB sizes that it will use. Splitting is implicit when the CB size is larger than the maximum TB size. Not splitting is implicit when splitting would result in aluma TB size smaller than the indicated minimum. The chroma TB size is half the luma TB size in each dimension, except when the luma TB size is 4×4, in which case a single 4×4 chroma TB is used for the region covered by four 4×4 luma TBs.  In the case of intra-picture-predicted CUs, the decoded samples of the nearest-neighboring TBs (within or outside the CB) are used as reference data for intra picture prediction.
In contrast to previous standards, the HEVC design allows a TB to span across multiple PBs for inter-picture predicted CUs to maximize the potential coding efficiency benefits of the quadtree-structured TB partitioning.
2.1.2.2. Parent and child nodes
A CTB is divided according to a quad-tree structure, the nodes of which are coding units. The plurality of nodes in a quad-tree structure includes leaf nodes and non-leaf nodes. The leaf nodes have no child nodes in the tree structure (i.e., the leaf nodes are not further split) . The, non-leaf nodes include a root node of the tree structure. The root node corresponds to an initial video block of the video data (e.g., a CTB) . For each respective non-root node of the plurality of nodes, the respective non-root node corresponds to a video block that is a sub-block of a video block corresponding to a parent node in the tree structure of the respective non-root node. Each respective non-leaf node of the plurality of non-leaf nodes has one or more child nodes in the tree structure.
2.1.3. Quadtree plus binary tree block structure with larger CTUs in JEM
To explore the future video coding technologies beyond HEVC, Joint Video Exploration Team (JVET) was founded by VCEG and MPEG jointly in 2015. Since then, many new methods have been adopted by JVET and put into the reference software named Joint Exploration Model (JEM) .
2.1.3.1. QTBT block partitioning structure
Different from HEVC, the QTBT structure removes the concepts of multiple partition types, i.e. it removes the separation of the CU, PU and TU concepts, and supports more flexibility for CU partition shapes. In the QTBT block structure, a CU can have either a square or rectangular shape. As shown in FIGS. 5A-5B, a coding tree unit (CTU) is first partitioned by a quadtree structure. The quadtree leaf nodes are further partitioned by a binary tree structure. There are two splitting types, symmetric horizontal splitting and symmetric vertical splitting, in the binary tree splitting. The binary tree leaf nodes are called coding units (CUs) , and that segmentation is used for prediction and transform processing without any further partitioning. This means that the CU, PU and TU have the same block size in the QTBT coding block structure. In the JEM, a CU sometimes consists of coding blocks (CBs) of different colour components, e.g. one CU contains one luma CB and two chroma CBs in the case of P and B slices of the 4: 2: 0 chroma format and  sometimes consists of a CB of a single component, e.g., one CU contains only one luma CB or just two chroma CBs in the case of I slices.
The following parameters are defined for the QTBT partitioning scheme.
- CTU size: the root node size of a quadtree, the same concept as in HEVC.
- MinQTSize: the minimally allowed quadtree leaf node size.
- MaxBTSize: the maximally allowed binary tree root node size.
- MaxBTDepth: the maximally allowed binary tree depth.
- MinBTSize: the minimally allowed binary tree leaf node size.
In one example of the QTBT partitioning structure, the CTU size is set as 128×128 luma samples with two corresponding 64×64 blocks of chroma samples, the MinQTSize is set as 16× 16, the MaxBTSize is set as 64×64, the MinBTSize (for both width and height) is set as 4×4, and the MaxBTDepth is set as 4. The quadtree partitioning is applied to the CTU first to generate quadtree leaf nodes. The quadtree leaf nodes may have a size from 16× 16 (i.e., the MinQTSize) to 128×128 (i.e., the CTU size) . If the leaf quadtree node is 128×128, it will not be further split by the binary tree since the size exceeds the MaxBTSize (i.e., 64×64) . Otherwise, the leaf quadtree node could be further partitioned by the binary tree. Therefore, the quadtree leaf node is also the root node for the binary tree and it has the binary tree depth as 0. When the binary tree depth reaches MaxBTDepth (i.e., 4) , no further splitting is considered. When the binary tree node has width equal to MinBTSize (i.e., 4) , no further horizontal splitting is considered. Similarly, when the binary tree node has height equal to MinBTSize, no further vertical splitting is considered. The leaf nodes of the binary tree are further processed by prediction and transform processing without any further partitioning. In the JEM, the maximum CTU size is 256×256 luma samples.
FIG. 5A illustrates an example of block partitioning by using QTBT, and FIG. 5B illustrates the corresponding tree representation. The solid lines indicate quadtree splitting and dotted lines indicate binary tree splitting. In each splitting (i.e., non-leaf) node of the binary tree, one flag is signalled to indicate which splitting type (i.e., horizontal or vertical) is used, where 0 indicates horizontal splitting and 1 indicates vertical splitting. For the quadtree splitting, there is no need to indicate the splitting type since quadtree splitting always splits a block both horizontally and vertically to produce 4 sub-blocks with an equal size.
In addition, the QTBT scheme supports the ability for the luma and chroma to have a separate QTBT structure. Currently, for P and B slices, the luma and chroma CTBs in one CTU share the same QTBT structure. However, for I slices, the luma CTB is partitioned into CUs by a QTBT structure, and the chroma CTBs are partitioned into chroma CUs by another QTBT structure. This means that a CU in an I slice consists of a coding block of the luma component or  coding blocks of two chroma components, and a CU in a P or B slice consists of coding blocks of all three colour components.
In HEVC, inter prediction for small blocks is restricted to reduce the memory access of motion compensation, such that bi-prediction is not supported for 4×8 and 8×4 blocks, and inter prediction is not supported for 4×4 blocks. In the QTBT of the JEM, these restrictions are removed.
2.1.4. Ternary-tree for VVC
As proposed, tree types other than quad-tree and binary-tree are supported. In the implementation, two more ternary tree (TT) partitions, i.e., horizontal and vertical center-side ternary-trees are introduced, as shown in FIG. 6 (d) and (e) .
FIG. 6 (a) quad-tree partitioning 6 (b) vertical binary-tree partitioning 6 (c) horizontal binary-tree partitioning 6 (d) vertical center-side ternary-tree partitioning 6 (e) horizontal center-side ternary-tree partitioning.
There are two levels of trees, region tree (quad-tree) and prediction tree (binary-tree or ternary-tree) . A CTU is firstly partitioned by region tree (RT) . A RT leaf may be further split with prediction tree (PT) . A PT leaf may also be further split with PT until max PT depth is reached. A PT leaf is the basic coding unit. It is still called CU for convenience. A CU cannot be further split. Prediction and transform are both applied on CU in the same way as JEM. The whole partition structure is named ‘multiple-type-tree’ .
2.2. Inter prediction in HEVC/H. 265
Each inter-predicted PU has motion parameters for one or two reference picture lists. Motion parameters include a motion vector and a reference picture index. Usage of one of the two reference picture lists may also be signalled using inter_pred_idc. Motion vectors may be explicitly coded as deltas relative to predictors, such a coding mode is called AMVP mode.
When a CU is coded with skip mode, one PU is associated with the CU, and there are no significant residual coefficients, no coded motion vector delta or reference picture index. A merge mode is specified whereby the motion parameters for the current PU are obtained from neighbouring PUs, including spatial and temporal candidates. The merge mode can be applied to any inter-predicted PU, not only for skip mode. The alternative to merge mode is the explicit transmission of motion parameters, where motion vector, corresponding reference picture index for each reference picture list and reference picture list usage are signalled explicitly per each PU.
When signalling indicates that one of the two reference picture lists is to be used, the PU is produced from one block of samples. This is referred to as ‘uni-prediction’ . Uni-prediction is available both for P-slices and B-slices.
When signalling indicates that both of the reference picture lists are to be used, the PU is produced from two blocks of samples. This is referred to as ‘bi-prediction’ . Bi-prediction is available for B-slices only.
The following text provides the details on the inter prediction modes specified in HEVC. The description will start with the merge mode.
2.2.1. Merge mode
2.2.1.1. Derivation of candidates for merge mode
When a PU is predicted using merge mode, an index pointing to an entry in the merge candidates list is parsed from the bitstream and used to retrieve the motion information. The construction of this list is specified in the HEVC standard and can be summarized according to the following sequence of steps:
● Step 1: Initial candidates derivation
○ Step 1.1: Spatial candidates derivation
○ Step 1.2: Redundancy check for spatial candidates
○ Step 1.3: Temporal candidates derivation
● Step 2: Additional candidates insertion
○ Step 2.1: Creation of bi-predictive candidates
○ Step 2.2: Insertion of zero motion candidates
These steps are also schematically depicted in FIG. 7. For spatial merge candidate derivation, a maximum of four merge candidates are selected among candidates that are located in five different positions. For temporal merge candidate derivation, a maximum of one merge candidate is selected among two candidates. Since constant number of candidates for each PU is assumed at decoder, additional candidates are generated when the number of candidates does not reach to maximum number of merge candidate (MaxNumMergeCand) which is signalled in slice header. Since the number of candidates is constant, index of best merge candidate is encoded using truncated unary binarization (TU) . If the size of CU is equal to 8, all the PUs of the current CU share a single merge candidate list, which is identical to the merge candidate list of the 2N×2N prediction unit.
In the following, the operations associated with the aforementioned steps are detailed.
2.2.1.2. Spatial candidates derivation
In the derivation of spatial merge candidates, a maximum of four merge candidates are selected among candidates located in the positions depicted in FIG. 8. The order of derivation is  A 1, B 1, B 0, A 0 and B 2. Position B 2 is considered only when any PU of position A 1, B 1, B 0, A 0 is not available (e.g. because it belongs to another slice or tile) or is intra coded. After candidate at position A 1 is added, the addition of the remaining candidates is subject to a redundancy check which ensures that candidates with same motion information are excluded from the list so that coding efficiency is improved. To reduce computational complexity, not all possible candidate pairs are considered in the mentioned redundancy check. Instead only the pairs linked with an arrow in FIG. 9 are considered and a candidate is only added to the list if the corresponding candidate used for redundancy check has not the same motion information. Another source of duplicate motion information is the “second PU” associated with partitions different from 2Nx2N. As an example, FIGS. 10A-10B depict the second PU for the case of N×2N and 2N×N, respectively. When the current PU is partitioned as N×2N, candidate at position A1 is not considered for list construction. In fact, by adding this candidate will lead to two prediction units having the same motion information, which is redundant to just have one PU in a coding unit. Similarly, position B 1 is not considered when the current PU is partitioned as 2N×N.
