EP4562851A1

EP4562851A1 - Method and apparatus of improving performance of convolutional cross-component model in video coding system

Info

Publication number: EP4562851A1
Application number: EP23845598.4A
Authority: EP
Inventors: Cheng-Yen Chuang; Ching-Yeh Chen; Chih-Wei Hsu; Tzu-Der Chuang
Original assignee: MediaTek Inc
Current assignee: MediaTek Inc
Priority date: 2022-07-27
Filing date: 2023-07-26
Publication date: 2025-06-04
Also published as: WO2024022390A1; TW202406341A; CN119547432A

Abstract

Methods and apparatus of CCCM for video coding are provided. According to one method, an auto-correlation matrix associated with a footprint of a convolutional filter for reference luma samples in a reference area is derived. A cross-correlation vector, associated with the footprint, between the reference luma samples and reference chroma samples in the reference area is derived. Coefficients of the convolutional filter are derived based on the auto-correlation matrix and the cross-correlation vector using Gaussian elimination scheme. For a target chroma sample in the chroma block, the convolutional filter with the coefficients derived is applied to a corresponding location of the luma block to form a convolutional cross-component model predictor for the target chroma sample. A final predictor is generated for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor. The target chroma sample is encoded or decoded using the final predictor.

Description

METHOD AND APPARATUS OF IMPROVING PERFORMANCE OF CONVOLUTIONAL CROSS-COMPONENT MODEL IN VIDEO CODING SYSTEM

CROSS REFERENCE TO RELATED APPLICATIONS
The present invention is a non-Provisional Application of and claims priority to U.S. Provisional Patent Application No. 63/369,524, filed on July 27, 2022. The U.S. Provisional Patent Application is hereby incorporated by reference in its entirety.

