Encoding / decoding video picture data
Patent Information
- Application Number
- PCT/CN2024/077148
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2024-01-08
- Filing Date
- 2024-02-09
- Publication Date
- 2026-08-27
Smart Images

Figure CN2024077148_27082026_PF_FP_ABST
Abstract
Description
ENCODING / DECODING VIDEO PICTURE DATAFIELD
[0001] The present application generally relates to video picture encoding and decoding. Particularly, but not exclusively, the technical field of the present application is related to template-based intra prediction of block of video picture.BACKGROUND
[0002] The present section is intended to introduce the reader to various aspects of art, which may be related to various aspects of at least one exemplary embodiment of the present application that is described and / or claimed below. This discussion is believed to be helpful in providing the reader with background information to facilitate a better understanding of the various aspects of the present application. Accordingly, it should be understood that these statements are to be read in this light, and not as admissions of prior art.
[0003] ECM (Algorithm description of Enhanced Compression Model 10 (ECM 10) , JVET-AE2025, Geneva, 11-19 July 2023) defines template-based intra prediction mode such as the so-called Decoder-side Intra Mode Derivation (DIMD) mode and the so-called Template-based Intra Mode Derivation (TIMD) modes for determining predicted samples of a current block for both luma and chroma samples.
[0004] Template-based intra prediction mode information is derived for a current block based on at least one reconstructed sample of a template defined around the current block. Deriving template-based intra prediction mode information at the time of decoding a current block introduces latency for decoding / reconstructing the current block.
[0005] One of the problem is to reduce latency for decoding a current block when the decoding requires template-based intra prediction mode information.SUMMARY
[0006] The following section presents a simplified summary of the at least one exemplary embodiment in order to provide a basic understanding of some aspects of the present application. This summary is not an extensive overview of an exemplary embodiment. It is not intended to identify key or critical elements of an exemplary embodiment. The following summary merely presents some aspects of the at least one exemplary embodiment in a simplified form as a prelude to the more detailed description provided elsewhere in the document.
[0007] According to a first aspect of the present application, there is provided a method of determining template-based intra prediction of a block of a video picture based on template-based intra prediction mode information derived based on at least one reconstructed sample of a template, the template comprising at least one reconstructed sample of the current block neighboring the template-based intra predicted block, wherein template-based intra prediction mode information is at least partially derived at the time of reconstructing the at least one sample of the current block.
[0008] In some embodiments, template-based intra prediction mode information is stored at block level.
[0009] In some embodiments, template-based intra prediction mode information is derived based on reconstructed samples of at least one sub-block of the template and template-based intra prediction mode information is stored at sub-block level, the sub-block size being lower than the neighboring block size.
[0010] In some embodiments, the size of the sub-blocks is fixed or is determined based on the template size or the current block size.
[0011] In some embodiments, at least one reconstructed sample of the template is used for deriving more than one template-based intra prediction mode information.
[0012] In some embodiments, template-based intra prediction mode information is derived based on all reconstructed samples of the template.
[0013] In some embodiments, deriving template-based intra prediction information excludes samples outside of a border of the current block or is based on reconstructed samples outside of a border of the current block or is based on padded samples outside of the current block borders based already reconstructing samples within the current block.
[0014] In some embodiments, template-based intra prediction mode information is derived only when all the samples of the template are reconstructed.
[0015] In some embodiments, template-based intra prediction mode information is derived once at least one sample within a reference block is reconstructed, the reference block referring to a block that the template for template-based intra prediction mode is located.
[0016] In some embodiments, template-based intra prediction mode information is derived at the time of reconstructing each block of the video picture.
[0017] In some embodiments, a flag or an index is signaled in a bitstream to indicate that template-based intra prediction mode information is derived based on reconstructed samples of a block at the time of reconstructing said block.
[0018] In some embodiments, the video picture being split into multiple adjacent coding tree units, at the time of reconstructing samples of a block located at a boundary between a current coding tree unit and a next coding tree unit, template-based intra prediction mode information is enforced to be derived at the time of reconstructing a block located at the boundary of the next coding tree unit.
[0019] In some embodiments, the video picture being split into multiple adjacent coding tree units, template-based intra prediction mode information is not derived at the time of reconstructing a block located at a border of a coding tree unit.
[0020] In some embodiments, the flag or index is signaled in sequence parameter set, picture parameter set or in picture header, or at slice, tile, coding tree unit, block, prediction unit, or transformation unit level.
[0021] In some embodiments, deriving template-based prediction mode information at the time of reconstructing blocks of a video picture is not enabled for blocks having sizes lower than a threshold.