2.2.1.3. Temporal candidate derivation
In this step, only one candidate is added to the list. Particularly, in the derivation of this temporal merge candidate, a scaled motion vector is derived based on co-located PU belonging to the picture which has the smallest POC difference with current picture within the given reference picture list. The reference picture list to be used for derivation of the co-located PU is explicitly signalled in the slice header. The scaled motion vector for temporal merge candidate is obtained as illustrated by the dotted line in 13, which is scaled from the motion vector of the co-located PU using the POC distances, tb and rd, where tb is defined to be the POC difference between the reference picture of the current picture and the current picture and td is defined to be the POC difference between the reference picture of the co-located picture and the co-located picture. The reference picture index of temporal merge candidate is set equal to zero. A practical realization of the scaling process is described in the HEVC specification. For a B-slice, two motion vectors, one is for reference picture list 0 and the other is for reference picture list 1, are obtained and combined to make the bi-predictive merge candidate.
FIG. 11 is an illustration of motion vector scaling for temporal merge candidate.
In the co-located PU (Y) belonging to the reference frame, the position for the temporal candidate is selected between candidates C 0 and C 1, as depicted in FIG. 12. If PU at position C 0 is not available, is intra coded, or is outside of the current CTU, position C 1 is used. Otherwise, position C 0 is used in the derivation of the temporal merge candidate.
2.2.1.4. Additional candidates insertion
Besides spatio-temporal merge candidates, there are two additional types of merge candidates: combined bi-predictive merge candidate and zero merge candidate. Combined bi-predictive merge candidates are generated by utilizing spatio-temporal merge candidates. Combined bi-predictive merge candidate is used for B-Slice only. The combined bi-predictive candidates are generated by combining the first reference picture list motion parameters of an initial candidate with the second reference picture list motion parameters of another. If these two tuples provide different motion hypotheses, they will form a new bi-predictive candidate. As an example, FIG. 13 depicts the case when two candidates in the original list (on the left) , which have mvL0 and refldxL0 or mvL1 and refldxL1, are used to create a combined bi-predictive merge candidate added to the final list (on the right) . There are numerous rules regarding the combinations which are considered to generate these additional merge candidates.
Zero motion candidates are inserted to fill the remaining entries in the merge candidates list and therefore hit the MaxNumMergeCand capacity. These candidates have zero spatial displacement and a reference picture index which starts from zero and increases every time a new zero motion candidate is added to the list. The number of reference frames used by these candidates is one and two for uni and bi-directional prediction, respectively. Finally, no redundancy check is performed on these candidates.
2.2.1.5. Motion estimation regions for parallel processing
To speed up the encoding process, motion estimation can be performed in parallel whereby the motion vectors for all prediction units inside a given region are derived simultaneously. The derivation of merge candidates from spatial neighbourhood may interfere with parallel processing as one prediction unit cannot derive the motion parameters from an adjacent PU until its associated motion estimation is completed. To mitigate the trade-off between coding efficiency and processing latency, HEVC defines the motion estimation region (MER) whose size is signalled in the picture parameter set using the “log2_parallel_merge_level_minus2” syntax element. When a MER is defined, merge candidates falling in the same region are marked as unavailable and therefore not considered in the list construction.
7.3.2.3 Picture parameter set RBSP syntax
7.3.2.3.1 General picture parameter set RBSP syntax
Figure PCTCN2020134287-appb-000001
log2_parallel_merge_level_minus2 plus 2 specifies the value of the variable Log2ParMrgLevel, which is used in the derivation process for luma motion vectors for merge mode as specified in clause 8.5.3.2.2 and the derivation process for spatial merging candidates as specified in clause 8.5.3.2.3. The value of log2_parallel_merge_level_minus2 shall be in the range of 0 to CtbLog2SizeY-2, inclusive.
The variable Log2ParMrgLevel is derived as follows:
Log2ParMrgLevel = log2_parallel_merge_level_minus2 + 2
(7-37)
NOTE 3 -The value of Log2ParMrgLevel indicates the built-in capability of parallel derivation of the merging candidate lists. For example, when Log2ParMrgLevel is equal to 6, the merging candidate lists for all the prediction units (PUs) and coding units (CUs) contained in a 64x64 block can be derived in parallel.
2.2.2. Motion vector prediction in AMVP mode
Motion vector prediction exploits spatio-temporal correlation of motion vector with neighbouring PUs, which is used for explicit transmission of motion parameters. It constructs a motion vector candidate list by firstly checking availability of left, above temporally neighbouring PU positions, removing redundant candidates and adding zero vector to make the candidate list to be constant length. Then, the encoder can select the best predictor from the candidate list and transmit the corresponding index indicating the chosen candidate. Similarly with merge index signalling, the index of the best motion vector candidate is encoded using truncated unary. The maximum value to be encoded in this case is 2 (see FIG. 14) . In the following sections, details about derivation process of motion vector prediction candidate are provided.
2.2.2.1. Derivation of motion vector prediction candidates
FIG. 14 summarizes derivation process for motion vector prediction candidate.
In motion vector prediction, two types of motion vector candidates are considered: spatial motion vector candidate and temporal motion vector candidate. For spatial motion vector candidate derivation, two motion vector candidates are eventually derived based on motion vectors of each PU located in five different positions as depicted in FIG. 8.
For temporal motion vector candidate derivation, one motion vector candidate is selected from two candidates, which are derived based on two different co-located positions. After the first list of spatio-temporal candidates is made, duplicated motion vector candidates in the list are removed. If the number of potential candidates is larger than two, motion vector candidates whose reference picture index within the associated reference picture list is larger than 1 are removed from the list. If the number of spatio-temporal motion vector candidates is smaller than two, additional zero motion vector candidates is added to the list.
2.2.2.2. Spatial motion vector candidates
In the derivation of spatial motion vector candidates, a maximum of two candidates are considered among five potential candidates, which are derived from PUs located in positions as depicted in FIG. 8, those positions being the same as those of motion merge. The order of derivation for the left side of the current PU is defined as A 0, A 1, and scaled A 0, scaled A 1. The order of derivation for the above side of the current PU is defined as B 0, B 1, B 2, scaled B 0, scaled B 1, scaled B 2. For each side there are therefore four cases that can be used as motion vector candidate, with two cases not required to use spatial scaling, and two cases where spatial scaling is used. The four different cases are summarized as follows.
● No spatial scaling
- (1) Same reference picture list, and same reference picture index (same POC)
- (2) Different reference picture list, but same reference picture (same POC)
● Spatial scaling
- (3) Same reference picture list, but different reference picture (different POC)
- (4) Different reference picture list, and different reference picture (different POC)
The no-spatial-scaling cases are checked first followed by the spatial scaling. Spatial scaling is considered when the POC is different between the reference picture of the neighbouring PU and that of the current PU regardless of reference picture list. If all PUs of left candidates are not available or are intra coded, scaling for the above motion vector is allowed to help parallel  derivation of left and above MV candidates. Otherwise, spatial scaling is not allowed for the above motion vector.
In a spatial scaling process, the motion vector of the neighbouring PU is scaled in a similar manner as for temporal scaling, as depicted as FIG. 15. The main difference is that the reference picture list and index of current PU is given as input; the actual scaling process is the same as that of temporal scaling.
2.2.2.3. Temporal motion vector candidates
Apart for the reference picture index derivation, all processes for the derivation of temporal merge candidates are the same as for the derivation of spatial motion vector candidates (see FIG. 12) The reference picture index is signalled to the decoder.
2.2.2.4. Signalling of Merge/AMVP information
For the AMVP mode, four parts may be signalled in the bitstream, i.e., prediction direction, reference index, MVD and mv predictor candidate index (highlighted in the syntax table below) . While for the merge mode, only a merge index may need to be signalled.
Syntax tables:
7.3.6.1 General slice segment header syntax
Figure PCTCN2020134287-appb-000002
7.3.8.6 Prediction unit syntax
Figure PCTCN2020134287-appb-000003
7.3.8.9 Motion vector difference syntax
Figure PCTCN2020134287-appb-000004
Semantics
five_minus_max_num_merge_cand specifies the maximum number of merging MVP candidates supported in the slice subtracted from 5. The maximum number of merging MVP candidates, MaxNumMergeCand is derived as follows:
MaxNumMergeCand = 5-five_minus_max_num_merge_cand        (2-1)
The value of MaxNumMergeCand shall be in the range of 1 to 5, inclusive.
merge_flag [x0] [y0] specifies whether the inter prediction parameters for the current prediction unit are inferred from a neighbouring inter-predicted partition. The array indices x0, y0 specify the location (x0, y0) of the top-left luma sample of the considered prediction block relative to the top-left luma sample of the picture.
When merge_flag [x0] [y0] is not present, it is inferred as follows:
- If CuPredMode [x0] [y0] is equal to MODE_SKIP, merge flag [x0] [y0] is inferred to be equal to 1.
- Otherwise, merge_flag [x0] [y0] is inferred to be equal to 0.
merge_idx [x0] [y0] specifies the merging candidate index of the merging candidate list where x0, y0 specify the location (x0, y0) of the top-left luma sample of the considered prediction block relative to the top-left luma sample of the picture.
2.3. Inter mode coding in VVC
In VVC, the translational motion and affine motion are handled in different ways. For the translational motion, the enhanced AMVP and Merge design are employed.
2.3.1. History-based Motion Vector Prediction (HMVP)
Different from the merge list design, in VVC, the history-based motion vector prediction (HMVP) method is adopted wherein HMVP tables are maintained to store motion information from previously coded blocks. Such a method is also known as look-up-table based MVP method.
LUT-based motion vector prediction techniques using one or more tables (e.g., look up tables) with at least one motion candidate stored to predict motion information of a block can be  implemented in various embodiments to provide video coding with higher coding efficiencies. A look up table is an example of a table which can be used to include motion candidates to predict motion information of a block and other implementations are also possible. Each LUT can include one or more motion candidates, each associated with corresponding motion information. Motion information of a motion candidate can include partial or all of the prediction direction, reference indices/pictures, motion vectors, LIC flags, affine flags, Motion Vector Derivation (MVD) precisions, and/or MVD values. Motion information may further include block position information to indicate from which the motion information is coming.