FIELD OF THE INVENTION

The present invention relates to video coding system. In particular, the present invention relates to schemes to improve performance or reducing the complexity of CCLM (Convolutional Cross-Component Model) related modes in a video coding system.
BACKGROUND AND RELATED ART
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.
Fig. 1A illustrates an exemplary adaptive Inter/Intra video coding system incorporating loop processing. For Intra Prediction, 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 of 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, are 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.
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.
The decoder, as shown in Fig. 1B, can use similar or portion of the same functional blocks as the encoder except for Transform 118 and Quantization 120 since the decoder only needs Inverse Quantization 124 and Inverse Transform 126. 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.
According to VVC, an input picture is partitioned into non-overlapped square block regions referred as CTUs (Coding Tree Units) , similar to HEVC. Each CTU can be partitioned into one or multiple smaller size coding units (CUs) . The resulting CU partitions can be in square or rectangular shapes. Also, VVC divides a CTU into prediction units (PUs) as a unit to apply prediction process, such as Inter prediction, Intra prediction, etc.
Partitioning of the CTUs Using a Tree Structure
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.
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.
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) is signalled to indicate 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
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.
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.
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
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., 64×64) . 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 vertical splitting is considered. Similarly, when the multi-type tree node has height equal to MinCbsize, no further horizontal splitting is considered.
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.
Virtual Pipeline Data Units (VPDUs)
Virtual pipeline data units (VPDUs) are defined as non-overlapping units in a picture. In hardware decoders, successive VPDUs are processed by multiple pipeline stages at the same time. The VPDU size is roughly proportional to the buffer size in most pipeline stages, so it is important to keep the VPDU size small. In most hardware decoders, the VPDU size can be set to maximum transform block (TB) size. However, in VVC, ternary tree (TT) and binary tree (BT) partition may lead to the increasing of VPDUs size.
In order to keep the VPDU size as 64x64 luma samples, the following normative partition restrictions (with syntax signalling modification) are applied in VTM, as shown in Fig. 5:
– TT split is not allowed (as indicated by “X” in Fig. 5) for a CU with either width or height, or both width and height equal to 128.
– For a 128xN CU with N ≤ 64 (i.e. width equal to 128 and height smaller than 128) , horizontal BT is not allowed.
For an Nx128 CU with N ≤ 64 (i.e. height equal to 128 and width smaller than 128) , vertical BT is not allowed. In Fig. 5, the luma block size is 128x128. The dashed lines indicate block size 64x64. According to the constraints mentioned above, examples of the partitions not allowed are indicated by “X” as shown in various examples (510-580) in Fig. 5.
Intra Chroma Partitioning and Prediction Restriction
In typical hardware video encoders and decoders, processing throughput drops when a picture has smaller intra blocks because of sample processing data dependency between neighbouring intra blocks. The predictor generation of an intra block requires top and left boundary reconstructed samples from neighbouring blocks. Therefore, intra prediction has to be sequentially processed block by block.
In HEVC, the smallest intra CU is 8x8 luma samples. The luma component of the smallest intra CU can be further split into four 4x4 luma intra prediction units (PUs) , but the chroma components of the smallest intra CU cannot be further split. Therefore, the worst case hardware processing throughput occurs when 4x4 chroma intra blocks or 4x4 luma intra blocks are processed. In VVC, in order to improve worst case throughput, chroma intra CBs smaller than 16 chroma samples (size 2x2, 4x2, and 2x4) and chroma intra CBs with width smaller than 4 chroma samples (size 2xN) are disallowed by constraining the partitioning of chroma intra CBs.
In single coding tree, a smallest chroma intra prediction unit (SCIPU) is defined as a coding tree node whose chroma block size is larger than or equal to 16 chroma samples and has at least one child luma block smaller than 64 luma samples, or a coding tree node whose chroma block size is not 2xN and has at least one child luma block 4xN luma samples. It is required that in each SCIPU, all CBs are inter, or all CBs are non-inter, i.e., either intra or intra block copy (IBC) . In case of a non-inter SCIPU, it is further required that chroma of the non-inter SCIPU shall not be further split and luma of the SCIPU is allowed to be further split. In this way, the small chroma intra CBs with size less than 16 chroma samples or with size 2xN are removed. In addition, chroma scaling is not applied in case of a non-inter SCIPU. Here, no additional syntax is signalled, and whether a SCIPU is non-inter can be derived by the prediction mode of the first luma CB in the SCIPU. The type of a SCIPU is inferred to be non-inter if the current slice is an I-slice or the current SCIPU has a 4x4 luma partition in it after further split one time (because no inter 4x4 is allowed in VVC) ; otherwise, the type of the SCIPU (inter or non-inter) is indicated by one flag before parsing the CUs in the SCIPU.
For the dual tree in intra picture, the 2xN intra chroma blocks are removed by disabling vertical binary and vertical ternary splits for 4xN and 8xN chroma partitions, respectively. The small chroma blocks with sizes 2x2, 4x2, and 2x4 are also removed by partitioning restrictions.
In addition, a restriction on picture size is considered to avoid 2x2/2x4/4x2/2xN intra chroma blocks at the corner of pictures by considering the picture width and height to be multiple of max (8, MinCbSizeY) .
Intra Mode Coding with 67 Intra Prediction Modes
