Methods and apparatus for implicit sub-block transform coding by gradients in video coding systems
By implicitly determining the final position of sub-parts with non-zero residuals using gradient values, the method addresses the inefficiencies in VVC's SBT processes, enhancing coding efficiency and reducing signaling overhead.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- MEDIATEK INC
- Filing Date
- 2025-10-21
- Publication Date
- 2026-05-07
AI Technical Summary
The existing video coding systems, particularly Versatile Video Coding (VVC), face challenges in efficiently determining the final position of sub-parts with non-zero residuals, leading to increased signaling overhead and reduced coding efficiency in subblock transform (SBT) processes.
The method involves implicitly determining the final position of sub-parts with non-zero residuals based on gradient values associated with the current block, using gradient filters and boundary matching, to apply inferred transforms and reduce signaling overhead.
This approach enhances coding efficiency by reducing signaling overhead and improving prediction accuracy in SBT processes, thereby optimizing video coding performance.
Smart Images

Figure CN2025128952_07052026_PF_FP_ABST
Abstract
Description
METHODS AND APPARATUS FOR IMPLICIT SUB-BLOCK TRANSFORM CODING BY GRADIENTS IN VIDEO CODING SYSTEMSCROSS REFERENCE TO RELATED APPLICATIONS
[0001] The present invention is a non-Provisional Application of and claims priority to U.S. Provisional Patent Application No. 63 / 712,558, filed on October 28, 2024. The U.S. Provisional Patent Application is hereby incorporated by reference in its entirety.FIELD OF THE INVENTION
[0002] The present invention relates to video coding system. In particular, the present invention relates to determine a final position of a sub-part with non-zero residuals implicitly or to determine the final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block in a video coding system. BACKGROUND AND RELATED ART
[0003] Versatile video coding (VVC) is the latest international video coding standard developed by the Joint Video Experts Team (JVET) of the ITU-T Video Coding Experts Group (VCEG) and the ISO / IEC Moving Picture Experts Group (MPEG) . The standard has been published as an ISO standard: ISO / IEC 23090-3: 2021, Information technology -Coded representation of immersive media -Part 3: Versatile video coding, published Feb. 2021. VVC is developed based on its predecessor HEVC (High Efficiency Video Coding) by adding more coding tools to improve coding efficiency and also to handle various types of video sources including 3-dimensional (3D) video signals.
[0004] Fig. 1A illustrates an exemplary adaptive Inter / Intra video encoding system incorporating loop processing. For Intra Prediction 110, the prediction data is derived based on previously coded video data in the current picture. For Inter Prediction 112, Motion Estimation (ME) is performed at the encoder side and Motion Compensation (MC) is performed based on the result of ME to provide prediction data derived from other picture (s) and motion data. Switch 114 selects Intra Prediction 110 or Inter Prediction 112 and the selected prediction data is supplied to Adder 116 to form prediction errors, also called residues. The prediction error is then processed by Transform (T) 118 followed by Quantization (Q) 120. The transformed and quantized residues are then coded by Entropy Encoder 122 to be included in a video bitstream corresponding to the compressed video data. The bitstream associated with the transform coefficients is then packed with side information such as motion and coding modes associated with Intra prediction and Inter prediction, and other information such as parameters associated with loop filters applied to underlying image area. The side information associated with Intra Prediction 110, Inter prediction 112 and in-loop filter 130, is provided to Entropy Encoder 122 as shown in Fig. 1A. When an Inter-prediction mode is used, a reference picture or pictures have to be reconstructed at the encoder end as well. Consequently, the transformed and quantized residues are processed by Inverse Quantization (IQ) 124 and Inverse Transformation (IT) 126 to recover the residues. The residues are then added back to prediction data 136 at Reconstruction (REC) 128 to reconstruct video data. The reconstructed video data may be stored in Reference Picture Buffer 134 and used for prediction of other frames.
[0005] As shown in Fig. 1A, incoming video data undergoes a series of processing in the encoding system. The reconstructed video data from REC 128 may be subject to various impairments due to a series of processing. Accordingly, in-loop filter 130 is often applied to the reconstructed video data before the reconstructed video data are stored in the Reference Picture Buffer 134 in order to improve video quality. For example, deblocking filter (DF) , Sample Adaptive Offset (SAO) and Adaptive Loop Filter (ALF) may be used. The loop filter information may need to be incorporated in the bitstream so that a decoder can properly recover the required information. Therefore, loop filter information is also provided to Entropy Encoder 122 for incorporation into the bitstream. In Fig. 1A, Loop filter 130 is applied to the reconstructed video before the reconstructed samples are stored in the reference picture buffer 134. The system in Fig. 1A is intended to illustrate an exemplary structure of a typical video encoder. It may correspond to the High Efficiency Video Coding (HEVC) system, VP8, VP9, H.264 or VVC.
[0006] The decoder, as shown in Fig. 1B, can use some of the functional blocks as the encoder. For example, the decoder can reuse Inverse Quantization 124 and Inverse Transform 126; however, Transform 118 and Quantization 120 are not needed at the decoder. Instead of Entropy Encoder 122, the decoder uses an Entropy Decoder 140 to decode the video bitstream into quantized transform coefficients and needed coding information (e.g. ILPF information, Intra prediction information and Inter prediction information) . The Intra prediction 150 at the decoder side does not need to perform the mode search. Instead, the decoder only needs to generate Intra prediction according to Intra prediction information received from the Entropy Decoder 140. Furthermore, for Inter prediction, the decoder only needs to perform motion compensation (MC 152) according to Inter prediction information received from the Entropy Decoder 140 without the need for motion estimation.