The LUT-based motion vector prediction based on the disclosed technology, which may enhance both existing and future video coding standards, is elucidated in the following examples described for various implementations. Because the LUTs allow the encoding/decoding process to be performed based on historical data (e.g., the blocks that have been processed) , the LUT-based motion vector prediction can also be referred to as History-based Motion Vector Prediction (HMVP) method. In the LUT-based motion vector prediction method, one or multiple tables with motion information from previously coded blocks are maintained during the encoding/decoding process. These motion candidates stored in the LUTs are named HMVP candidates. During the encoding/decoding of one block, the associated motion information in LUTs may be added to the motion candidate lists (e.g., merge/AMVP candidate lists) , and after encoding/decoding one block, LUTs may be updated. The updated LUTs are then used to code the subsequent blocks. That is, the updating of motion candidates in the LUTs are based on the encoding/decoding order of blocks.
In HMVP, the previously coded motion information is stored. The motion information of a previously coded block is defined as an HMVP candidate. Multiple HMVP candidates are stored in a table, named as the HMVP table, and this table is maintained during the encoding/decoding process on-the-fly. The HMVP table is emptied when starting coding/decoding a new tile/LCU row/a slice. Whenever there is an inter-coded block and non-sub-block, non-triangular prediction mode (non-TPM) , the associated motion information is added to the last entry of the table as a new HMVP candidate. The overall coding flow is depicted in FIG. 16.
2.3.1.1. Regular merge list construction process
The construction of the regular merge list (for translational motion) can be summarized according to the following sequence of steps:
● Step 1: Derivation of spatial merge candidates
● Step 1: Derivation of temporal merge candidates (TMVP, temporal motion vector prediction)
● Step 2: Insertion of HMVP candidates
● Step 3: Insertion of pairwise average candidates
● Step 4: default motion candidates
HMVP candidates could be used in both AMVP and merge candidate list construction processes. FIG. 17 depicts the modified merge candidate list construction process (highlighted in blue) . When the merge candidate list is not full after the TMVP candidate insertion, HMVP candidates stored in the HMVP table could be utilized to fill in the merge candidate list. Considering that one block usually has a higher correlation with the nearest neighbourring block in terms of motion information, the HMVP candidates in the table are inserted in a descending order of indices. The last entry in the table is firstly added to the list, while the first entry is added in the end. Similarly, redundancy removal is applied on the HMVP candidates. Once the total number of available merge candidates reaches the maximal number of merge candidates allowed to be signaled, the merge candidate list construction process is terminated.
It is noted that all the spatial/temporal/HMVP candidate shall be coded with non-IBC mode. Otherwise, it is not allowed to be added to the regular merge candidate list.
HMVP table contains up to 5 motion candidates and each of them is unique.
2.3.1.2. Pruning processes
A candidate is only added to the list if the corresponding candidate used for redundancy check has not the same motion information. Such comparison process is called pruning process.
The pruning process among the spatial candidates is dependent on the usage of TPM for current block.
When current block is coded without TPM mode (e.g., regular merge, MMVD, CIIP) , the HEVC pruning process (i.e., five pruning) for the spatial merge candidates is utilized.
2.3.1.3. Updating of HMVP table
After decoding an inter-coded block with non-TPM, non-sub-block mode, the HMVP table is updated. Motion information of current block may be firstly pruned to all available HMVP candidates in the HMVP table.
A counter is assigned to the HMVP table which records number of available HMVP candidates in the table. It is initialized to be 0 if a HMVP table is reset.
There are three cases during the updating process, and how to handle each case is described as follows:
1. If there is one identical one found, the redundant one is removed from the table, and all the remaining are pushed forward with index minus 1. Then the motion information of current block is added to the HMVP table after all existing ones. Counter is kept unchanged.
2. If there is no identical one found and table is not full, the motion information of current block is added to the HMVP table after all existing ones. Counter is increased by 1.
3. If there is no identical one found and table is full, the first entry in the HMVP table is removed, all the remaining are pushed forward with index minus 1. the motion information of current block is added to the HMVP table after all existing ones. Counter is kept unchanged.
The first case and last two cases are further depicted in the FIGS. 18A-18B and FIGS. 19A-19B, respectively.
FIG. 18A shows the case when redundant candidate found, and the LUT is full before adding a new motion candidate.
FIG. 18B shows the case when redundant candidate found, and the LUT is not full before adding a new motion candidate.
FIGS. 18A-18B show redundancy-removal based HMVP updating method (with one redundancy motion candidate removed) .
(FIG. 19A) Redundant candidate NOT found, and the LUT is full before adding a new motion candidate, and FIG. 19B Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
The related specification is defined as follows:
8.5.2.16 Updating process for the history-based motion vector predictor candidate list
Inputs to this process are:
- luma motion vectors in 1/16 fractional-sample accuracy mvL0 and mvL1,
- reference indices refldxL0 and refldxL1,
- prediction list utilization flags predFlagL0 and predFlagL1,
- the bi-prediction weight index bcwldx,
- the half-sample interpolation filter index hpellfldx.
The MVP candidate hMvpCand consists of the luma motion vectors mvL0 and mvL1, the reference indices refldxL0 and refldxL1, the prediction list utilization flags predFlagL0 and predFlagL1, the bi-prediction weight index bcwldx, and the half sample interpolation filter index hpellfldx.
The candidate list HmvpCandList is modified using the candidate hMvpCand by the following ordered steps:
1. The variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
2. When NumHmvpCand is greater than 0, for each index hMvpldx with hMvpldx = 0.. NumHmvpCand -1, the following steps apply until identicalCandExist is equal to TRUE:
- When hMvpCand and HmvpCandList [hMvpldx] have the same motion vectors and the same reference indices, identicalCandExist is set equal to TRUE and removeldx is set equal to hMvpldx.
3. The candidate list HmvpCandList is updated as follows:
- If identicalCandExist is equal to TRUE or NumHmvpCand is equal to 5, the following applies:
- For each index i with i = (removeldx + 1) .. (NumHmvpCand -1) , HmvpCandList [i -1] is set equal to HmvpCandList [i] .
- HmvpCandList [NumHmvpCand -1] is set equal to hMvpCand.
- Otherwise (identicalCandExist is equal to FALSE and NumHmvpCand is less than 5) , the following applies:
- HmvpCandList [NumHmvpCand++] is set equal to hMvpCand.
2.3.2. AMVP list construction for translational motion
HMVP is introduced to the AMVP list without pruning.
2.4. Intra Block Copy (IBC) mode in VVC
In IBC, the AMVP and Merge list is shared, that is, the same construction process is utilized. For 4x4 blocks, only HMVP candidates in the IBC HMVP table and zero block vectors (BVs) are added to the candidate list. For other cases, two spatial IBC candidates, HMVP candidates, and zero BVs are added in order with partial pruning.
After decoding an IBC coded block, the IBC HMVP table is updated accordingly, in a similar way for the updating process of HMVP table used for inter mode.
The corresponding specification is provided as follows:
8.5.2.6 Updating process for the history-based block vector predictor candidate list
Inputs to this process are:
- luma block vector bvL in 1/16 fractional-sample accuracy.
The candidate list HmvplbcCandList is modified by the following ordered steps:
1. The variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
2. When NumHmvplbcCand is greater than 0, for each index hMvpldx with hMvpldx = 0.. NumHmvplbcCand -1, the following steps apply until identicalCandExist is equal to TRUE:
- When bvL is equal to HmvplbcCandList [hMvpldx] , identicalCandExist is set equal to TRUE and removeldx is set equal to hMvpldx.
3. The candidate list HmvplbcCandList is updated as follows:
- If identicalCandExist is equal to TRUE or NumHmvplbcCand is equal to 5, the following applies:
- For each index i with i = (removeldx + 1) .. (NumHmvplbcCand -1) , HmvplbcCandList [i -1] is set equal to HmvplbcCandList [i] .
- HmvplbcCandList [NumHmvplbcCand -1] is set equal to bvL.
- Otherwise (identicalCandExist is equal to FALSE and NumHmvplbcCand is less than 5) , the following applies:
- HmvplbcCandList [NumHmvplbcCand ++] is set equal to bvL.
2.5. Coding structure in AVS2
Similar to HEVC, AVS2 also adopts the concept of CU, PU and TU-based coding/prediction/transform structure. First, pictures are split into largest coding units (LCUs) , which consist of 2N x2N samples. One LCU can be a single CU or can be split into four smaller CUs with a quad-tree partition structure; a CU can be recursively split until it reaches the smallest CU size limit, as shown in FIG. 20. once the splitting of the Cu hierarchical tree is finished, the leaf node CUs can be further split into PUs. PU is the basic unit for intra-and inter prediction and allows multiple different shapes to encoder irregular image patterns, as shown in FIG. 20.
FIG. 20 shows (a) The maximum possible recursive CU structure in AVS2. (LCU size = 64, maximum hierarchical depth=4) . (b) possible PU splitting for skip, intra modes, and inter modes in AVS2, including symmetric and asymmetric prediction (d=1, 2 for intra prediction; and d=0, 1, 2 for inter prediction) .
2.6. Inter prediction in AVS3
Similar to HEVC, there are two inter-coded frame types, P frame and B frame. A P frame is a forward-predicted frame using a single reference picture, while a B frame is a bipredicted frame that consists of forward, backward, bi-prediction, and symmetric prediction, using two reference frames. In a B frame, in addition to the conventional forward, backward, bi-directional, and skip/direct prediction modes, symmetric prediction is defined as a special bi-prediction mode, wherein only one forward motion vector (MV) is coded and the backward MV is derived from the forward MV based on the picture-order-counter (POC) distances. The symmetry mode could efficiently represent the linear motion model of an object.
2.6.1. Multi-motion model based Skip/direct modes
The SKIP mode’s motion information of the current block is derived from previously decoded blocks and no residual information is encoded. Similar to SKIP mode, DIRECT mode has no motion information to transmit while prediction residuals and mode information are transmitted.
As an expansion of traditional SKIP and DIRECT modes wherein motion information from only one temporal prediction block (CuSubtypeldx equal to 0) is utilized, four additional motion models based SKIP and DIRECT modes are introduced in AVS2 and AVS3, i.e. bi-direction DIRECT/SKIP mode (CuSubtypeldx equal to 1) , symmetry DIRECT/SKIP mode (CuSubtypeldx equal to 2) , backward DIRECT/SKIP mode (CuSubtypeldx equal to 3) and forward DIRECT/SKIP mode (CuSubtypeldx equal to 4) .
A priority-based motion information derivation method is designed which takes block’s motion model (prediction direction) into consideration. A higher priority is assigned to the motion information of neighbor blocks with the same motion model as current block.