[0022] According to a second aspect of the present application, there is provided an apparatus comprising means for performing one of the methods according to the first aspect of the present disclosure.
[0023] According to a third aspect of the present application, there is provided a computer program product including instructions which, when the program is executed by one or more processors, causes the one or more processors to carry out a method according to the first aspect of the present application.
[0024] According to a fourth aspect of the present application, there is provided a non-transitory storage medium carrying instructions of program code for executing a method according to the first aspect of the present application.
[0025] The specific nature of at least one of the exemplary embodiments as well as other objects, advantages, features and uses of said at least one of exemplary embodiments will become evident from the following description of examples taken in conjunction with the accompanying drawings.BRIEF DESCRIPTION OF THE DRAWINGS
[0026] Reference will now be made, by way of example, to the accompanying drawings which show exemplary embodiments of the present application, and in which:
[0027] Figure 1 shows an example of coding-tree unit in accordance with HEVC;
[0028] Figure 2 shows an example of partitioning coding units into prediction units in accordance with HEVC;
[0029] Figure 3 shows an example of a CTU division in accordance with VVC;
[0030] Figure 4 shows examples of split modes supported in the multi-type tree partitioning in accordance with VVC;
[0031] Figure 5 shows a schematic block diagram of a method 100 of encoding a video picture VP in accordance with prior art;
[0032] Figure 6 shows a schematic block diagram of a method 200 of decoding a video picture VP in accordance with prior art;
[0033] Figure 7 shows an example of template for computing a Histogram of Gradients in accordance with prior art;
[0034] Figure 8 shows an example of calculating an intra-predictor of a current block of the video picture based on a weighted linear combination of three intra-predictors in accordance with prior art;
[0035] Figure 9 shows an example of co-located luma block corresponding to a chroma block in accordance with prior art;
[0036] Figure 10 shows an example of definition of non-adjacent spatial neighboring blocks in accordance with prior art;
[0037] Figure 11 shows an example of the first template of a block used for template prediction, and an example of the second template used for calculating a prediction cost;
[0038] Figure 12 illustrates examples of reference lines;
[0039] Figure 13 shows an example of template of a current block used for deriving template-based intra prediction modes for predicting the current block;
[0040] Figure 14 shows a block diagram of a method of reconstructing a current block based on template-based intra prediction information mode in accordance with prior art;
[0041] Figure 15 illustrates an example of template in accordance with at least one embodiment of the present disclosure;
[0042] Figure 16 illustrates an example of HoGs for deriving DIMD Information in accordance with at least one embodiment of the present disclosure;
[0043] Figure 17 illustrates an example of DIMD-based predicted block of a template of a current block in accordance with at least one exemplary embodiment of the present disclosure;
[0044] Figure 18 illustrates an example of deriving multiple template-based intra prediction information in accordance with at least one exemplary embodiment of the present disclosure;
[0045] Figure 19 shows an example template-based intra prediction mode information derivation at sub-block level in accordance with at least one exemplary embodiment of the present disclosure;
[0046] Figure 20 shows an example of a 3x3 Sobel filter that is applied on reconstructed samples of a block of template to derive a template-based intra prediction mode information in accordance with at least one exemplary embodiment of the present disclosure.
[0047] Figure 21 illustrates an example of deriving template-based intra prediction mode information once at least one required sample of the template is reconstructed in accordance with at least one exemplary embodiment of the present disclosure;
[0048] Figure 22 illustrates an example of decoding a current block located at a boundary of a current CTU in accordance with at least one exemplary embodiment of the present disclosure;
[0049] Figure 23 illustrates an example signalling template-based intra prediction mode derivation information at sequence parameter set (SPS) in accordance with at least one exemplary embodiments of the present disclosure;
[0050] Figure 24 illustrates an example signalling template-based intra prediction mode derivation information at sequence parameter set (SPS) in accordance with at least one exemplary embodiments of the present disclosure;
[0051] Figure 25 shows a block diagram of an example of a system in which various aspects and exemplary embodiments are implemented.
[0052] Similar or same elements are referenced with the same reference numbers.
[0053] DESCRIPTION OF EXEMPLARY EMBODIMENTS
[0054] At least one of the exemplary embodiments is described more fully hereinafter with reference to the accompanying figures, in which examples of at least one of the exemplary embodiments are depicted. An exemplary embodiment may, however, be embodied in many alternate forms and should not be construed as limited to the examples set forth herein. Accordingly, it should be understood that there is no intent to limit exemplary embodiments to the particular forms disclosed. On the contrary, the present application is intended to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the present application.