[0007] In VVC, the Sequence Parameter Set (SPS) and the Picture Parameter Set (PPS) contain high-level syntax elements that apply to entire coded video sequences and pictures, respectively. The Picture Header (PH) and Slice Header (SH) contain high-level syntax elements that apply to a current coded picture and a current coded slice, respectively.
[0008] In VVC, a coded picture is partitioned into non-overlapped square block regions represented by the associated coding tree units (CTUs) . A coded picture can be represented by a collection of slices, each comprising an integer number of CTUs. The individual CTUs in a slice are processed in raster-scan order. A bi-predictive (B) slice may be decoded using intra prediction or inter prediction with at most two motion vectors and reference indices to predict the sample values of each block. A predictive (P) slice is decoded using intra prediction or inter prediction with at most one motion vector and reference index to predict the sample values of each block. An intra (I) slice is decoded using intra prediction only.
[0009] I. 1 Partitioning of the CTUs Using a Tree Structure
[0010] In HEVC, a CTU is split into CUs by using a quaternary-tree (QT) 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 leaf CU level. Each leaf 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 leaf CU can be partitioned into transform units (TUs) according to another quaternary-tree 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.
[0011] In VVC, a quadtree with nested multi-type tree using binary and ternary splits segmentation structure replaces the concepts of multiple partition unit types, i.e. it removes the separation of the CU, PU and TU concepts except as needed for CUs that have a size too large for the maximum transform length, and supports more flexibility for CU partition shapes. In the coding tree structure, a CU can have either a square or rectangular shape. A coding tree unit (CTU) is first partitioned by a quaternary tree (a.k.a. quadtree) structure. Then the quaternary tree leaf nodes can be further partitioned by a multi-type tree structure. As shown in Fig. 2, there are four splitting types in multi-type tree structure, vertical binary splitting (SPLIT_BT_VER 210) , horizontal binary splitting (SPLIT_BT_HOR 220) , vertical ternary splitting (SPLIT_TT_VER 230) , and horizontal ternary splitting (SPLIT_TT_HOR 240) . The multi-type tree leaf nodes are called coding units (CUs) , and unless the CU is too large for the maximum transform length, this segmentation is used for prediction and transform processing without any further partitioning. This means that, in most cases, the CU, PU and TU have the same block size in the quadtree with nested multi-type tree coding block structure. The exception occurs when maximum supported transform length is smaller than the width or height of the colour component of the CU.
[0012] Fig. 3 illustrates the signalling mechanism of the partition splitting information in quadtree with nested multi-type tree coding tree structure. A coding tree unit (CTU) is treated as the root of a quaternary tree and is first partitioned by a quaternary tree structure. Each quaternary tree leaf node (when sufficiently large to allow it) is then further partitioned by a multi-type tree structure. In quadtree with nested multi-type tree coding tree structure, for each CU node, a first flag (split_cu_flag) is signalled to indicate whether the node is further partitioned. If the current CU node is a quadtree CU node, a second flag (split_qt_flag) whether it's a QT partitioning or MTT partitioning mode. When a node is partitioned with MTT partitioning mode, a third flag (mtt_split_cu_vertical_flag) is signalled to indicate the splitting direction, and then a fourth flag (mtt_split_cu_binary_flag) is signalled to indicate whether the split is a binary split or a ternary split. Based on the values of mtt_split_cu_vertical_flag and mtt_split_cu_binary_flag, the multi-type tree slitting mode (MttSplitMode) of a CU is derived as shown in Table 1. Table 1 -MttSplitMode derivation based on multi-type tree syntax elements
[0013] Fig. 4 shows a CTU divided into multiple CUs with a quadtree and nested multi-type tree coding block structure, where the bold block edges represent quadtree partitioning and the remaining edges represent multi-type tree partitioning. The quadtree with nested multi-type tree partition provides a content-adaptive coding tree structure comprised of CUs. The size of the CU may be as large as the CTU or as small as 4×4 in units of luma samples. For the case of the 4: 2: 0 chroma format, the maximum chroma CB size is 64×64 and the minimum size chroma CB consist of 16 chroma samples.
[0014] In VVC, the maximum supported luma transform size is 64×64 and the maximum supported chroma transform size is 32×32. When the width or height of the CB is larger the maximum transform width or height, the CB is automatically split in the horizontal and / or vertical direction to meet the transform size restriction in that direction.
[0015] The following parameters are defined for the quadtree with nested multi-type tree coding tree scheme. These parameters are specified by SPS syntax elements and can be further refined by picture header syntax elements. – CTU size: the root node size of a quaternary tree – MinQTSize: the minimum allowed quaternary tree leaf node size – MaxBtSize: the maximum allowed binary tree root node size – MaxTtSize: the maximum allowed ternary tree root node size – MaxMttDepth: the maximum allowed hierarchy depth of multi-type tree splitting from a quadtree leaf – MinCbSize: the minimum allowed coding block node size
[0016] In one example of the quadtree with nested multi-type tree coding tree structure, the CTU size is set as 128×128 luma samples with two corresponding 64×64 blocks of 4: 2: 0 chroma samples, the MinQTSize is set as 16×16, the MaxBtSize is set as 128×128 and MaxTtSize is set as 64×64, the MinCbsize (for both width and height) is set as 4×4, and the MaxMttDepth is set as 4. The quaternary tree partitioning is applied to the CTU first to generate quaternary tree leaf nodes. The quaternary tree leaf nodes may have a size from 16×16 (i.e., the MinQTSize) to 128×128 (i.e., the CTU size) . If the leaf QT node is 128×128, it will not be further split by the binary tree since the size exceeds the MaxBtSize and MaxTtSize (i.e., 128×128) . Otherwise, the leaf qdtree node could be further partitioned by the multi-type tree. Therefore, the quaternary tree leaf node is also the root node for the multi-type tree and it has multi-type tree depth (mttDepth) as 0. When the multi-type tree depth reaches MaxMttDepth (i.e., 4) , no further splitting is considered. When the multi-type tree node has width equal to MinCbsize, no further horizontal splitting is considered. Similarly, when the multi-type tree node has height equal to MinCbsize, no further vertical splitting is considered.