For each of the four additional motion model based skip/direct mode, the motion information derivation process can be divided into three steps performed in order.
1) Motion model-matched search: As shown in FIG. 21, an initial process of finding neighbor blocks with the same motion model as current block at positions F, G, C, A, B, D is conducted in that order. Once the first block sharing the same motion model with current block is found, the motion information of that block is assigned to current block.
2) Artificial search: In case of there isn’t a neighbor block sharing the same motion model with current block (i.e., no result found from step 1) ) , a second process of artificial constructing motion information of current block is conducted. This process varies from one motion model to another.
a. For bi-direction DIRECT/SKIP mode, count how many spatial neighbors are coded with forward-direction, denoted by N fw and how many spatial neighbors are coded with backward-direction, denoted by N bw. If N fw and N bw are both equal to or larger than 1, the combination of the first forward and the first backward prediction blocks’ motion information is assigned to current block. The searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
b. For symmetry DIRECT/SKIP mode, the following steps are performed in order
- count how many spatial neighbors are coded with bi-direction, denoted by N bi
- If N bi is equal to or larger than 2, the motion information of neighbor bi-direction predicted blocks’ motion information of the last block in scanning order of F, G, C, A, B and D (which is equal to the first block in the scanning order of D, B, A, C, G and F) is assigned to current block.
- else (if N bi is smaller than 2) if N bw is equal to or larger than 1 wherein N bw denotes how many spatial neighbors are coded with backward-direction, the backward motion vector (denoted by MvE1) of the first backward prediction blocks’ motion information is assigned to current block and the  forward motion vector MvE0 is set equal to Clip3 (-32768, 32767, -MvE1) . The searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
- else (if N bi is smaller than 2 and N bw is smaller than 1) if N fw is equal to or larger than 1 wherein N fw denotes how many spatial neighbors are coded with forward-direction, , the forward motion vector (denoted by MvE0) of the first forward prediction blocks’ motion information is assigned to current block and the backward motion vector MvE1 is set equal to Clip3 (-32768, 32767, -MvE0) . The searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
c. For backward DIRECT/SKIP mode, count how many spatial neighbors are coded with bi-direction, denoted by N bi. the backward motion information of the last neighbor bi-direction predicted block is assigned to current block. The searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
d. For forward DIRECT/SKIP mode, count how many spatial neighbors are coded with bi-direction, denoted by N bi. the forward motion information of the last neighbor bi-direction predicted block is assigned to current block. The searching order is the same as the 1st step, i.e., from positions F, G, C, A, B, and D.
3) Default MVs construction: this step is invoked only when both step 1) and step 2) fail to find available motion vectors. In AVS2/AVS3, the following applies:
a. For bi-direction/symmetry skip/direct modes, zero motion vectors are set for both forward and backward directions.
b. For backward DIRECT/SKIP mode, the backward motion vector is set to be a zero MV, i.e., (0, 0)
c. For forward DIRECT/SKIP mode, the forward motion vector is set to be a zero MV, i.e., (0, 0) .
Since there are multiple SKIP and DIRECT modes, it would be required to bring some side information in signaling SKIP and DIRECT modes’ motion model. In AVS2 reference software RD3.0, the Rate-distortion optimization method is used to select a mode to encode a block. The side information is taken into consideration to yield a more accurate cost value when calculate the RD-cost of the new SKIP and DIRECT modes.
2.6.2. HMVP for AVS3
The basic idea of the HMVP design in AVS3 and VVC are the same. However, different from VVC, the HMVP candidate index in a merge/amvp list is dynamically changed due to pruning among spatial/temporal candidates, in AVS3, the HMVP candidates are fixed to be from the index equal to 4 to 11 for B slices, and 2 to 9 for P slices.
Table 1. sub-CU types for P pictures
Figure PCTCN2020134287-appb-000005
Figure PCTCN2020134287-appb-000006
Table 2. sub-CU types for B pictures
Figure PCTCN2020134287-appb-000007
The updating process is same as that in VVC.
3. Examples of technical problems solved by disclosed technical solutions
The current design of HMVP updating process has the following problems:
1. It follows the assumption that the earlier added candidate has lower chance to be selected as MVP for subsequent blocks. However, it is not always true.
4. A listing of techniques and embodiments
The list below should be considered as examples to explain general concepts. These items should not be interpreted in a narrow way. Furthermore, these items can be combined in any manner. Combinations between this IDF and other IDFs are also applicable.
In this document, the HMVP table is a table to store candidates wherein a candidate is not limited to a motion candidate (e.g., with the prediction direction, reference indices/pictures, motion vectors, LIC flag, affine flag, MVD precision, MVD values, block position information, block vectors (BVs) for IBC coded blocks) , it could also be a intra mode candidate or others.
Suppose a HMVP table size is denoted by L, and each of candidate in the table is associated with an index, in the range of [0, L-1] .
1. During the HMVP table updating processing, instead of removing the firstly (the earliest) added candidate (e.g. maybe with candidate index set to 0) in a HMVP table, it is proposed to remove the K-th entry in the HMVP table (i.e., the HMVP candidate with index equal to K) wherein K is a non-zero integer value.
a. In one example, K is pre-defined, e.g., it is set to 1.
b. In one example, K is dependent on the HMVP table size, i.e., maximum number of candidates to be stored in the HMVP table.
c. In one example, K is dependent on how the HMVP table is utilized.
i. In one example, when updating the HMVP tables used for inter-coded blocks, K may be set to K0, and when updating the HMVP tables used for IBC-coded blocks, K may be set to K1 wherein K0 is unequal to K1.
ii. In one example, when updating the HMVP tables used for intra-coded blocks, K may be set equal to K2. K2 maybe unequal to K1 or/and K0.
d. In one example, K is dependent on the coded information, such as block dimension, block position, how many times the table has been updated, the parity of times the table has been updated.
e. In one example, K may be signaled in a bitstream, such as in sequence/video/picture/slice/tile/subpicture/brick/other video processing unit-level.
i. In one example, K may be signaled in SPS/PPS/Picture header/slice header/VPS/DPS.
ii. In one example, zero is excluded from the possible values to be signaled for K.
f. In one example, for each entry in a HMVP table, a variable may be further assigned, such as, to indicate the frequency of the candidate.
i. Alternatively, furthermore, which entry (e.g., decision of K) to be removed from the table may be determined according to the variable.
g. In one example, K may be adaptively changed on-the-fly.
i. In one example, K may be set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
ii. Alternatively, furthermore, the first or the second value may be equal to 0.
h. Alternatively, furthermore, the above method may be applied under certain conditions.
i. In one example, when updating a HMVP table with a new candidate, if the table is full, and there is no redundant candidate in the HMVP table (i.e., a candidate in the HMVP table is identical to or similar to the new candidate) , then the K-th entry in the HMVP table is removed.
ii. In one example, when updating a HMVP table with a new candidate, if there is a redundant candidate in the HMVP table (i.e., a candidate in the HMVP table is identical to or similar to the new candidate) and the redundant one is associated with index equal to 0, then the K-th entry in the HMVP table is removed.
2. How to update an HMVP table with a new candidate may depend on where the redundant (or similar) candidate is in the table (e.g., the index of the redundant candidate) .
a. In one example, if the redundant candidate is associated with an index equal to N (e.g., N = 0 or L-1) , the redundant candidate may be firstly modified and stored in the HMVP table.
i. In one example, the motion vectors/BVs associated with the redundant candidate may be modified by adding one or multiple offsets or being scaled.
ii. In one example, in the updated HMVP table, the modified redundant candidate may be associated with the same index, i.e., N.
3. When a new candidate is used to update a HMVP table, instead of always adding the new candidate after all existing ones after pruning, it is proposed to add the new candidate to the K-th entry (e.g., with index equal to L-1-K) to the last which is not the last one.
a. In one example, K is smaller or equal to T, wherein T is the number of candidates already in the table.
b. How to add a new candidate may depend on the index of the redundant HMVP candidate in the HMVP table.
c. In one example, K is pre-defined, e.g., 1.
i. In one example, the new one is added as a candidate with index set to (L-2) wherein K=1.
d. In one example, K is dependent on the HMVP table size, i.e., maximum number of candidates to be stored in the HMVP table.
e. In one example, K is dependent on how the HMVP table is utilized.
i. In one example, when updating the HMVP tables used for inter-coded blocks, K may be set to K0, and when updating the HMVP tables used for IBC-coded blocks, K may be set to K1 wherein K0 is unequal to K1.
ii. In one example, when updating the HMVP tables used for intra-coded blocks, K may be set equal to K2. K2 maybe unequal to K1 or/and K0.
f. In one example, K is dependent on the coded information, such as block dimension, block position, how many times the table has been updated, the parity of times the table has been updated.
g. In one example, K may be signaled in a bitstream, such as in sequence/video/picture/slice/tile/subpicture/brick/other video processing unit-level.
i. In one example, K may be signaled in SPS/PPS/Picture header/slice header/VPS/DPS.
h. In one example, for each entry in a HMVP table, a variable may be further assigned, such as, to indicate the frequency of the candidate.
i. Alternatively, furthermore, which entry (e.g., decision of K) to be removed from the table may be determined according to the variable.
i. In one example, K may be adaptively changed on-the-fly.
i. In one example, K may be set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
ii. Alternatively, furthermore, the first or the second value may be equal to 0.
j. In one example, K may be dependent on the pruning results, e.g., whether there is one identical or similar one in the HMVP table before being updated.
4. The rules for updating HMVP tables may be changed from one video processing unit to another video processing unit.
a. In one example, for a first rule, pruning is applied, and for a second rule, non-pruning is applied.
b. In one example, for a first rule, partial pruning is applied (i.e., compare the new candidate to be added with partial of existing ones in the table) , and for a second rule, full pruning is applied (i.e., compare the new candidate to be added with all of existing ones in the table) .
c. In one example, it may depend on how the HMVP table is going to be used, e.g., for blocks with which prediction mode the HMVP table is utilized.
d. In one example, the coded information may include the coded mode information.
i. In one example, when updating the HMVP tables used for inter-coded blocks, the above method or the existing method (adaptive order according to pruning results) may be applied, and when updating the HMVP tables used for IBC-coded blocks, the FIFO rule may be applied.