To capture the arbitrary edge directions presented in natural video, the number of directional intra modes in VVC is extended from 33, as used in HEVC, to 65. The new directional modes not in HEVC are depicted as red dotted arrows in Fig. 6, and the planar and DC modes remain the same. These denser directional intra prediction modes apply for all block sizes and for both luma and chroma intra predictions.
In VVC, several conventional angular intra prediction modes are adaptively replaced with wide-angle intra prediction modes for the non-square blocks.
In HEVC, every intra-coded block has a square shape and the length of each of its side is a power of 2. Thus, no division operations are required to generate an intra-predictor using DC mode. In VVC, blocks can have a rectangular shape that necessitates the use of a division operation per block in the general case. To avoid division operations for DC prediction, only the longer side is used to compute the average for non-square blocks.
To keep the complexity of the most probable mode (MPM) list generation low, an intra mode coding method with 6 MPMs is used by considering two available neighbouring intra modes. The following three aspects are considered to construct the MPM list:
– Default intra modes
– Neighbouring intra modes
– Derived intra modes.
A unified 6-MPM list is used for intra blocks irrespective of whether MRL and ISP coding tools are applied or not. The MPM list is constructed based on intra modes of the left and above neighbouring block. Suppose the mode of the left is denoted as Left and the mode of the above block is denoted as Above, the unified MPM list is constructed as follows:
– When a neighbouring block is not available, its intra mode is set to Planar by default.
– If both modes Left and Above are non-angular modes:
– MPM list → {Planar, DC, V, H, V -4, V + 4}
– If one of modes Left and Above is angular mode, and the other is non-angular:
– Set a mode Max as the larger mode in Left and Above
– MPM list → {Planar, Max, Max -1, Max + 1, Max –2, Max + 2}
– If Left and Above are both angular and they are different:
– Set a mode Max as the larger mode in Left and Above
– If Max –Min is equal to 1:
● MPM list → {Planar, Left, Above, Min –1, Max + 1, Min –2}
– Otherwise, if Max –Min is greater than or equal to 62:
● MPM list → {Planar, Left, Above, Min + 1, Max –1, Min + 2}
– Otherwise, if Max –Min is equal to 2:
● MPM list → {Planar, Left, Above, Min + 1, Min –1, Max + 1}
– Otherwise:
● MPM list → {Planar, Left, Above, Min –1, Min + 1, Max –1}
– If Left and Above are both angular and they are the same:
– MPM list → {Planar, Left, Left -1, Left + 1, Left –2, Left + 2}
Besides, the first bin of the MPM index codeword is CABAC context coded. In total three contexts are used, corresponding to whether the current intra block is MRL enabled, ISP enabled, or a normal intra block.
During 6 MPM list generation process, pruning is used to remove duplicated modes so that only unique modes can be included into the MPM list. For entropy coding of the 61 non-MPM modes, a Truncated Binary Code (TBC) is used.
Wide-Angle Intra Prediction for Non-Square Blocks
Conventional angular intra prediction directions are defined from 45 degrees to -135 degrees in clockwise direction. In VVC, several conventional angular intra prediction modes are adaptively replaced with wide-angle intra prediction modes for non-square blocks. The replaced modes are signalled using the original mode indexes, which are remapped to the indexes of wide angular modes after parsing. The total number of intra prediction modes is unchanged, i.e., 67, and the intra mode coding method is unchanged.
To support these prediction directions, the top reference with length 2W+1, and the left reference with length 2H+1, are defined as shown in Fig. 7A and Fig. 7B respectively. The Dia. mode in Fig. 7A and Fig. 7B are diagonal modes, i.e., mode 34.
The number of replaced modes in wide-angular direction mode depends on the aspect ratio of a block. The replaced intra prediction modes are illustrated in Table 2.
Table 2 –Intra prediction modes replaced by wide-angular modes
In VVC, 4: 2: 2 and 4: 4: 4 chroma formats are supported as well as 4: 2: 0. Chroma derived mode (DM) derivation table for 4: 2: 2 chroma format was initially ported from HEVC extending the number of entries from 35 to 67 to align with the extension of intra prediction modes. Since HEVC specification does not support prediction angle below -135° and above 45°, luma intra prediction modes ranging from 2 to 5 are mapped to 2. Therefore, chroma DM derivation table for 4: 2: 2: chroma format is updated by replacing some values of the entries of the mapping table to convert prediction angle more precisely for chroma blocks.
Cross-Component Linear Model (CCLM) Prediction
To reduce the cross-component redundancy, a cross-component linear model (CCLM) prediction mode is used in the VVC, for which the chroma samples are predicted based on the reconstructed luma samples of the same CU by using a linear model as follows:
pred_C (i, j) =α·rec_L′ (i, j) + β (1)
where pred_C (i, j) represents the predicted chroma samples in a CU and rec_L (i, j) represents the downsampled reconstructed luma samples of the same CU.
The CCLM parameters (α and β) are derived with at most four neighbouring chroma samples and their corresponding down-sampled luma samples. Suppose the current chroma block dimensions are W×H, then W’ and H’ are set as
– W’ = W, H’ = H when LM_LA mode is applied;
– W’ =W + H when LM_A mode is applied;
– H’ = H + W when LM_L mode is applied.
The above neighbouring positions are denoted as S [0, -1 ] …S [W’ -1, -1 ] and the left neighbouring positions are denoted as S [-1, 0 ] …S [-1, H’ -1 ] . Then the four samples are selected as
- S [W’ /4, -1 ] , S [3 *W’ /4, -1 ] , S [-1, H’ /4 ] , S [-1, 3 *H’ /4 ] when LM mode is applied and both above and left neighbouring samples are available;
- S [W’ /8, -1 ] , S [3 *W’ /8, -1 ] , S [5 *W’ /8, -1 ] , S [7 *W’ /8, -1 ] when LM-A mode is applied or only the above neighbouring samples are available;
- S [-1, H’ /8 ] , S [-1, 3 *H’ /8 ] , S [-1, 5 *H’ /8 ] , S [-1, 7 *H’ /8 ] when LM-L mode is applied or only the left neighbouring samples are available.
The four neighbouring luma samples at the selected positions are down-sampled and compared four times to find two larger values: x⁰ _A and x¹ _A, and two smaller values: x⁰ _B and x¹ _B. Their corresponding chroma sample values are denoted as y⁰ _A, y¹ _A, y⁰ _B and y¹ _B. Then x_A, x_B, y_A and y_B are derived as:
Xa= (x0A + x1A +1) >>1;
Xb= (x0B + x1B +1) >>1;
Ya= (y0A + y1A +1) >>1;
Yb= (y0B + y1B +1) >>1 (2)
Finally, the linear model parameters α and β are obtained according to the following equations.