[0055] At least one of the aspects generally relates to video picture encoding and decoding, one other aspect generally relates to transmitting a bitstream provided or encoded and one other aspect relates to receiving / accessing a decoded bitstream.
[0056] At least one of the exemplary embodiments is described for encoding / decoding a video picture but extends to the encoding / decoding of video pictures (sequences of pictures) because each video picture is sequentially encoded / decoded as described below.
[0057] Moreover, the at least one exemplary embodiments are not limited to MPEG standards such as AVC (ISO / IEC 14496-10 Advanced Video Coding for generic audio-visual services, ITU-T Recommendation H. 264, https: / / www. itu. int / rec / T-REC-H. 264-202108-P / en) , EVC (ISO / IEC 23094-1 Essential video coding) , HEVC (ISO / IEC 23008-2 High Efficiency Video Coding, ITU-T Recommendation H. 265, https: / / www. itu. int / rec / T-REC-H. 265-202108-P / en, VVC (ISO / IEC 23090-3 Versatile Video Coding, ITU-T Recommendation H. 266, https: / / www. itu. int / rec / T-REC-H. 266-202008-I / en but may be applied to other standards and recommendations such as AV1 (AOMedia Video 1, http: / / aomedia. org / av1 / specification / ) for example. The at least one exemplary embodiment may apply to pre-existing or future-developed, and extensions of any such standards and recommendations. Unless indicated otherwise, or technically precluded, the aspects described in the present application may be used individually or in combination.
[0058] A pixel corresponds to the smallest display unit on a screen, which can be composed of one or more sources of light (1 for monochrome screen or 3 or more for colour screens) .
[0059] A video picture, also denoted frame or picture frame, comprises at least one component (also called picture component, or channel) determined by a specific picture / video format which specifies all information relative to pixel values and all information which may be used by a display unit and / or any other device to display and / or to decode video picture data related to said video picture.
[0060] A video picture comprises at least one component usually expressed in the shape of an array of samples.
[0061] A monochrome video picture comprises a single component and a color video picture may comprise three components.
[0062] For example, a color video picture may comprise a luma (or luminance) component and two chroma components when the picture / video format is the well-known (Y, Cb, Cr) format or may comprise three color components (one for Red, one for Green and one for Blue) when the picture / video format is the well-known (R, G, B) format.
[0063] Each component of a video picture may comprise a number of samples relative to a number of pixels of a screen on which the video picture is intended to be display. In variants, the number of samples comprised in a component may be a multiple (or fraction) of a number of samples comprised in another component of a same video picture.
[0064] For example, in the case of a video format comprising a luma component and two chroma component like the (Y, Cb, Cr) format, dependent on the color format considered, the chroma component may contain half the number of samples in width and / or height, relative to the luma component.
[0065] A sample is the smallest visual information unit of a component composing a video picture. A sample value may be, for example a luma or chroma value or a colour value of a (R, G, B) format.
[0066] A pixel value is the value of a pixel of a screen. A pixel value may be represented by one sample for monochrome video picture and by multiple co-located samples for color video picture. Co-located samples associated with a pixel mean samples corresponding to the location of a pixel in the screen.
[0067] It is common to consider a video picture as being a set of pixel values, each pixel being represented by at least one sample.
[0068] A block of a video picture is a set of samples of one component of the video picture. A block of at least one luma sample or a block of at least one chroma sample may be considered when the picture / video format is the well-known (Y, Cb, Cr) format, or a block of at least one color sample when the picture / video format is the well-known (R, G, B) format.
[0069] The at least one exemplary embodiment is not limited to a particular picture / video format.
[0070] In the state-of-the-art video compression systems such as HEVC or VVC low-level and high-level picture partitioning are provided to divide a video picture into picture areas so-called Coding-Tree Units (CTU) which size may be typically between 16x16 and 64x64 pixels for HEVC and 32x32, 64x64, or 128x128 pixels for VVC.
[0071] The CTU division of a video picture forms a grid of fixed size CTUs, namely a CTU grid, which upper and left bounds spatially coincide with the top and left borders of the video picture. The CTU grid represents a spatial partition of the video picture.
[0072] In VVC and HEVC, the CTU size (CTU width and CTU height) of all the CTUs of a CTU grid equals a same default CTU size (default CTU width CTU DW and default CTU height CTU DH) . For example, the default CTU size (default CTU height, default CTU width) may equal to 128 (CTU DW=CTU DH=128) . A default CTU size (height, width) is encoded into the bitstream, for example at a sequence level in the Sequence Parameter Set (SPS) .
[0073] The spatial position of a CTU in a CTU grid is determined from a CTU address ctuAddr defining a spatial position of the top-left corner of a CTU from an origin. As illustrated on Figure 1, the CTU address may define the spatial position from the top-left corner of a higher-level spatial structure S containing the CTU.