[0017] In VVC, the coding tree scheme supports the ability for the luma and chroma to have a separate block tree structure. For P and B slices, the luma and chroma CTBs in one CTU have to share the same coding tree structure. However, for I slices, the luma and chroma can have separate block tree structures. When the separate block tree mode is applied, luma CTB is partitioned into CUs by one coding tree structure, and the chroma CTBs are partitioned into chroma CUs by another coding tree structure. This means that a CU in an I slice may consist of a coding block of the luma component or coding blocks of two chroma components, and a CU in a P or B slice always consists of coding blocks of all three colour components unless the video is monochrome.
[0018] I. 2 Inter Prediction
[0019] For each inter-predicted CU, motion parameters consisting of motion vectors, reference picture indices and reference picture list usage index, and additional information needed for the new coding feature of VVC to be used for inter-predicted sample generation. The motion parameter can be signalled in an explicit or implicit manner. When a CU is coded with skip mode, the CU is associated with one PU and has no significant residual coefficients, nor coded motion vector delta or reference picture index. A merge mode is specified whereby the motion parameters for the current CU are obtained from neighbouring CUs, including spatial and temporal candidates, and additional schedules introduced in VVC. The merge mode can be applied to any inter-predicted CU, 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 flag and other needed information are signalled explicitly per each CU.
[0020] I. 3 Large Block-Size Transforms with High-Frequency Zeroing
[0021] In VVC, large block-size transforms, up to 64×64 in size, are enabled, which is primarily useful for higher resolution video, such as 1080p and 4K sequences. High frequency transform coefficients are zeroed out for the transform blocks with size (width or height, or both width and height) equal to 64, so that only the lower-frequency coefficients are retained. For example, for an M×N transform block, with M as the block width and N as the block height, when M is equal to 64, only the left 32 columns of transform coefficients are kept. Similarly, when N is equal to 64, only the top 32 rows of transform coefficients are kept. When transform skip mode is used for a large block, the entire block is used without zeroing out any values. In addition, transform shift is removed in transform skip mode. The VTM also supports configurable max transform size in SPS, such that encoder has the flexibility to choose up to 32-length or 64-length transform size depending on the need of specific implementation. VTM (Versatile Video Coding Test Model) is the reference software for the VVC.
[0022] I. 4 Subblock Transform (SBT)
[0023] In VVC, subblock transform is introduced for an inter-predicted CU. In SBT, only a sub-part of the residual block is coded for the CU. When inter-predicted CU with cu_coded_flag equal to 1, cu_sbt_flag may be signalled to indicate whether the whole residual block or a sub-part of the residual block is coded with transformation process. In the former case, inter MTS information is further parsed to determine the transform type of the CU. In the latter case, a part of the residual block is adaptively coded with inferred transform type by side (e.g. left or right side, top or bottom side) and the other part of the residual block is zeroed out.
[0024] When SBT is used for an inter-coded CU, SBT type and SBT position information are signalled in the bitstream. SBT type information indicates the TU split types (i.e., split likes a binary tree split or an asymmetric binary tree split) and the split direction (i.e., horizontal split or vertical split) , where the semantic names are cu_sbt_quad_flag and cu_sbt_horizontal_flag in VVC. SBT position information indicates which TU has non-zero residual, where the semantic name is cu_sbt_pos_flag in VVC. For example, two SBT types and two SBT positions are illustrated in Fig. 5. For SBT-V 510 and 520 (or SBT-H 530 and 540) , the TU width (or height) may equal to half of the CU width (or height) or 1 / 4 of the CU width (or height) , resulting in 2: 2 split or 1: 3 / 3: 1 split. The 2: 2 split is like a binary tree (BT) split while the 1: 3 / 3: 1 split is like an asymmetric binary tree (ABT) split. In ABT splitting, only the small region contains the non-zero residual. If one dimension of a CU is 8 in luma samples, the 1: 3 / 3: 1 split along that dimension is disallowed. There are at most 8 SBT modes for a CU.
[0025] Position-dependent transform core selection is applied on luma transform blocks in SBT-V and SBT-H (chroma TB always using DCT-2) . The two positions of SBT-H and SBT-V are associated with different core transforms. More specifically, the horizontal and vertical transforms for each SBT position is specified in Fig. 5. For example, the horizontal and vertical transforms for SBT-V position 0 is DCT-8 and DST-7, respectively. When one side of the residual TU is greater than 32, the transform for both dimensions is set as DCT-2. Therefore, the subblock transform jointly specifies the TU tiling, cbf, and horizontal and vertical core transform type of a residual block. Note, the SBT is not applied to the CU coded with combined inter-intra mode in VVC.