β=Y_b-α·X_b (4)
Fig. 8 shows an example of the location of the left and above samples and the sample of the current block involved in the LM_LA mode. Fig. 8 shows the relative sample locations of N × N chroma block 810, the corresponding 2N × 2N luma block 820 and their neighbouring samples (shown as filled circles) .
The division operation to calculate parameter α is implemented with a look-up table. To reduce the memory required for storing the table, the diff value (difference between maximum and minimum values) and the parameter α are expressed by an exponential notation. For example, diff is approximated with a 4-bit significant part and an exponent. Consequently, the table for 1/diff is reduced into 16 elements for 16 values of the significand as follows:
DivTable [] = {0, 7, 6, 5, 5, 4, 4, 3, 3, 2, 2, 1, 1, 1, 1, 0 } (5)
This would have a benefit of both reducing the complexity of the calculation as well as the memory size required for storing the needed tables.
Besides the above template and left template can be used to calculate the linear model coefficients together, they also can be used alternatively in the other 2 LM modes, called LM_A, and LM_L modes.
In LM_A mode, only the above template is used to calculate the linear model coefficients. To get more samples, the above template is extended to (W+H) samples. In LM_L mode, only left template are used to calculate the linear model coefficients. To get more samples, the left template is extended to (H+W) samples.
In LM_LA mode, left and above templates are used to calculate the linear model coefficients.
To match the chroma sample locations for 4: 2: 0 video sequences, two types of down-sampling filter are applied to luma samples to achieve 2 to 1 down-sampling ratio in both horizontal and vertical directions. The selection of down-sampling filter is specified by a SPS level flag. The two down-sampling filters are as follows, which are corresponding to “type-0” and “type-2” content, respectively.
Rec_L′ (i, j) = [rec_L (2i-1, 2j-1) +2·rec_L (2i-1, 2j-1) +rec_L (2i+1, 2j-1) +
rec_L (2i-1, 2j) +2·rec_L (2i, 2j) +rec_L (2i+1, 2j) +4] ＞＞3 (6)
Rec_L′ (i, j) =rec_L (2i, 2j-1) +rec_L (2i-1, 2j) +4·rec_L (2i, 2j) +rec_L (2i+1, 2j) +
rec_L (2i, 2j+1) +4] ＞＞3 (7)
Note that only one luma line (general line buffer in intra prediction) is used to make the down-sampled luma samples when the upper reference line is at the CTU boundary.
This parameter computation is performed as part of the decoding process, and is not just as an encoder search operation. As a result, no syntax is used to convey the α and β values to the decoder.
For chroma intra mode coding, a total of 8 intra modes are allowed for chroma intra mode coding. Those modes include five traditional intra modes and three cross-component linear model modes (LM_LA, LM_A, and LM_L) . Chroma mode signalling and derivation process are shown in Table 3 Chroma mode coding directly depends on the intra prediction mode of the corresponding luma block. Since separate block partitioning structure for luma and chroma components is enabled in I slices, one chroma block may correspond to multiple luma blocks. Therefore, for Chroma DM mode, the intra prediction mode of the corresponding luma block covering the centre position of the current chroma block is directly inherited.
Table -3 –Derivation of chroma prediction mode from luma mode when cclm_is enabled
A single binarization table is used regardless of the value of sps_cclm_enabled_flag as shown in Table 4.
Table 4–Unified binarization table for chroma prediction mode
In Table 4, the first bin indicates whether it is regular (0) or CCLM modes (1) . If it is LM mode, then the next bin indicates whether it is LM_LA (0) or not. If it is not LM_LA, next 1 bin indicates whether it is LM_L (0) or LM_A (1) . For this case, when sps_cclm_enabled_flag is 0, the first bin of the binarization table for the corresponding intra_chroma_pred_mode can be discarded prior to the entropy coding. Or, in other words, the first bin is inferred to be 0 and hence not coded. This single binarization table is used for both sps_cclm_enabled_flag equal to 0 and 1 cases. The first two bins in Table 4 are context coded with its own context model, and the rest bins are bypass coded.
In addition, in order to reduce luma-chroma latency in dual tree, when the 64x64 luma coding tree node is partitioned with Not Split (and ISP is not used for the 64x64 CU) or QT, the chroma CUs in 32x32 /32x16 chroma coding tree node are allowed to use CCLM in the following way:
– If the 32x32 chroma node is not split or partitioned QT split, all chroma CUs in the 32x32 node can use CCLM
– If the 32x32 chroma node is partitioned with Horizontal BT, and the 32x16 child node does not split or uses Vertical BT split, all chroma CUs in the 32x16 chroma node can use CCLM.
In all the other luma and chroma coding tree split conditions, CCLM is not allowed for chroma CU.
Multiple Model CCLM (MMLM)
In the JEM (J. Chen, E. Alshina, G.J. Sullivan, J. -R. Ohm, and J. Boyce, Algorithm Description of Joint Exploration Test Model 7, document JVET-G1001, ITU-T/ISO/IEC Joint Video Exploration Team (JVET) , Jul. 2017) , multiple model CCLM mode (MMLM) is proposed for using two models for predicting the chroma samples from the luma samples for the whole CU. In MMLM, neighbouring luma samples and neighbouring chroma samples of the current block are classified into two groups, each group is used as a training set to derive a linear model (i.e., a particular α and β are derived for a particular group) . Furthermore, the samples of the current luma block are also classified based on the same rule for the classification of neighbouring luma samples.
Fig. 9 shows an example of classifying the neighbouring samples into two groups. Threshold is calculated as the average value of the neighbouring reconstructed luma samples. A neighbouring sample with Rec′_L [x, y] <= Threshold is classified into group 1; while a neighbouring sample with Rec′_L [x, y] > Threshold is classified into group 2.
Local illumination compensation (LIC)
Local Illumination Compensation (LIC) is a method of inter prediction by using neighbouring samples of current block and reference block. It is based on a linear model using a scaling factor a and an offset b. It derives the scaling factor a and the offset b by referring to the neighbouring samples of current block and reference block. Moreover, the coding tool is enabled or disabled adaptively for each CU.
For more detail for LIC, it can refer to the document “JVET-C1001 (J. Chen, et al., “Algorithm Description of Joint Exploration Test Model 3” , Joint Video Exploration Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, 3rd Meeting: Geneva, CH, 26 May –1 June 2016, document JVET-C1001) .
Convolutional cross-component model (CCCM)
In CCCM, a convolutional model is applied to improve the chroma prediction performance. The convolutional model uses a 7-tap filter consisting of a 5-tap plus sign shape spatial component, a nonlinear term and a bias term. The input to the spatial 5-tap components of the filter consist of a centre (C) luma sample which is collocated with the chroma sample to be predicted and its above/north (N) , below/south (S) , left/west (W) and right/east (E) neighbours as shown in Fig. 10.
The nonlinear term (denoted as P) is represented as power of two of the centre luma sample C and scaled to the sample value range of the content:
P = (C*C + midVal ) >> bitDepth (9)
Accordingly, for 10-bit contents, it is calculated as:
P = (C*C + 512 ) >> 10 (10)
The bias term (denoted as B) represents a scalar offset between the input and output (similarly to the offset term in CCLM) and is set to middle chroma value (e.g. 512 for 10-bit contents) .
Output of the filter at a current pixel location (i.e., “C” in Fig. 10) is calculated as a convolution between the filter coefficients c_i and the input values and clipped to the range of valid chroma samples:
predChromaVal = c₀C + c₁N + c₂S + c₃E + c₄W + c₅P + c₆B (11)
The filter coefficients c_i are calculated by minimising MSE between predicted and reconstructed chroma samples in the reference area. Fig. 11 illustrates the reference area which consists of 6 lines of chroma samples above and left of the PU. Reference area extends one PU width to the right and one PU height below the PU boundaries. Area is adjusted to include only available samples. The extensions to the area shown in grey are needed to support the “side samples” of the plus shaped spatial filter and are padded if unavailable.