[0074] A coding tree is associated with each CTU to determine a tree-division of the CTU.
[0075] As illustrated on Figure 1, in HEVC, the coding tree is a quad-tree division of a CTU, where each node is called a Coding Unit (CU) . The spatial position of a CU in the video picture is defined by a CU index cuIdx indicating a spatial position from the top-left corner of the CTU. A CU is spatially partitioned into one or more Prediction Units (PU) . The spatial position of a PU in the video picture VP is defined by a PU index puIdx defining a spatial position from the top-left corner of the CTU and the spatial position of an element of a partitioned PU is defined by a PU partition index puPartIdx defining a spatial position from the top-left corner of a PU. Each PU is assigned some intra or inter prediction data.
[0076] The coding mode intra or inter is assigned on the CU level. That means that a same intra / inter coding mode is assigned to each PU of a CU, though the prediction parameters varying from PU to PU.
[0077] A CU may be also spatially partitioned into one or more Transform Units (TU) , according to a quad-tree called the transform tree. Transform Units are the leaves of the transform tree. The spatial position of a TU in the video picture is defined by a TU index tuIdx defining a spatial position from the top-left corner of a CU. Each TU is assigned some transform parameters. The transform type is assigned on the TU level, and 2D separate transform is performed at TU level during the coding or decoding of a picture block.
[0078] The PU Partition types existing in HEVC are illustrated on Figure 2. They include square partitions (2Nx2N and NxN) , which are the only ones used in both Intra and Inter predicted CUs, symmetric non-square partitions (2NxN, Nx2N, used only in Inter predicted CUs) , and asymmetric Partitions (used only in Inter predicted CUs) . For instance, the PU type 2NxnU stands for an asymmetric horizontal partitioning of the PU, where the smaller partition lies on the top of the PU. According to another example, PU type 2NxnL stands for an asymmetric horizontal partitioning of the PU, where the smaller partition lies on the top of the PU.
[0079] As illustrated on Figure 3, in VVC, the coding tree starts from a root node, i.e., the CTU. Next, a quad-tree (or quaternary tree) split divides the root node into 4 nodes corresponding to 4 sub-blocks of equal sizes (solid lines) . Next, the quaternary tree (or quad-tree) leaves can then be further partitioned by a so-called multi-type tree, which involves a binary or ternary split according to one of 4 split modes illustrated on Figure 4. These split types are the vertical and horizontal binary split modes, noted SBTV and SBTH and the vertical and horizontal ternary split modes SPTTV and STTH.
[0080] The leaves of the coding tree of a CTU are CU in the case of a joint coding tree shared by luma and chroma components.
[0081] Contrary to HEVC, in VVC, in most cases, CU, PU and TU have equal size, which means coding units are generally not partitioned into PU or TU, except in some specific coding modes.
[0082] Figures 5 and 6 provide an overview of video encoding / decoding methods used in current video standard compression systems like HEVC or VVC for example.
[0083] Figure 5 shows a schematic block diagram of steps of a method 100 of encoding a video picture VP in accordance with prior art.
[0084] In step 110, a video picture VP is partitioned into blocks of samples and partitioning information data is signaled into a bitstream. Each block comprises samples of one component of the video picture VP. The blocks thus comprise samples of each component defining the video picture VP.
[0085] For example, in HEVC, a picture is divided into Coding Tree Units (CTU) . Each CTU may be further subdivided using a quad-tree division, where each leaf of the quad-tree is denoted a Coding Unit (CU) . The partitioning information data may then comprise data describing the CTU and the quad-tree subdivision of each CTU.
[0086] Each block of samples, in short block, may then be either a CU (if the CU comprises a single PU) or a PU of a CU.
[0087] Each block is encoded along an encoding loop also called “in loop” using either an intra or inter prediction mode.
[0088] Intra prediction (step 120) used intra prediction data. Intra prediction consists in predicting a current block by means of an intra-predicted block based on already encoded, decoded and reconstructed samples located around the current block, typically on the top and on the left of the current block. Intra prediction is performed in the spatial domain.