[0026] To reduce encoder run-time, some fast algorithms have been developed in RDO (Rate-Distortion Optimization) of SBT: - Algorithm 1: For each SBT mode, a RD (Rate-Distortion) cost is estimated based on the SSD of the residual-skipped part. An SBT mode is skipped in RDO if the estimated RD cost of the SBT mode is larger than the actual RD cost of the best mode. In addition, only the best 4 SBT modes in terms of the estimated RD cost are tried in RDO. - Algorithm 2: a transform mode save and load is applied (which is improved from that proposed in JVET-K0358) . The residual energy (i.e., Sum of Squared Differences (SSD) ) and the best transform mode (one among whole block transform with DCT-2, whole block transform with inter MTS (Multiple Transform Selection) , and sub-block transform) of a PU is saved as history information. When the residual energy of a PU matches a previous case, the best transform mode associated with the residual energy is tried while the other transform modes are skipped. This fast algorithm reduces the encoding time of both SBT and inter MTS. - Algorithm 3: if the RD cost of whole residual block being transformed by DCT-2 is much worse than the current best RD cost, the SBT is skipped. - Algorithm 4: if the RD cost of whole residual block being transformed by DCT-2 is small enough, the SBT is skipped.
[0027] In the present invention, methods and apparatus to determine a final position of a sub-part with non-zero residuals implicitly are disclosed to reduce signalling overhead for the final position of the sub-part with non-zero residuals. BRIEF SUMMARY OF THE INVENTION
[0028] A method and apparatus for video coding using SBT (Subblock Transform) mode are disclosed. According to the method, input data associated with a current block is received, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is split into multiple sub-parts according to an SBT (Subblock Transform) mode. A final position of a sub-part with non-zero residuals in the current block is determined implicitly based on one or more gradient values associated with the current block, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type. Inferred transforms are applied to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.
[0029] In one embodiment, one or more gradient filters are applied to current prediction to derive said one or more gradient values.
[0030] In one embodiment, the final position of the sub-part with the non-zero residuals corresponds to an area covering one or more maximum gradient values. In one embodiment, the final position of the sub-part with the non-zero residuals corresponds to an area having a maximum sum of gradients, and wherein a sum of gradients is calculated by summing said one or more gradient values over the area. In one embodiment, the sum of gradients is calculated partially from an offset location of the area with a first step size in a horizontal direction and / or a second step size in a vertical direction, and wherein the first step size and the second step size are integers greater than 1. In one embodiment, the sum of gradients is calculated based on said one or more gradient values greater than a threshold.
[0031] In one embodiment, the final position of the sub-part with the non-zero residuals is selected based on said one or more gradient values at a specified location for each of the multiple sub-parts, and a target sub-part having a largest gradient value at the specified location is selected as the final position of the sub-part with the non-zero residuals. In one embodiment, the specified location for each of the multiple sub-parts corresponds to a centre position or a nearest position to the centre position for each of the multiple sub-parts.
[0032] In one embodiment, said one or more gradient values are calculated from a reference block having a shortest temporal reference picture distance, less quantization parameter, or shortest spatial distance. In one embodiment, the reference block is from short-term reference pictures.
[0033] In one embodiment, if N candidate residual block positions are allowed for the current block, first k out of the N candidate residual block positions with highest sum of gradients are chosen first and the final position of the sub-part with the non-zero residuals is selected from the k candidate residual block positions, and wherein N is an integer greater than or equal to 2 and k is a positive integer smaller than N. In one embodiment, if a target candidate residual block having block boundaries aligned with deblocking grid size, the target candidate residual block is eligible to be chosen. In one embodiment, a syntax is signalled or parsed to indicate selection of the final position of the sub-part with the non-zero residuals from the k candidate residual block positions.
[0034] In one embodiment, if N candidate residual block sizes or areas are allowed for the current block and each of the N candidate residual block sizes or areas is determined implicitly according to a position condition, a final block size or area is selected from k candidate residual block sizes or areas out of the N candidate residual block sizes or areas, and wherein N is an integer greater than or equal to 2 and k is a positive integer smaller than N. In one embodiment, the position condition comprises each of the N candidate residual block sizes or areas covering a best position corresponding to a maximum gradient value of the current block. In one embodiment, a syntax is signalled or parsed to indicate selection of the final position of the sub-part with the non-zero residuals from the k candidate residual block positions.
[0035] According to another method, input data associated with a current block is receiving, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is coded using an SBT (Subblock Transform) mode. A final position of a sub-part with non-zero residuals in the current block is determined by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type. Inferred transforms is applied to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.BRIEF DESCRIPTION OF THE DRAWINGS
[0036] Fig. 1A illustrates an exemplary adaptive Inter / Intra video coding system incorporating loop processing.
[0037] Fig. 1B illustrates a corresponding decoder for the encoder in Fig. 1A.
[0038] Fig. 2 illustrates examples of a multi-type tree structure corresponding to vertical binary splitting (SPLIT_BT_VER) , horizontal binary splitting (SPLIT_BT_HOR) , vertical ternary splitting (SPLIT_TT_VER) , and horizontal ternary splitting (SPLIT_TT_HOR) .
[0039] Fig. 3 illustrates an example of the signalling mechanism of the partition splitting information in quadtree with nested multi-type tree coding tree structure.
[0040] Fig. 4 shows an example of a CTU divided into multiple CUs with a quadtree and nested multi-type tree coding block structure, where the bold block edges represent quadtree partitioning and the remaining edges represent multi-type tree partitioning.
[0041] Fig. 5 shows position-dependent transform core selection for SBT.