The MSE minimization is performed by calculating autocorrelation matrix for the luma input and a cross-correlation vector between the luma input and chroma output. Autocorrelation matrix is LDL decomposed and the final filter coefficients are calculated using back-substitution. The process follows roughly the calculation of the ALF filter coefficients in ECM, however LDL decomposition was chosen instead of Cholesky decomposition to avoid using square root operations. As is known in linear algebra, the Cholesky decomposition or Cholesky factorization is a decomposition of a Hermitian, positive-definite matrix into the product of a lower triangular matrix and its conjugate transpose, which is useful for efficient numerical solutions. The Cholesky decomposition is very efficient for solving systems of linear equations. On the other hand, the LDL decomposition is a closely related variant of the classical Cholesky decomposition. The LDL decomposition decomposes a matrix A as A=LDL^T, where L is a lower unit triangular (unitriangular) matrix, and D is a diagonal matrix.
The Convolutional Cross-Component Model (CCCM) has been disclosed for consideration of next generation video coding beyond the VVC and has shown performance improvement. It is desirable to further improve the performance or to reduce the complexity of CCCM are disclosed in the present invention. Accordingly, the present invention disclosed some scheme to further improve the performance of CCCM. In additional, some schemes to reduce the complexity of CCCM are also disclosed.
BRIEF SUMMARY OF THE INVENTION
A method and apparatus of convolutional cross-component prediction model (CCCM) for video coding are disclosed. According to this method, input data associated with a current block comprising a luma block and a chroma block are received, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side. An auto-correlation matrix for reference luma samples in a reference area is derived, wherein the auto-correlation matrix is associated with a footprint of a convolutional filter. A cross-correlation vector between the reference luma samples and reference chroma samples in the reference area is derived, wherein the cross-correlation vector is associated with the footprint of the convolutional filter. Coefficients of the convolutional filter are derived based on the auto-correlation matrix and the cross-correlation vector using Gaussian elimination scheme. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample is generated by applying the convolutional filter with the coefficients derived to a corresponding location of the luma block. A final predictor is generated for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor. The target chroma sample is encoded or decoded using the final predictor.
According to another method, a number of filter taps for a target convolutional filter is determined depending on one or more conditions. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample by applying the target convolutional filter to a corresponding location of the luma block. A final predictor is generated for the target chroma sample comprising the convolutional cross-component model predictor. The target chroma sample is encoded or decoded using the final predictor.
In one embodiment, said one or more conditions correspond to one or more pre-defined implicit rules. In one embodiment, the target convolutional filter is determined implicitly according to a current block size.
In one embodiment, the target convolutional filter is derived by setting a derived coefficient of the reference convolutional filter to zero if the derived coefficient is smaller than a pre-defined threshold value. In another embodiment, the target convolutional filter is derived by setting a derived coefficient of the reference convolutional filter to zero if the derived coefficient is smaller than a sum of coefficients of the reference convolutional filter multiplied by a pre-defined threshold value.
In one embodiment, the target convolutional filter is selected from multiple convolutional filters with different numbers of filter taps.
In one embodiment, the target convolutional filter is selected from multiple convolutional filters generated by separating the reference convolutional filter. In one embodiment, one of said multiple convolutional filters achieving a best performance is explicitly signalled as the target convolutional filter.
In one embodiment, the target convolutional filter is selected from multiple convolutional filters with different shapes according to a block-level syntax.
In one embodiment, the target convolutional filter comprises an optional non-linear tap and a syntax is used to indicate whether the optional non-linear tap is used.
According to another method, one or more syntax elements to indicate a target convolutional cross-component filter selected from multiple convolutional cross-component filters are signalled or parsed, wherein each convolutional cross-component filter uses at least two different luma samples from two different positions. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample is generated by applying the target convolutional cross-component filter to a corresponding location of the luma block. A final predictor is generated for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor. The target chroma sample is encoded or decoded using the final predictor.
In one embodiment, said one or more syntax elements are signalled or parsed at an SPS (Sequence Parameter Set) , PPS (Picture Parameter Set) , PH (Picture Header) , SH (Slice Header) , or CTU (Coding Tree Unit) level.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1A illustrates an exemplary adaptive Inter/Intra video coding system incorporating loop processing.
Fig. 1B illustrates a corresponding decoder for the encoder in Fig. 1A.
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) .
Fig. 3 illustrates an example of the signalling mechanism of the partition splitting information in quadtree with nested multi-type tree coding tree structure.
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.
Fig. 5 shows some examples of TT split forbidden when either width or height of a luma coding block is larger than 64.
Fig. 6 shows the intra prediction modes as adopted by the VVC video coding standard.
Figs. 7A-B illustrate examples of wide-angle intra prediction a block with width larger than height (Fig. 7A) and a block with height larger than width (Fig. 7B) .
Fig. 8 shows an example of the location of the left and above samples and the sample of the current block involved in the LM_LA mode.
Fig. 9 shows an example of classifying the neighbouring samples into two groups according to multiple mode CCLM.
Fig. 10 illustrates an example of spatial part of the convolutional filter for CCCM.
Fig. 11 illustrates an example of reference area (with its paddings) used to derive the CCCM filter coefficients.
Fig. 12 illustrates the 3x2 down-sampling filter used for down-sampling the luma samples for YUV420 colour format.
Figs. 13A-B illustrate two kinds of filter shapes with numbers of filter taps according to one embodiment of the present invention.
Figs. 14A-D illustrate four kinds of filter shapes with numbers of filter taps according to one embodiment of the present invention.
Figs. 15A-D illustrate four kinds of filter shapes with numbers of filter taps according to one embodiment of the present invention.
Fig. 16 illustrates a flowchart of an exemplary video coding system that incorporates a CCCM (Convolutional Cross-Component Model) related mode using Gaussian elimination according to an embodiment of the present invention.
Fig. 17 illustrates a flowchart of an exemplary video coding system that utilises multiple CCCM models with numbers of filter taps according to an embodiment of the present invention.
Fig. 18 illustrates a flowchart of an exemplary video coding system that incorporates signalling for selecting a target filter from multiple CCCM models according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