5. Whether to and/or how to apply the above-mentioned bullets may be dependent on coded information.
a. In one example, the coded information may include the coded mode information.
i. In one example, when updating the HMVP tables used for inter-coded blocks, the above method may be applied, and when updating the HMVP tables used for IBC-coded blocks, the existing method may be applied.
5. Embodiment Examples
5.1. Examples of figures to depict the HMVP table updating process
In FIG. 22C the HMVP table updating process is modified for the case when the table is full, and no redundant candidate is found among the existing candidates in the HMVP table.
FIG. 22A shows the case of: Redundant candidate found, and the LUT is full before adding a new motion candidate.
FIG. 22B shows the case of: Redundant candidate found, and the LUT is not full before adding a new motion candidate.
FIG. 22C shows the case of: Redundant candidate NOT found, and the LUT is full before adding a new motion candidate.
FIG. 22D shows the case of: Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
In FIGS. 23A-23F the HMVP table updating process (corresponding to K=1 in bullet 1 of section 4) is modified for the following cases:
when the table is full, and no redundant candidate is found among the existing candidates in the HMVP table, pls refer to FIG. 23E.
when the table is full, and redundant candidate is found to be the first candidate (HMVP0) in the HMVP table, pls refer to FIG. 23C.
when the table is not full, and redundant candidate is found to be the first candidate (HMVP0) in the HMVP table, pls refer to FIG. 23D.
FIG. 23A shows the case of: Redundant candidate (with index unequal to 0) found, and the LUT is full before adding a new motion candidate.
FIG. 23B shows the case of: Redundant candidate (with index unequal to 0) found, and the LUT is not full before adding a new motion candidate.
FIG. 23C shows the case of: Redundant candidate (with index equal to 0) found, and the LUT is full before adding a new motion candidate.
FIG. 23D shows the case of: Redundant candidate (with index equal to 0) found, and the LUT is not full before adding a new motion candidate.
FIG. 23E shows the case of: Redundant candidate NOT found, and the LUT is full before adding a new motion candidate.
FIG. 23F shows the case of: Redundant candidate NOT found, and the LUT is NOT full before adding a new motion candidate.
5.2. Embodiment #1 on top of AVS3 specification
The deleted text is marked with
Figure PCTCN2020134287-appb-000008
and the newly added is highlighted.
9.16更新历史运动信息表
完成当前预测单元的解码后,如果当前预测单元是帧内预测单元且不是仿射预测单元,当NumOfHmvpCand大于0时,根据当前预测块的运动信息更新历史运动信息表HmvpCandidateList;否则,不执行本条定义的操作。
a)将hmvpIdx初始化为0。
b)如果CntHmvp等于0,则HmvpCandidateList[CntHmvp]等于当前预测单元的运动信息,CntHmvp加1。
c)否则,根据9.15定义的方法判断当前预测块的运动信息和HmvpCandidateList[hmvpIdx]是否相同:
1)如果运动信息相同,执行步骤d),否则,hmvpIdx加1。
2)如果hmvpIdx小于CntHmvp,执行步骤c);否则,执行步骤d)。
d)如果hmvpIdx小于CntHmvp,则:
1)i从hmvpIdx到CntHmvp-1,令HmvpCandidateList[i]等于HmvpCandidateList[i+1];
2)HmvpCandidateList[CntHmvp-1]等于当前预测单元的运动信息。
否则,如果hmvpIdx等于CntHmvp且CntHmvp等于NumOfHmvpCand,则:
1)i从
Figure PCTCN2020134287-appb-000009
1到CntHmvp-1,令HmvpCandidateList[i]等于HmvpCandidateList[i+1];
2)HmvpCandidateList[CntHmvp-1]等于当前预测单元的运动信息。
否则,如果hmvpIdx等于CntHmvp且CntHmvp小于NumOfHmvpCand,则HmvpCandidateList[CntHmvp]等于当前预测单元的运动信息,CntHmvp加1。
5.3. Embodiment #1 on top of VVC specification
8.5.2.6 Updating process for the history-based block vector predictor candidate list
Inputs to this process are:
- luma block vector bvL in 1/16 fractional-sample accuracy.
The candidate list HmvplbcCandList is modified by the following ordered steps:
4. The variable identicalCandExist is set equal to FALSE and the variable removeldx is set equal to 0.
5. When NumHmvplbcCand is greater than 0, for each index hMvpldx with hMvpldx = 0..NumHmvplbcCand - 1, the following steps apply until identicalCandExist is equal to TRUE:
- When bvL is equal to HmvplbcCandList [ hMvpldx] , identicalCandExist is set equal to TRUE and removeldx is set equal to hMvpldx.
6. The candidate list HmvplbcCandList is updated as follows:
If identicalCandExist is equal to FALSE and NumHmvplbcCand is equal to 5, removeldx is set  to 1.
- If identicalCandExist is equal to TRUE or NumHmvplbcCand is equal to 5, the following applies:
- For each index i with i = (removeldx + 1) .. (NumHmvplbcCand - 1) , HmvplbcCandList [i - 1] is set equal to HmvplbcCandList [i] .
- HmvplbcCandList [NumHmvplbcCand - 1] is set equal to bvL.
- Otherwise (identicalCandExist is equal to FALSE and NumHmvplbcCand is less than 5) , the following applies:
- HmvplbcCandList [NumHmvplbcCand ++] is set equal to bvL.
8.5.2.16 Updating process for the history-based motion vector predictor candidate list
Inputs to this process are:
- luma motion vectors in 1/16 fractional-sample accuracy mvL0 and mvL1,
- reference indices refIdxL0 and refIdxL1,
- prediction list utilization flags predFlagL0 and predFlagL1,
- the bi-prediction weight index bcwIdx,
- the half-sample interpolation filter index hpelIfIdx.
The MVP candidate hMvpCand consists of the luma motion vectors mvL0 and mvL1, the reference indices refIdxL0 and refIdxL1, the prediction list utilization flags predFlagL0 and predFlagL1, the bi-prediction weight index bcwIdx, and the half sample interpolation filter index hpelIfIdx.
The candidate list HmvpCandList is modified using the candidate hMvpCand by the following ordered steps:
1. The variable identicalCandExist is set equal to FALSE and the variable removeIdx is set equal to 0.
2. When NumHmvpCand is greater than 0, for each index hMvpIdx with hMvpIdx = 0.. NumHmvpCand-1, the following steps apply until identicalCandExist is equal to TRUE:
- When hMvpCand and HmvpCandList [hMvpIdx] have the same motion vectors and the same reference indices, identicalCandExist is set equal to TRUE and removeIdx is set equal to hMvpIdx.
3. The candidate list HmvpCandList is updated as follows:
If identicalCandExist is equal to FALSE and NumHmvplbcCand is equal to 5, removeldx is set  to 1.
- If identicalCandExist is equal to TRUE or NumHmvpCand is equal to 5, the following applies:
- For each index i with i = (removeIdx + 1) .. (NumHmvpCand-1) , HmvpCandList [i-1] is set equal to HmvpCandList [i] .
- HmvpCandList [NumHmvpCand -1] is set equal to hMvpCand.
- Otherwise (identicalCandExist is equal to FALSE and NumHmvpCand is less than 5) , the following applies:
- HmvpCandList [NumHmvpCand++] is set equal to hMvpCand.
FIG. 24 FIG. is a block diagram showing an example video processing system 1900 in which various techniques disclosed herein may be implemented. Various implementations may include some or all of the components of the system 1900. The system 1900 may include input 1902 for receiving video content. The video content may be received in a raw or uncompressed format, e.g., 8 or 10 bit multi-component pixel values, or may be in a compressed or encoded format. The input 1902 may represent a network interface, a peripheral bus interface, or a storage interface. Examples of network interface include wired interfaces such as Ethernet, passive optical network (PON) , etc. and wireless interfaces such as Wi-Fi or cellular interfaces.
The system 1900 may include a coding component 1904 that may implement the various coding or encoding methods described in the present document. The coding component 1904 may reduce the average bitrate of video from the input 1902 to the output of the coding component 1904 to produce a coded representation of the video. The coding techniques are therefore sometimes called video compression or video transcoding techniques. The output of the coding component 1904 may be either stored, or transmitted via a communication connected, as represented by the component 1906. The stored or communicated bitstream (or coded) representation of the video received at the input 1902 may be used by the component 1908 for  generating pixel values or displayable video that is sent to a display interface 1910. The process of generating user-viewable video from the bitstream representation is sometimes called video decompression. Furthermore, while certain video processing operations are referred to as “coding” operations or tools, it will be appreciated that the coding tools or operations are used at an encoder and corresponding decoding tools or operations that reverse the results of the coding will be performed by a decoder.
Examples of a peripheral bus interface or a display interface may include universal serial bus (USB) or high definition multimedia interface (HDMI) or Displayport, and so on. Examples of storage interfaces include SATA (serial advanced technology attachment) , PCI, IDE interface, and the like. The techniques described in the present document may be embodied in various electronic devices such as mobile phones, laptops, smartphones or other devices that are capable of performing digital data processing and/or video display.
FIG. 25 is a block diagram of a video processing apparatus 3600. The apparatus 3600 may be used to implement one or more of the methods described herein. The apparatus 3600 may be embodied in a smartphone, tablet, computer, Internet of Things (IoT) receiver, and so on. The apparatus 3600 may include one or more processors 3602, one or more memories 3604 and video processing hardware 3606. The processor (s) 3602 may be configured to implement one or more methods described in the present document. The memory (memories) 3604 may be used for storing data and code used for implementing the methods and techniques described herein. The video processing hardware 3606 may be used to implement, in hardware circuitry, some techniques described in the present document.
A listing of solutions preferred by some embodiments is provided next.
The following solutions show example embodiments of techniques discussed in the previous section (e.g., item 1) .
1. A method of video processing (e.g., method 2600 depicted in FIG. 26) , comprising: performing (2602) a conversion between a video block of a video and a bitstream representation of the video; and selectively updating (2604) a table of motion candidates used for another video block by removing a Kth entry of the table, wherein K is a positive integer based on a rule. For example, the rule specifies conditions under which the updating is done and conditions under which the updating is not performed (e.g., when there is no new candidate to be added) .
2. The method of solution 1, wherein the rule specifies that K = 1.
3. The method of solution 1, wherein the rule specifies that K is dependent on a size of the table.
4. The method of solution 1, wherein the rule specifies that K depends on a characteristics of the conversion of the video block.
5. The method of solution 1, wherein the rule specifies a condition under which the updating is performed.
6. The method of solution 5, wherein the condition is that the table has no redundant candidate that is similar to a new candidate, and the rule specifies to remove the Kth entry in the table.
7. The method of solution 5, wherein the condition is that the table includes a redundant candidate that is similar to a new candidate and the redundant candidate corresponds to an index 0, and wherein the rule specifies to remove the Kth entry in the table.
The following solutions show example embodiments of techniques discussed in the previous section (e.g., item 2) .
8. A method of video processing, comprising: performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate by removing an existing candidate in the table, wherein the rule depends on a characteristic of the existing candidate.
9. The method of solution 8, wherein the characteristic of the existing candidate is an index of the existing candidate.
The following solutions show example embodiments of techniques discussed in the previous section (e.g., item 3) .
10. A method of video processing, comprising: performing a conversion between a video block of a video and a bitstream representation of the video; and updating, selectively based on a rule, a table of motion candidates used for another video block by adding a new candidate based on the conversion at a Kth position from 0th position in the table, wherein K is dependent on a rule.