[0042] Fig. 6A-D shows an example for regions (i.e., a to f) used to calculate the boundary matching difference value.
[0043] Fig. 7 shows various examples of gradient filters for calculating the gradient of the current block, where the dash circle is the centre position to apply a filter.
[0044] Fig. 8A illustrates an example of an area containing gradient values.
[0045] Fig. 8B illustrates an example of current block size being w×h, the inner rectangle being the residual block with size m×n at the left-top position (x, y) .
[0046] Fig. 9 illustrates a flowchart of an exemplary video coding system that determines a final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block according to an embodiment of the present invention.
[0047] Fig. 10 illustrates a flowchart of an exemplary video coding system that determines a final position of a sub-part with non-zero residuals in the current block by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size according to an embodiment of the present invention.DETAILED DESCRIPTION OF THE INVENTION
[0048] It will be readily understood that the components of the present invention, as generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the following more detailed description of the embodiments of the systems and methods of the present invention, as represented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. References throughout this specification to “one embodiment, ” “an embodiment, ” or similar language mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present invention. Thus, appearances of the phrases “in one embodiment” or “in an embodiment” in various places throughout this specification are not necessarily all referring to the same embodiment.
[0049] Furthermore, the described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. One skilled in the relevant art will recognize, however, that the invention can be practiced without one or more of the specific details, or with other methods, components, etc. In other instances, well-known structures, or operations are not shown or described in detail to avoid obscuring aspects of the invention. The illustrated embodiments of the invention will be best understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of apparatus and methods that are consistent with the invention as claimed herein.
[0050] In order to improve the prediction accuracy or coding performance of cross-component prediction, various schemes related to inheriting cross-component models are disclosed.
[0051] PROPOSED METHODS
[0052] Several methods are proposed to reduce the signalling overhead related to SBT or to improve the coding efficiency for SBT.
[0053] The final position of the sub-part of a residual block in SBT (e.g. cu_sbt_pos_flag in VVC) can be implicitly derived. In one embodiment, the final position is determined through boundary matching, by evaluating the boundary matching cost between the current prediction samples and the neighbouring reconstructed samples of each sub-part TU. If a sub-part TU has the maximum boundary matching difference compared to the other sub-part TUs, it is implicitly inferred to have non-zero residuals and should undergo the transform or inverse transform process using inferred transforms, while the other sub-part TU is assumed to have all-zero residuals.
[0054] For example, if the current SBT type is SBT-H with BT split (e.g. as shown in Fig. 6A) , the boundary matching difference of TU “C” is diffC=∑|recoe-predf|, where recoe is the neighbouring reconstruction samples of “e” , and predf is the prediction samples of “f” . Similarly, the boundary matching difference of TU “D” is diffD=∑|recog-predh|, where recog is the neighbouring reconstruction samples of “g” , and predh is the prediction samples of “h” . If diffC>diffD, TU “A” is the sub-part with non-zero residuals. Otherwise, if diffD>diffC, TU “B” is the sub-part with non-zero residuals.
[0055] For another example, if the current SBT type is SBT-V with BT split (e.g. as shown in Fig. 6C) , the boundary matching difference of TU “A” is diffA=∑|recoa-predb|, where recoa is the neighbouring reconstruction samples of “a” , and predb is the prediction samples of “b” . Similarly, the boundary matching difference of TU “B” is diffB=∑|recoc-predd|, where recoc is the neighbouring reconstruction samples of “c” , and predd is the prediction samples of “d” . If diffA>diffB, TU “A” is the sub-part with non-zero residuals. Otherwise, if diffB>diffA, TU “B” is the sub-part with non-zero residuals.
[0056] For another example, if the current SBT type is SBT-H with ABT split (e.g. as shown in Fig. 6B) , the boundary matching difference of TU “E” is diffE=∑|recoi-predj|, where recoi is the neighbouring reconstruction samples of “i” , and predj is the prediction samples of “j” . Similarly, the boundary matching difference of TU “F” is diffF=∑|recoo-predp|, where recoo is the neighbouring reconstruction samples of “o” , and predp is the prediction samples of “p” . If diffE>diffF, TU “E” is the sub-part with non-zero residuals. Otherwise, if diffF>diffE, TU “F” is the sub-part with non-zero residuals. In this example, if ∑|recok-predl|>∑|recom-predn|, TU “F” is the sub-part with non-zero residuals. If ∑|recok-predl|<∑|recom-predn|, TU “E” is the sub-part with non-zero residuals, where recok is the neighbouring reconstruction samples of “k” , predl is the prediction samples of “l” , recom is the neighbouring reconstruction samples of “m” , and predn is the prediction samples of “n” .
[0057] For another example, if the current SBT type is SBT-V with ABT split (e.g. as shown in Fig. 6D) , the boundary matching difference of TU “G” is diffG=∑|recoq-predr|, where recoq is the neighbouring reconstruction samples of “q” , and predr is the prediction samples of “r” . Similarly, the boundary matching difference of TU “H” is diffH=∑|recow-predx|, where recow is the neighbouring reconstruction samples of “w” , and predx is the prediction samples of “x” . If diffG>diffH, TU “G” is the sub-part with non-zero residuals. Otherwise, if diffH>diffG, TU “H” is the sub-part with non-zero residuals. In this example, if ∑|recos-predt|>∑|recou-predv|, TU “H” is the sub-part with non-zero residuals. If ∑|recos-predt|<∑|recou-predv|, TU “G” is the sub-part with non-zero residuals, where recos is the neighbouring reconstruction samples of “s” , predt is the prediction samples of “t” , recou is the neighbouring reconstruction samples of “u” , and predv is the prediction samples of “v” .