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.
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.
The following methods are proposed to reduce the complexity of CCCM.
In current CCCM, a method of LDL decomposition followed by forward-substitution and back-substitution is used to derive the optimal filter coefficients. However, most of the process in this method are sequential, which introduces some latency issues or acceleration difficulties. We proposed to use elimination-based method to derive the optimal coefficients of CCCM model, which has lower latency and is easier parallel processing.
In one embodiment, Gaussian elimination method can be used to derive the optimal coefficients of CCCM model. In mathematics, Gaussian elimination, also known as row reduction, is an algorithm for solving systems of linear equations. It consists of a sequence of operations performed on the corresponding matrix of coefficients. To perform row reduction on a matrix, one uses a sequence of elementary row operations to modify the matrix until the lower left-hand corner of the matrix is filled with zeros, as much as possible. The Gaussian elimination method has lower computational complexity than LDL decomposition and Cholesky decomposition.
According to the present invention, the LDL or Cholesky decomposition as used by the conventional CCCM for MSE minimization is replaced by the Gaussian elimination to reduce the computational complexity. Accordingly, the autocorrelation matrix for the luma input and a cross-correlation vector between the luma input and chroma output are calculated first. The autocorrelation matrix for the luma input can be calculated according to the footprint of the reference convolutional filter (e.g. the 7-tap CCCM filter) using luma data in the reference area (e.g. the reference area in Fig. 11) . The cross-correlation vector between the luma input and chroma output can be calculated according to the footprint of the reference convolutional filter (e.g. the 7-tap CCCM filter) using luma data and chroma data in the reference area (e.g. the reference area in Fig. 11) .
CCCM uses a 7-tap filter to formulate a chroma prediction model, which has been shown to achieve noticeable coding gain. However, solving the optimal parameters for a 7-tap filter can be quite complicated, especially on tiny chroma blocks. Overfitting on the reference samples can also be a problem. As a result, several methods are proposed to reduce the number of filter taps from an initial or a reference convolutional filter (e.g. the 7-tap filter used in the conventional CCCM) implicitly or explicitly.
In one embodiment, CCCM model with adaptive filter shape is proposed to reduce the number of filter taps. It will adaptively reduce the number of filter taps by one or more pre-defined implicit rules.
In one embodiment, the implicit selection of CCCM model with adaptive filter shapes is dependent on the CU size. In one embodiment, one or more predefined values are used for the filter taps instead of deriving it at the decoder side.
In one embodiment, if a derived coefficient of CCCM model is smaller than a pre-defined threshold value, the coefficient will be set it to zero to reduce the number of filter taps.
In one embodiment, the sum of all derived coefficients of the reference convolutional filter is calculated first. If one coefficient is smaller than the sum of all coefficients of the reference convolutional filter multiplied by a pre-defined threshold value, we will set this coefficient to zero to reduce the number of filter taps.
In one embodiment, there are multiple CCCM models with different numbers of filter taps. When the model with more filter taps performs much better than the model with less filter taps, the model with more filter taps will be selected. Otherwise, we would rather select the model with less filter taps, even if it is slightly worse than the model with more filter taps.
Another method is proposed to reduce the number of filter taps. Original 7-tap CCCM filter can be separated into multiple filter shapes, and then explicitly signal the best one by RDO process.
In one embodiment, there are two kinds of filter shapes. The first one consists of centre, left, right, non-linear and bias terms (Fig. 13A) ; the second one consists of centre, above, below, non-linear and bias terms (Fig. 13B) .
predChromaVal₁ = c₀C + c₁E + c₂W + c₃P + c₄B, (12)
predChromaVal₂ = c₀C + c₁N + c₂S + c₃P + c₄B. (13)
In one embodiment, there are four kinds of filter shapes. The first one consists of centre, right, non-linear and bias terms (Fig. 14A) ; the second one consists of centre, left, non-linear and bias terms (Fig. 14B) ; the third one consists of centre, above, non-linear and bias terms (Fig. 14C) ; the fourth one consists of centre, below, non-linear and bias terms (Fig. 14D) .
predChromaVal₁ = c₀C + c₁E + c₂P + c₃B,              (14)
predChromaVal₂ = c₀C + c₁W + c₂P + c₃B,           (15)
predChromaVal₃ = c₀C + c₁N + c₂P + c₃B,              (16)
predChromaVal₄ = c₀C + c₁S + c₂P + c₃B.            (17)
In one embodiment, there are four kinds of filter shapes. The first one consists of centre, right, above, non-linear and bias terms (Fig. 15A) ; the second one consists of centre, left, above, non-linear and bias terms (Fig. 15B) ; the third one consists of centre, right, below, non-linear and bias terms (Fig. 15C) ; the fourth one consists of centre, left, below, non-linear and bias terms (Fig. 15D) .
predChromaVal₁ = c₀C + c₁E + c₂N + c₃P + c₄B,                (18)
predChromaVal₂ = c₀C + c₁W + c₂N + c₃P + c₄B,              (19)
predChromaVal₃ = c₀C + c₁W + c₂S + c₃P + c₄B,              (20)
predChromaVal₄ = c₀C + c₁E + c₂S + c₃P + c₄B.               (21)
In one embodiment, the non-linear term is removed to further reduce the model complexity.
In one embodiment, the non-linear term is optional and decided by the encoder. Additional flag is signalled to enable the non-linear term.
In one embodiment, there is another mode consisting of centre, non-linear and bias terms, while the non-linear term in other modes can be removed.
In one embodiment, allowing a CCCM model with multiple filter shapes can depend on a high-level syntax. For example, a flag can be signalled in the SPS (Sequence Parameter Set) , PPS (Picture Parameter Set) , PH (Picture Header) , SH (Slice Header) , or CTU (Coding Tree Unit) level.
In one embodiment, allowing a CCCM model with multiple filter shapes can depend on a block-level syntax. For example, a flag can be signalled for each CU, or be signalled for each chroma block.