11. The method of solution 10, wherein the rule specifies that K is dependent on an index of a redundant candidate in the table.
12. The method of solution 10, wherein K = 1 or K is dependent on a size of the table.
The following solutions show example embodiments of techniques discussed in the previous section (e.g., item 5) .
13. A method of video processing, comprising: performing a conversion between a video block of a video processing unit of a video and a bitstream representation of the video; and  updating, selectively based on a rule, a table of motion candidates used for another video block, wherein the rule depends on the video processing unit.
14. The method of solution 13, wherein the rule specifies whether pruning is applied to the updating for the video processing unit.
15. The method of any of solutions 13-14, wherein the rule is one of a first rule that specifies applying partial pruning or a second rule that specifies applying full pruning, depending on a characteristic of the video processing unit.
16. The method of any of solutions 1-15, wherein the table is a history based motion vector prediction table.
17. The method of any of solutions 1-16, wherein the performing the conversion comprising encoding the video to generate the coded representation.
18. The method of any of solutions 1-16, wherein the performing the conversion comprises parsing and decoding the coded representation to generate the video.
19. A video decoding apparatus comprising a processor configured to implement a method recited in one or more of solutions 1 to 18.
20. A video encoding apparatus comprising a processor configured to implement a method recited in one or more of solutions 1 to 18.
21. A computer program product having computer code stored thereon, the code, when executed by a processor, causes the processor to implement a method recited in any of solutions 1 to 18.
22. A method, apparatus or system described in the present document.
FIG. 27 shows a flowchart of an example method for video processing. The method includes maintaining (2702) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2704) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2706) the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.
In some examples, the performing a conversion is at least in part based on the HMVP table.
In some examples, the predetermined rule specifies that K is pre-defined.
In some examples, K is set to 1.
In some examples, the predetermined rule specifies that K is dependent on a size of the HMVP table.
In some examples, the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
In some examples, when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
In some examples, when the HMVP table is used for intra-coded blocks, K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
In some examples, the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
In some examples, K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
In some examples, for each entry in the table, a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
In some examples, the predetermined rule specifies that K is adaptively changed on-the-fly.
In some examples, K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
In some examples, the first value or the second value is equal to 0.
In some examples, the updating is applied under certain conditions.
In some examples, when updating the table with the motion candidate, if the table is full, and there is no redundant candidate in the table, the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
In some examples, when updating the table with the motion candidate, if there is a redundant candidate in the table and the redundant candidate is associated with index equal to 0, the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
FIG. 28 shows a flowchart of an example method for video processing. The method includes maintaining (2802) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2804) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2806) the table  with a motion candidate based on position of a redundant candidate in the HMVP table, wherein the motion candidate is derived from motion information derived during the conversion and the redundant candidate is identical to or similar to the motion candidate.
In some examples, the performing a conversion is at least in part based on the HMVP table.
In some examples, if the redundant candidate is associated with an index equal to N in the HMVP table, the redundant candidate is firstly modified and stored in the table, N being an integer.
In some examples, N=0 or L-1, wherein L is a size of the HMVP table.
In some examples, motion vectors or block vectors associated with the redundant candidate are modified by adding one or multiple offsets or being scaled.
In some examples, in the updated HMVP table, the modified redundant candidate is associated with the same index N.
FIG. 29 shows a flowchart of an example method for video processing. The method includes maintaining (2902) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (2904) a conversion between a current video block of a video and a bitstream representation of the video; and updating (2906) the table by adding a motion candidate to a Kth entry in the HMVP table, wherein the motion candidate added into the HMVP table is derived from motion information derived during the conversion and K is a positive integer and is determined based on a predetermined rule.
In some examples, the performing a conversion is at least in part based on the HMVP table.
In some examples, the predetermined rule specifies that K is smaller or equal to T, wherein T is the number of candidates already in the HMVP table.
In some examples, how to add the motion candidate depends on an index of a redundant candidate in the HVMP table, wherein the redundant candidate is identical to or similar to the motion candidate.
In some examples, the predetermined rule specifies that K is pre-defined.
In some examples, K is set to 1.
In some examples, the motion candidate is added as a candidate with index set to (L-2) , wherein L is a size of the HMVP table.
In some examples, the predetermined rule specifies that K is dependent on a size of the HMVP table.
In some examples, the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
In some examples, when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
In some examples, when HMVP table is used for intra-coded blocks, K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
In some examples, the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
In some examples, K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
In some examples, for each entry in the HMVP table, a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
In some examples, the predetermined rule specifies that K is adaptively changed on-the-fly.
In some examples, K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
In some examples, the first value or the second value is equal to 0.
In some examples, the predetermined rule specifies that K is dependent on pruning results indicating whether there is one identical or similar candidate in the HMVP table before being updated.
FIG. 30 shows a flowchart of an example method for video processing. The method includes maintaining (3002) a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate; performing (3004) a conversion between a current video block of a video and a bitstream representation of the video; and updating (3006) the HMVP table based on one or more rules associated with the video processing units, wherein the rules change from one video processing unit to another video processing unit.
In some examples, the performing a conversion is at least in part based on the HMVP table.
In some examples, for a first rule, pruning is applied, and for a second rule, non-pruning is applied.
In some examples, for a first rule, partial pruning is applied, and for a second rule, full pruning is applied.
In some examples, the rules are determined dependent on one or more characteristics of the conversion of the video processing unit.
In some examples, the one or more characteristics include a prediction mode of the processing unit.
In some examples, the rules are determined based on coded mode information of the video processing unit.
In some examples, when the HMVP table is used for inter-coded blocks, the rules change from one video processing unit to another video processing unit or the rules specify adaptive order according to pruning results.
In some examples, when the HMVP table is used for IBC-coded blocks, the rules are FIFO rule.
In some examples, the updating is dependent of coded information associated with the video blocks.
In some examples, the coded information includes coded mode information.
In some examples, the conversion generates the video blocks of video from the bitstream representation.
In some examples, the conversion generates the bitstream representation from the video blocks of video.
In the present document, the term “video processing” may refer to video encoding, video decoding, video compression or video decompression. For example, video compression algorithms may be applied during conversion from pixel representation of a video to a corresponding bitstream representation or vice versa. The bitstream representation of a current video block may, for example, correspond to bits that are either co-located or spread in different places within the bitstream, as is defined by the syntax. For example, a macroblock may be encoded in terms of transformed and coded error residual values and also using bits in headers and other fields in the bitstream.
The disclosed and other solutions, examples, embodiments, modules and the functional operations described in this document can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, including the structures disclosed in this document and their structural equivalents, or in combinations of one or more of them. The disclosed and  other embodiments can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer readable medium for execution by, or to control the operation of, data processing apparatus. The computer readable medium can be a machine-readable storage device, a machine-readable storage substrate, a memory device, a composition of matter effecting a machine-readable propagated signal, or a combination of one or more them. The term “data processing apparatus” encompasses all apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus can include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them. A propagated signal is an artificially generated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus.
A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document) , in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code) . A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a communication network.
The processes and logic flows described in this document can be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit) .
Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read only memory or a random-access memory or both. The essential elements of a computer are a  processor for performing instructions and one or more memory devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto optical disks, or optical disks. However, a computer need not have such devices. Computer readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto optical disks; and CD ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.
While this patent document contains many specifics, these should not be construed as limitations on the scope of any subject matter or of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular techniques. Certain features that are described in this patent document in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.
Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. Moreover, the separation of various system components in the embodiments described in this patent document should not be understood as requiring such separation in all embodiments.
Only a few implementations and examples are described and other implementations, enhancements and variations can be made based on what is described and illustrated in this patent document.