[0058] In another embodiment, the final position of the sub-part with non-zero residuals in SBT can be implicitly derived by the gradient of the current block. For example, by applying gradient filter to the current prediction and given the residual block size in SBT, the location sub-part of a residual block in SBT is the area can cover the maximum gradient values. For another example, the gradient filter can be but not limited to the examples in the following figure. The gradient value of a position of the current block can be one or more combination of gradient filters (e.g. the combination of a. 1 and a. 2, b. 1 and b. 2, e. 1 and e. 2, or f. 1 and f. 2 in Fig. 7) . The gradient filter selection can also depend on the valid of neighbouring samples of the centre position or the position inside the current block. For example, if the left samples are not valid, the filters e. 1 and / or e. 2 in Fig. 7 are used to calculate gradient values. If the above samples are not valid, the filters c. 1, c. 2, d. 1, and / or d. 2 in Fig. 7 are used to calculate gradient values. For still another example, the sample padding or sample repeat is used to represent the invalid of neighbouring samples when applying a gradient filter. In still another embodiment, variance of each sample locations calculate by the current prediction of the current block can be used to determine the final position of the sub-part with non-zero residuals in SBT. In still another embodiment, a high-pass filter is applied to each sample location of the current prediction to determine the final position of the sub-part with non-zero residuals in SBT.
[0059] In still the same embodiment, by applying gradient filter (s) to the current prediction and given the residual block size in SBT, the location of the residual block is identified by the area containing the maximum sum of gradient values. For example, as shown in Fig. 8A and Fig. 8B, the current block width and height are denoted as w and h, the gradient value at position (i, j) inside the current block is denoted as Gi, j, where 0≤i<w and 0≤j<h. Besides, given the size of the sub-part with residual is m×n, the left-top position (i.e., denoted as (x, y) ) of the sub-part with residual inside the current block has the sum of gradient values equal to The target is to implicitly derive a location (x, y) *that has the maximum as the final position of the sub-part with residual, where 0≤x< (w-m) and 0≤y< (h-n) . In still another embodiment, when summing the gradient values, it can start from an offset position and then step to the next position with a distance greater than 1 sample. For example, if the start offset position is u, and the step distance is v, then the selected horizontal positions are x+u, x+u+v, x+u+2×v, x+u+3×v, …, etc. In the vertical direction, if the start offset position is u’ , and the step distance is v’ , then the selected vertical positions are y+u′, y+u′+v′, y+u′+2×v′, y+u′+3×v′, …, etc.
[0060] In another embodiment, all gradient values can be summed horizontally as a 1-D matrix (e.g., and ) , then the final top position of the sub-part with residual inside the current block is the position that has the maximum sum of gradient values in the 1-D matrix (e.g. Ghor) . For example, the target above position (i.e., denoted as y) of the sub-part with residual inside the current block has the maximum where 0≤y< (h-n) . Similarly, all gradient values can be summed vertically as a 1-D matrix (e.g. and ) , then the final left position of the sub-part with residual inside the current block is the position that has the maximum sum of gradient values in the 1-D matrix (e.g., Gver) . For example, the target left position (i.e., denoted as x) of the sub-part with residual inside the current block has the maximum where 0≤x< (w-m) . Similarly, to reduce computations, only partial gradient values are selected to calculate the sum of gradient values. It can select gradient values with a step distance (e.g. start from a starting position, and select one gradient value for every k values) . In still another embodiment, only the gradient values greater than or equal to a threshold are used to calculate the sum of gradient values. The threshold value can be a predefined value, a gradient value at a specified position inside or outside the current block, or the mean value of selected gradient values of the current block.
[0061] In still another embodiment, the location of the residual block is a specified location that has the maximum gradient value. For example, the target is to identify a location that has the maximum gradient value at the centre position of the current block. The centre position of a sub-part can be defined at (x + (m / 2) -1, y + (n / 2) -1) , (x + (m / 2) -1, y + (n / 2) ) , (x + (m / 2) , y + (n / 2) -1) , (x + (m / 2) , y + (n / 2) ) , or their combinations. For another example, the target is to identify a location that has the maximum gradient value at the left-top position (i.e., (x, y) *) of the sub-part. For still another example, the target is to identify a location that has the maximum gradient value at the left-bottom position (i.e., (x, y + n -1) ) of the sub-part. For still another example, the target is to identify a location that has the maximum gradient value at the right-top position (i.e., (x + m -1, y) ) of the sub-part. For still another example, the target is to identify a location that has the maximum gradient value at the right-bottom position (i.e., (x + m -1, y + n -1) ) of the sub-part.
[0062] Moreover, the selection of x and y (i.e., the left-top position of the residual block inside the current block) can be subject to the deblocking grid size. In one embodiment, if the start position of the deblocking grid is at the left-top position of the current block and the deblocking grid size is 4x4, the selected x and y shall be the multiples of 4. In another embodiment, if the start position of the deblocking grid is at the left-top position of the current block and the deblocking grid size is 8x8, the selected x and y shall be the multiples of 8.
[0063] In another embodiment, suppose the left-top position of the current block inside the current picture is (s, t) , the left-top position of the residual block inside the current block is (s+x, t + y) , and if deblocking grid size is 4x4, the selected (s+ x) and (t + y) , and (s+ x + m) and (t + y + n) shall be multiples of 4. For another example, if deblocking grid size is 8x8, the selected (s + x) and (t + y) , and (s+ x + m) and (t + y + n) shall be the multiples of 8.