In one embodiment, indicating which filter shape is used in CCCM can depend on a syntax, and the syntax can be coded by truncated unary coding or truncated binary coding.
Instead of directly using the sample values, the method of using a differential form to build the CCCM model is proposed. All or part of adjacent spatial terms are replaced by differential terms, which are the differences between respective adjacent spatial terms and the central term. In addition, a clipping operation can be applied to the differential terms to reduce the noise.
In one embodiment, a 6-tap CCCM model is used, which consists of four spatial differential terms, one non-linear term and one bias term. The differential terms are the difference between respective adjacent sample values and the central sample value.
predChromaVal = c0 (N-C) + c1 (S-C) + c2 (E-C) + c3 (W-C) + c4P + c5B (22)
In one embodiment, a 6-tap CCCM model is used, which consists of four spatial differential terms, one non-linear term and one bias term. The differential terms correspond to the clipped differences between respectively adjacent sample values and the central sample value, where the clipping operation is performed after computing the difference. A threshold value is pre-defined for the clipping operation.
predChromaVal = c₀*Clip (N-C) + c₁*Clip (S-C) + c₂*Clip (E-C)
+ c₃*Clip (W-C) + c₄P + c₅B (23)
In one embodiment, clipped differential term is used in CCCM model. The pre-defined clipping threshold value of each differential term can be different or partially different.
In one embodiment, clipped differential term is used in CCCM model. There are several pre-defined clipping threshold values, and a syntax can be signalled for indicating the selected threshold.
In one embodiment, we have two 6-tap CCCM models. The first one use differential terms, and the second one use clipped differential terms. A flag can be signalled for indicating the selected model.
The CCCM (Convolutional Cross-Component Model) with Gaussian elimination or with multiple CCCM models as described above can be implemented in an encoder side and/or a decoder side. For example, any of the proposed CCCM methods can be implemented in an Intra coding module (e.g. Intra pred. 150 in Fig. 1B) and/or an inter prediction module (e.g. MC 152 in Fig. 1B) in a decoder or an Intra coding module (e.g. Intra Pred. 110 in Fig. 1A) and/or an inter prediction module (e.g. Inter Pred. 112 in Fig. 1A) in an encoder. Any of the proposed CCCM methods 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 CCCM processing. While the Intra Pred. units (e.g. unit 110 in Fig. 1A and unit 150 in Fig. 1B) and/or inter prediction module (e.g. Inter Pred. 112 in Fig. 1Aand MC 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) ) .
Fig. 16 illustrates a flowchart of an exemplary video coding system that incorporates a CCCM (Convolutional Cross-Component Model) related mode using Gaussian elimination 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 this method, input data associated with a current block comprising a luma block and a chroma block are received in step 1610, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side. An auto-correlation matrix for reference luma samples in a reference area is derived in step 1620, wherein the auto-correlation matrix is associated with a footprint of a convolutional filter. A cross-correlation vector between the reference luma samples and reference chroma samples in the reference area is derived in step 1630, wherein the cross-correlation vector is associated with the footprint of the convolutional filter. Coefficients of the convolutional filter are derived based on the auto-correlation matrix and the cross-correlation vector using Gaussian elimination scheme in step 1640. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample is generated by applying the convolutional filter with the coefficients derived to a corresponding location of the luma block in step 1650. A final predictor is generated for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor in step 1660. The target chroma sample is encoded or decoded using the final predictor in step 1670.
Fig. 17 illustrates a flowchart of an exemplary video coding system that utilises multiple CCCM models with numbers of filter taps according to an embodiment of the present invention. According to this method, input data associated with a current block comprising a luma block and a chroma block are received in step 1710, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side. A number of filter taps for target convolutional filter is determined depending on one or more conditions in step 1720. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample is generated by applying the target convolutional filter to a corresponding location of the luma block in step 1730. A final predictor is generated for the target chroma sample is generated comprising the convolutional cross-component model predictor in step 1740. The target chroma sample is encoded or decoded using the final predictor in step 1750.
Fig. 18 illustrates a flowchart of an exemplary video coding system that incorporates signalling for selecting a target filter from multiple CCCM models according to an embodiment of the present invention. According to this method, input data associated with a current block comprising a luma block and a chroma block are received in step 1810, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side. One or more syntax elements to indicate a target convolutional cross-component filter selected from multiple convolutional cross-component filters are signalled or parsed in step 1820, wherein each convolutional cross-component filter uses at least two different luma samples from two different positions. For a target chroma sample in the chroma block, a convolutional cross-component model predictor for the target chroma sample is generated by applying the target convolutional filter to a corresponding location of the luma block in step 1830. A final predictor is generated for the target chroma sample is generated comprising the convolutional cross-component model predictor in step 1840. The target chroma sample is encoded or decoded using the final predictor in step 1850.
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.
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.
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.
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