Claims (57)

  1. A method for video processing, comprising:
    maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate;
    performing a conversion between a current video block of a video and a bitstream representation of the video; and
    updating the table by removing a Kth entry in the table and adding a motion candidate into the HMVP table, wherein K is a positive integer and is determined based on a predetermined rule and the motion candidate added into the HMVP table is derived from motion information derived during the conversion.
  2. The method of claim 1, wherein the performing a conversion is at least in part based on the HMVP table.
  3. The method of claim 1 or 2, wherein the predetermined rule specifies that K is pre-defined.
  4. The method of claim 3, wherein K is set to 1.
  5. The method of claim 1, wherein the predetermined rule specifies that K is dependent on a size of the HMVP table.
  6. The method of claim 1, wherein the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
  7. The method of claim 1, wherein when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
  8. The method of claim 7, wherein when the HMVP table is used for intra-coded blocks, K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
  9. The method of claim 1, wherein the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the  table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
  10. The method of any of claims 1-9, wherein K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
  11. The method of claim 1, wherein for each entry in the table, a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
  12. The method of claim 1, wherein the predetermined rule specifies that K is adaptively changed on-the-fly.
  13. The method of claim 12, wherein K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  14. The method of claim 13, wherein the first value or the second value is equal to 0.
  15. The method of any of claims 1-14, wherein the updating is applied under certain conditions.
  16. The method of claims 15, wherein when updating the table with the motion candidate, if the table is full, and there is no redundant candidate in the table, the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
  17. The method of claim 15, wherein when updating the table with the motion candidate, if there is a redundant candidate in the table and the redundant candidate is associated with index equal to 0, the K-th entry in the table is removed, wherein the redundant candidate is identical to or similar to the new candidate.
  18. A method for video processing, comprising:
    maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate;
    performing a conversion between a current video block of a video and a bitstream representation of the video; and
    updating the table with a motion candidate based on position of a redundant candidate in the HMVP table, wherein the motion candidate is derived from motion information derived during the conversion and the redundant candidate is identical to or similar to the motion candidate.
  19. The method of claim 1, wherein the performing a conversion is at least in part based on the HMVP table.
  20. The method of claim 18 or 19, wherein if the redundant candidate is associated with an index equal to N in the HMVP table, the redundant candidate is firstly modified and stored in the table, N being an integer.
  21. The method of claim 20, wherein N=0 or L-1, wherein L is a size of the HMVP table.
  22. The method of claim 21, wherein motion vectors or block vectors associated with the redundant candidate are modified by adding one or multiple offsets or being scaled.
  23. The method of claim 22, wherein in the updated HMVP table, the modified redundant candidate is associated with the same index N.
  24. A method for video processing, comprising:
    maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate;
    performing a conversion between a current video block of a video and a bitstream representation of the video; and
    updating the table by adding a motion candidate to a Kth entry in the HMVP table, wherein the motion candidate added into the HMVP table is derived from motion information derived during the conversion and K is a positive integer and is determined based on a predetermined rule.
  25. The method of claim 24, wherein the performing a conversion is at least in part based on the HMVP table.
  26. The method of claim 24 or 25, wherein the predetermined rule specifies that K is smaller or equal to T, wherein T is the number of candidates already in the HMVP table.
  27. The method of claim 24, wherein how to add the motion candidate depends on an index of a redundant candidate in the HVMP table, wherein the redundant candidate is identical to or similar to the motion candidate.
  28. The method of claim 24, wherein the predetermined rule specifies that K is pre-defined.
  29. The method of claim 28, wherein K is set to 1.
  30. The method of claim 29, wherein the motion candidate is added as a candidate with index set to (L-2) , wherein L is a size of the HMVP table.
  31. The method of claim 24, wherein the predetermined rule specifies that K is dependent on a size of the HMVP table.
  32. The method of claim 24, wherein the predetermined rule specifies that K is dependent on one or more characteristics of the conversion of the video blocks.
  33. The method of claim 32, wherein when the HMVP table is used for inter-coded blocks, K is set to K0, and when the HMVP table is used for IBC-coded blocks, K is set to K1, wherein K0 is unequal to K1.
  34. The method of claim 33, wherein when HMVP table is used for intra-coded blocks, K is set equal to K2, wherein K2 is unequal to K1 or/and K0.
  35. The method of claim 24, wherein the predetermined rule specifies that K is dependent on at least one of coded information associated with the current video block, how many times the table has been updated and the parity of times that the table has been updated, wherein the coded information including at least one of block dimension and block position.
  36. The method of any of claims 24-35, wherein K is signaled in a bitstream in at least one of sequence level, video level, picture level, slice level, tile level, subpicture level, brick level or other video processing unit level.
  37. The method of claim 24, wherein for each entry in the HMVP table, a variable is assigned to indicate a frequency of the candidate associated with the entry, wherein the predetermined rule specifies that K is determined according to the variable.
  38. The method of claim 24, wherein the predetermined rule specifies that K is adaptively changed on-the-fly.
  39. The method of claim 38, wherein K is set to a first value for a first video region and set to a second value for a second video region, wherein the first and second values are different.
  40. The method of claim 39, wherein the first value or the second value is equal to 0.
  41. The method of claim 24, wherein the predetermined rule specifies that K is dependent on pruning results indicating whether there is one identical or similar candidate in the HMVP table before being updated.
  42. A method for video processing, comprising:
    maintaining a history-based motion vector prediction (HMVP) table which comprises at least one motion candidate;
    performing a conversion between video processing units of video blocks of a video and a bitstream representation of the video; and
    updating the HMVP table based on one or more rules associated with the video processing units, wherein the rules change from one video processing unit to another video processing unit.
  43. The method of claim 42, wherein the performing a conversion is at least in part based on the HMVP table.
  44. The method of claim 42 or 43, wherein for a first rule, pruning is applied, and for a second rule, non-pruning is applied.
  45. The method of claim 42 or 43, wherein for a first rule, partial pruning is applied, and for a second rule, full pruning is applied.
  46. The method of claim 42 or 43, wherein the rules are determined dependent on one or more characteristics of the conversion of the video processing unit.
  47. The method of claim 46, wherein the one or more characteristics include a prediction mode of the processing unit.
  48. The method of claim 42, wherein the rules are determined based on coded mode information of the video processing unit.
  49. The method of claim 48, wherein when the HMVP table is used for inter-coded blocks, the rules change from one video processing unit to another video processing unit or the rules specify adaptive order according to pruning results.
  50. The method of claim 48, wherein when the HMVP table is used for IBC-coded blocks, the rules are FIFO rule.
  51. The method of any of claims 1-48, wherein the updating is dependent of coded information associated with the video blocks.
  52. The method of claim 51, wherein the coded information includes coded mode information.
  53. The method of any of claims 1-52, wherein the conversion generates the video blocks of video from the bitstream representation.
  54. The method of anyone of claims 1-52, wherein the conversion generates the bitstream representation from the video blocks of video.
  55. An apparatus in a video system comprising a processor and a non-transitory memory with instructions thereon, wherein the instructions upon execution by the processor, cause the processor to implement the method in any one of claims 1 to 54.
  56. A computer program product stored on a non-transitory computer readable media, the computer program product including program code for carrying out the method in any one of claims 1 to 54.
  57. A non-transitory computer-readable recording medium storing a bitstream representation which is generated by the method in any one of claims 1-54 performed by a video processing apparatus.
PCT/CN2020/134287 2019-12-06 2020-12-07 Updating of hmvp tables WO2021110170A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202080084219.8A CN114747218A (en) 2019-12-06 2020-12-07 Update of HMVP tables