[0064] If the current SBT block is coded in inter mode, or the current SBT block has associated motion information, one or more reference blocks can be located based on the motion information of the current block. Traditionally, the prediction of the current block is the combination of these reference blocks. However, to reduce the possible latency or computation, the prediction of the current block used to calculate gradient can be from the reference block that has the shortest temporal reference picture distance, less quantization parameter, or shortest spatial distance. In one embodiment, the reference block is from short-term reference pictures rather than long-term reference pictures.
[0065] If two or more residual block positions are allowed for the current block (assuming there are N candidate block positions in total) , the first k out of N candidates with higher gradient sums are chosen. Then, the final position out of these k candidate positions (e.g., k equal to 2, 3, 4, …, or N-1) is signalled in the bitstream.
[0066] If two or more residual block sizes / areas are allowed for the current block (assuming there are N candidate block sizes / areas in total) , and each of the N candidate residual block sizes or areas is determined implicitly according to a position condition (e.g. the position condition comprising a specified block size / area covering the maximum gradient value in the current block) , and the final block size / area along with the associated position of these k candidate sizes / areas (e.g. k equal to 2, 3, 4, …, or N-1) is signalled in the bitstream. In a special case, if only one block size / area is valid for the current SBT encoding or decoding, k is not signalled and the best position of the current residual block is implicitly determined. In other words, in this special case, only SBT on / off flag is signalled, and the residual block position and size are implicitly determined.
[0067] If two or more residual block positions associated with multiple block sizes / areas are allowed for the current block (assuming there are N candidates in total) , the first k out of N candidates with higher gradient sums are chosen. Then, the final position associated with a derived block size / area out of these k candidate positions (e.g., k equal to 2, 3, 4, …, or N-1) is signalled in the bitstream.
[0068] In another embodiment, a cost is calculated based on the boundary matching difference value and the sum of gradient values for each of candidate residual block positions associated with multiple block sizes / areas (assuming there are N candidates in total) , the first k out of N candidates with higher costs are chosen. Then, the final position associated with a derived block size / area out of these k candidate positions (e.g. k equal to 2, 3, 4, …, or N-1) is signalled in the bitstream.
[0069] The methods of deriving a final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block or by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size as described above can be implemented in an encoder side or a decoder side. For example, any of the proposed candidate derivation method can be implemented in an Intra / Inter coding module (e.g. Intra Pred. 150 / MC 152 in Fig. 1B) in a decoder or an Intra / Inter coding module is an encoder (e.g. Intra Pred. 110 / Inter Pred. 112 in Fig. 1A) . Any of the proposed shared buffer to store coding information among multiple coding tools including the CCM mode can also be implemented as a circuit coupled to the intra / inter coding module at the decoder or the encoder. However, the decoder or encoder may also use additional processing unit to implement the required cross-component prediction processing. While the Intra Pred. units (e.g. unit 110 / 112 in Fig. 1A and unit 150 / 152 in Fig. 1B) are shown as individual processing units, they may correspond to executable software or firmware codes stored on a media, such as hard disk or flash memory, for a CPU (Central Processing Unit) or programmable devices (e.g. DSP (Digital Signal Processor) or FPGA (Field Programmable Gate Array) ) .
[0070] Fig. 9 illustrates a flowchart of an exemplary video coding system that determines a final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block according to an embodiment of the present invention. The steps shown in the flowchart may be implemented as program codes executable on one or more processors (e.g., one or more CPUs) at the encoder side. The steps shown in the flowchart may also be implemented based hardware such as one or more electronic devices or processors arranged to perform the steps in the flowchart. According to the method, input data associated with a current block is received in step 910, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is split into multiple sub-parts according to an SBT (Subblock Transform) mode. A final position of a sub-part with non-zero residuals in the current block is determined implicitly based on one or more gradient values associated with the current block in step 920, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type. Inferred transforms are applied to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side in step 930.
[0071] Fig. 10 illustrates a flowchart of an exemplary video coding system that determines a final position of a sub-part with non-zero residuals in the current block by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size according to an embodiment of the present invention. According to this method, input data associated with a current block is receiving in step 1010, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is coded using an SBT (Subblock Transform) mode. A final position of a sub-part with non-zero residuals in the current block is determined by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size in step 1020, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type. Inferred transforms according to the target SBT is applied to the sub-part with the non-zero residuals in the current block to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side in step 1030.
[0072] The flowcharts shown are intended to illustrate an example of video coding according to the present invention. A person skilled in the art may modify each step, re-arranges the steps, split a step, or combine steps to practice the present invention without departing from the spirit of the present invention. In the disclosure, specific syntax and semantics have been used to illustrate examples to implement embodiments of the present invention. A skilled person may practice the present invention by substituting the syntax and semantics with equivalent syntax and semantics without departing from the spirit of the present invention.
[0073] The above description is presented to enable a person of ordinary skill in the art to practice the present invention as provided in the context of a particular application and its requirement. Various modifications to the described embodiments will be apparent to those with skill in the art, and the general principles defined herein may be applied to other embodiments. Therefore, the present invention is not intended to be limited to the particular embodiments shown and described, but is to be accorded the widest scope consistent with the principles and novel features herein disclosed. In the above detailed description, various specific details are illustrated in order to provide a thorough understanding of the present invention. Nevertheless, it will be understood by those skilled in the art that the present invention may be practiced.