A method of video coding for colour pictures using cross-component prediction, the method comprising:

receiving input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

deriving an auto-correlation matrix for reference luma samples in a reference area, wherein the auto-correlation matrix is associated with a footprint of a convolutional filter;

deriving a cross-correlation vector between the reference luma samples and reference chroma samples in the reference area, wherein the cross-correlation vector is associated with the footprint of the convolutional filter;

deriving coefficients of the convolutional filter based on the auto-correlation matrix and the cross-correlation vector using Gaussian elimination scheme;

for a target chroma sample in the chroma block, generating a convolutional cross-component model predictor for the target chroma sample by applying the convolutional filter with the coefficients derived to a corresponding location of the luma block;

generating a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encoding or decoding the target chroma sample using the final predictor.
An apparatus of video coding for colour pictures using cross-component prediction, the apparatus comprising one or more electronics or processors arranged to:

receive input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

derive an auto-correlation matrix for reference luma samples in a reference area, wherein the auto-correlation matrix is associated with a footprint of a convolutional filter;

deriving a cross-correlation vector between the reference luma samples and reference chroma samples in the reference area, wherein the cross-correlation vector is associated with the footprint of the convolutional filter;

derive coefficients of the convolutional filter based on the auto-correlation matrix and the cross-correlation vector using Gaussian elimination scheme;

for a target chroma sample in the chroma block, generate a convolutional cross-component model predictor for the target chroma sample by applying the convolutional filter with the coefficients derived to a corresponding location of the luma block;

generate a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encode or decode the target chroma sample using the final predictor.
A method of video coding for colour pictures using cross-component prediction, the method comprising:

receiving input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

determining, depending on one or more conditions, a number of filter taps for a target convolutional filter;

for a target chroma sample in the chroma block, generating a convolutional cross-component model predictor for the target chroma sample by applying the target convolutional filter to a corresponding location of the luma block;

generating a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encoding or decoding the target chroma sample using the final predictor.
The method of Claim 3, wherein said one or more conditions correspond to one or more pre-defined implicit rules.
The method of Claim 4, wherein the target convolutional filter is determined implicitly according to a current block size.
The method of Claim 3, wherein the target convolutional filter is derived by setting a derived coefficient of a reference convolutional filter to zero if the derived coefficient is smaller than a pre-defined threshold value.
The method of Claim 3, wherein the target convolutional filter is derived by setting a derived coefficient of a reference convolutional filter to zero if the derived coefficient is smaller than a sum of coefficients of the reference convolutional filter multiplied by a pre-defined threshold value.
The method of Claim 3, wherein the target convolutional filter is selected from multiple convolutional filters with different numbers of filter taps.
The method of Claim 3, wherein the target convolutional filter is selected from multiple convolutional filters generated by separating a reference convolutional filter.
The method of Claim 9, wherein one of said multiple convolutional filters achieving a best performance is explicitly signalled as the target convolutional filter.
The method of Claim 3, wherein the target convolutional filter comprises an optional non-linear tap and a syntax is used to indicate whether the optional non-linear tap is used.
An apparatus of video coding for colour pictures using cross-component prediction, the apparatus comprising one or more electronics or processors arranged to:

receive input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

determine, depending on one or more conditions, a number of filter taps for a target convolutional filter;

for a target chroma sample in the chroma block, generate a convolutional cross-component model predictor for the target chroma sample by applying the target convolutional filter to a corresponding location of the luma block;

generate a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encode or decode the target chroma sample using the final predictor.
A method of video coding for colour pictures using cross-component prediction, the method comprising:

receiving input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

signalling or parsing one or more syntax elements to indicate a target convolutional cross-component filter selected from multiple convolutional cross-component filters, wherein each convolutional cross-component filter uses at least two different luma samples from two different positions;

for a target chroma sample in the chroma block, generating a convolutional cross-component model predictor for the target chroma sample by applying the target convolutional cross-component filter to a corresponding location of the luma block;

generating a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encoding or decoding the target chroma sample using the final predictor.
The method of Claim 13, wherein the target convolutional filter is selected from multiple convolutional filters with different shapes according to a block-level syntax.
The method of Claim 13, wherein said one or more syntax elements are signalled or parsed at an SPS (Sequence Parameter Set) , PPS (Picture Parameter Set) , PH (Picture Header) , SH (Slice Header) , or CTU (Coding Tree Unit) level.
An apparatus of video coding for colour pictures using cross-component prediction, the apparatus comprising one or more electronics or processors arranged to:

receive input data associated with a current block comprising a luma block and a chroma block, wherein the input data comprise pixel data to be encoded at an encoder side or coded data associated with the current block to be decoded at a decoder side;

signal or parse one or more syntax elements to indicate a target convolutional cross-component filter selected from multiple convolutional cross-component filters, wherein each convolutional cross-component filter uses at least two different luma samples from two different positions;

for a target chroma sample in the chroma block, generate a convolutional cross-component model predictor for the target chroma sample by applying the target convolutional cross-component filter to a corresponding location of the luma block;

generate a final predictor for the target chroma sample from a set of prediction candidates comprising the convolutional cross-component model predictor; and

encode or decode the target chroma sample using the final predictor.