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2019123670 2019-12-06
CNPCT/CN2019/123670 2019-12-06

Publications (1)

Publication Number Publication Date
WO2021110170A1 true WO2021110170A1 (en) 2021-06-10

Family

ID=76221508

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/134287 WO2021110170A1 (en) 2019-12-06 2020-12-07 Updating of hmvp tables

Country Status (2)

Country Link
CN (1) CN114747218A (en)
WO (1) WO2021110170A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023072216A1 (en) * 2021-10-28 2023-05-04 Beijing Bytedance Network Technology Co., Ltd. Method, apparatus, and medium for video processing

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107710764A (en) * 2015-06-09 2018-02-16 高通股份有限公司 It is determined that the system and method for the illumination compensation state for video coding
US10491902B1 (en) * 2018-07-16 2019-11-26 Tencent America LLC Method and apparatus for history-based motion vector prediction

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107710764A (en) * 2015-06-09 2018-02-16 高通股份有限公司 It is determined that the system and method for the illumination compensation state for video coding
US10491902B1 (en) * 2018-07-16 2019-11-26 Tencent America LLC Method and apparatus for history-based motion vector prediction

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
L. ZHANG (BYTEDANCE), K. ZHANG (BYTEDANCE), H. LIU (BYTEDANCE), Y. WANG, P. ZHAO, D. HONG (BYTEDANCE): "CE4: History-based Motion Vector Prediction (Test 4.4.7)", 124. MPEG MEETING; 20181008 - 20181012; MACAO; (MOTION PICTURE EXPERT GROUP OR ISO/IEC JTC1/SC29/WG11), 2 October 2018 (2018-10-02), XP030191679 *
Y. HAN (QUALCOMM), W.-J. CHIEN (QUALCOMM), C.-C. CHEN (QUALCOMM), H. HUANG (QUALCOMM), C.-H. HUNG, Y. ZHANG, M. KARCZEWICZ (QUALCO: "CE4: Modification on History-based Motion Vector Prediction", 13. JVET MEETING; 20190109 - 20190118; MARRAKECH; (THE JOINT VIDEO EXPLORATION TEAM OF ISO/IEC JTC1/SC29/WG11 AND ITU-T SG.16 ), 19 January 2019 (2019-01-19), XP030202546 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023072216A1 (en) * 2021-10-28 2023-05-04 Beijing Bytedance Network Technology Co., Ltd. Method, apparatus, and medium for video processing

Also Published As

Publication number Publication date
CN114747218A (en) 2022-07-12

Similar Documents

Publication Publication Date Title
US11909951B2 (en) Interaction between lut and shared merge list
US11997253B2 (en) Conditions for starting checking HMVP candidates depend on total number minus K
US11871023B2 (en) Multi-motion model based video coding and decoding
US11962799B2 (en) Motion candidates derivation
US11589071B2 (en) Invoke of LUT updating
US11641483B2 (en) Interaction between merge list construction and other tools
AU2020250609B2 (en) Using interpolation filters for history based motion vector prediction
WO2021047542A1 (en) Recursive splitting of video coding blocks
CN113196777B (en) Reference pixel padding for motion compensation
WO2021110170A1 (en) Updating of hmvp tables
WO2021219144A1 (en) Entropy coding for partition syntax
TWI840400B (en) Multi-motion model based video coding and decoding
TWI840401B (en) Pruning in multi-motion model based skip and direct mode coded video blocks

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20895728

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 31-10-2022)

122 Ep: pct application non-entry in european phase

Ref document number: 20895728

Country of ref document: EP

Kind code of ref document: A1