[0074] Embodiment of the present invention as described above may be implemented in various hardware, software codes, or a combination of both. For example, an embodiment of the present invention can be one or more circuit circuits integrated into a video compression chip or program code integrated into video compression software to perform the processing described herein. An embodiment of the present invention may also be program code to be executed on a Digital Signal Processor (DSP) to perform the processing described herein. The invention may also involve a number of functions to be performed by a computer processor, a digital signal processor, a microprocessor, or field programmable gate array (FPGA) . These processors can be configured to perform particular tasks according to the invention, by executing machine-readable software code or firmware code that defines the particular methods embodied by the invention. The software code or firmware code may be developed in different programming languages and different formats or styles. The software code may also be compiled for different target platforms. However, different code formats, styles and languages of software codes and other means of configuring code to perform the tasks in accordance with the invention will not depart from the spirit and scope of the invention.
[0075] The invention may be embodied in other specific forms without departing from its spirit or essential characteristics. The described examples are to be considered in all respects only as illustrative and not restrictive. The scope of the invention is therefore, indicated by the appended claims rather than by the foregoing description. All changes which come within the meaning and range of equivalency of the claims are to be embraced within their scope.
Claims
1.A method of video coding, the method comprising:receiving input data associated with a current block, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is split into multiple sub-parts according to an SBT (Subblock Transform) mode;determining a final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type; andapplying inferred transforms to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.2.The method of Claim 1, wherein one or more gradient filters are applied to current prediction to derive said one or more gradient values.3.The method of Claim 2, wherein the final position of the sub-part with the non-zero residuals corresponds to an area covering one or more maximum gradient values.4.The method of Claim 2, wherein the final position of the sub-part with the non-zero residuals corresponds to an area having a maximum sum of gradients, and wherein a sum of gradients is calculated by summing said one or more gradient values over the area.5.The method of Claim 4, wherein the sum of gradients is calculated partially from an offset location of the area with a first step size in a horizontal direction and / or a second step size in a vertical direction, and wherein the first step size and the second step size are integers greater than 1.6.The method of Claim 4, wherein the sum of gradients is calculated based on said one or more gradient values greater than a threshold.7.The method of Claim 2, wherein the final position of the sub-part with the non-zero residuals is selected based on said one or more gradient values at a specified location for each of the multiple sub-parts, and a target sub-part having a largest gradient value at the specified location is selected as the final position of the sub-part with the non-zero residuals.8.The method of Claim 7, wherein the specified location for each of the multiple sub-parts corresponds to a centre position or a nearest position to the centre position for each of the multiple sub-parts.9.The method of Claim 2, wherein said one or more gradient values are calculated from a reference block having a shortest temporal reference picture distance, less quantization parameter, or shortest spatial distance.10.The method of Claim 9, wherein the reference block is from short-term reference pictures.11.The method of Claim 2, wherein if N candidate residual block positions are allowed for the current block, first k out of the N candidate residual block positions with highest sum of gradients are chosen first and the final position of the sub-part with the non-zero residuals is selected from the k candidate residual block positions, and wherein N is an integer greater than or equal to 2 and k is a positive integer smaller than N.12.The method of Claim 11, wherein if a target candidate residual block having block boundaries aligned with deblocking grid size, the target candidate residual block is eligible to be chosen.13.The method of Claim 11, wherein a syntax is signalled or parsed to indicate selection of the final position of the sub-part with the non-zero residuals from the k candidate residual block positions.14.The method of Claim 2, wherein if N candidate residual block sizes or areas are allowed for the current block and each of the N candidate residual block sizes or areas is determined implicitly according to a position condition, a final block size or area is selected from k candidate residual block sizes or areas out of the N candidate residual block sizes or areas, and wherein N is an integer greater than or equal to 2 and k is a positive integer smaller than N.15.The method of Claim 14, wherein the position condition comprises each of the N candidate residual block sizes or areas covering a best position corresponding to a maximum gradient value of the current block.16.The method of Claim 14, wherein a syntax is signalled or parsed to indicate selection of the final position of the sub-part with the non-zero residuals from the k candidate residual block positions.17.An apparatus of video coding, the apparatus comprising one or more electronics or processors arranged to:receive input data associated with a current block, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is split into multiple sub-parts according to an SBT (Subblock Transform) mode;determine a final position of a sub-part with non-zero residuals in the current block implicitly based on one or more gradient values associated with the current block, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type; andapply inferred transforms to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.18.A method of video coding, the method comprising:receiving input data associated with a current block, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is coded using an SBT (Subblock Transform) mode;determining a final position of a sub-part with non-zero residuals in the current block by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type; andapplying inferred transforms to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.19.An apparatus of video coding, the apparatus comprising one or more electronics or processors arranged to:receive input data associated with a current block, wherein the input data comprises residual data to be encoded at an encoder side or transformed residual data associated with the current block to be decoded at a decoder side, wherein the current block is coded using an SBT (Subblock Transform) mode;determine a final position of a sub-part with non-zero residuals in the current block by aligning the final position of the sub-part with the non-zero residuals with deblocking grid size, wherein the final position of the sub-part with the non-zero residuals is associated with a target SBT type; andapply inferred transforms to the sub-part with the non-zero residuals in the current block according to the target SBT to derive the transformed residual data at the encoder side or reconstructed residual data at the decoder side.
Citation Information
Patent Citations
Partitions on sub-block transform mode
CN113574880A
Sub-block transform in transform skip mode
CN113632493A
Implicit Transform Settings
US20190320203A1
Methods and apparatus for implicit sub-block transform coding
WO2023246901A1