[0089] In inter-prediction mode, motion estimation (step 130) and motion compensation (135) are performed. Motion estimation searches, in one or more reference video picture (s) used to predictively encode the current video picture, a candidate reference block that is a good predictor of the current block. For instance, a good predictor of the current block is a predictor which is similar to the current block. The output of the motion estimation step 130 is inter-prediction data comprising motion information (typically one or more motion vectors and one or more reference video picture indices) associated to the current block and other information used for obtaining a same predicted block at the encoding / decoding side. Next, motion compensation (step 135) obtains a predicted block by means of the motion vector (s) and reference video picture index (indices) determined by the motion estimation step 130. Basically, the block belonging to a selected reference video picture and pointed to by a motion vector may be used as the predicted block of the current block. Furthermore, since motion vectors are expressed in fractions of integer pixel positions (which is known as sub-pel accuracy motion vector representation) , motion compensation generally involves a spatial interpolation of some reconstructed samples of the reference video picture to compute the predicted block.
[0090] Prediction information data is signaled into the bitstream. The prediction information may comprise prediction mode (intra or inter or skip) , intra / inter prediction data and any other information used for obtaining a same predicted CU at the decoding side.
[0091] The method 100 selects one prediction mode (the intra or inter prediction mode) by optimizing a rate-distortion trade-off taking into account the encoding of a prediction residual block calculated, for example, by subtracting a candidate predicted block from the current block, and the signaling of prediction information data required for determining said candidate predicted block at the decoding side.
[0092] Usually, the intra and inter prediction modes are set in a list of Most Probable candidate Modes (MPM list) and the best prediction mode is given as being the prediction mode p*for a current block given by: p*=Argminp∈P {RDcost (p) } (1)
[0093] where P is the MPM list for the current block, p represents a candidate coding mode in that MPM list, RDcost (p) is a rate-distortion cost of candidate prediction mode p, typically expressed as: RDcost (p) =D (p) +λ·R (p)
[0094] D (p) is the distortion between the current block and a reconstructed block obtained after encoding / decoding the current block based on the candidate prediction mode p, R (p) is a rate cost associated with the coding of the current block based on the prediction mode p, and λ is the Lagrange parameter representing the rate constraint for coding the current block and typically computed from a quantization parameter Qp used for encoding the current block.
[0095] The current block is usually encoded from a prediction residual block PR. More precisely, a prediction residual block PR is calculated, for example, by subtracting the best predicted block from the current block. The prediction residual block PR is then transformed (step 140) by using, for example, a DCT (discrete cosine transform) or DST (Discrete Sinus transform) type transform, or any other appropriate transform, and the obtained transformed coefficient block is quantized (step 150) .
[0096] In variant, the method 100 may also skip the transform step 140 and apply quantization (step 150) directly to the prediction residual block PR, according to the so-called transform-skip coding mode.
[0097] Quantized transform coefficient block (or quantized prediction residual block) is entropy encoded into the bitstream (step 160) .
[0098] Next, the quantized transform coefficient block (or the quantized residual block) is de-quantized (step 170) and inverse transformed (180) (or not) as part of the encoding loop, leading to a decoded prediction residual block. The decoded prediction residual block and the predicted block are then combined, typically summed, which provides the reconstructed block.
[0099] Other information data may also be entropy encoded in step 160 for encoding a current block of the video picture VP.
[0100] In-loop filters (step 190) may be applied to a reconstructed picture (comprising reconstructed blocks) to reduce compression artefacts. Loop filters may apply after all picture blocks are reconstructed. For instance, they consist in deblocking filter, Sample Adaptive Offset (SAO) or adaptive loop filter.
[0101] The reconstructed blocks or the filtered reconstructed blocks form a reference picture that may be stored into a decoded picture buffer (DPB) so that it can be used as a reference picture for the encoding of a next current block of the video picture VP, or of a next vide picture to encode.
[0102] Figure 6 shows a schematic block diagram of steps of a method 200 of decoding a video picture VP in accordance with prior art.
[0103] In step 210, partitioning information data, prediction information data and quantized transform coefficient block (or quantized residual block) are obtained by entropy decoding a bitstream of encoded video picture data. For instance, this bitstream has been generated in accordance with the method 100.
[0104] Other information data may also be entropy decoded for decoding from the bitstream a current block of the video picture VP.
[0105] In step 220, a reconstructed picture is divided into current blocks based on the partitioning information. Each current block is entropy decoded from the bitstream along a decoding loop also called “in loop” . Each decoded current block is either a quantized transform coefficient block or quantized prediction residual block.
[0106] In step 230, the current block is de-quantized and possibly inverse transformed (step 240) , to obtain a decoded prediction residual block.
[0107] On the other hand, the prediction information data is used to predict the current block. A predicted block is obtained through its intra prediction (step 250) or its motion-compensated temporal prediction (step 260) . The prediction process performed at the decoding side is identical to that of the encoding side.
[0108] Next, the decoded prediction residual block and the predicted block are then combined, typically summed, which provides a reconstructed block.
[0109] In step 270, in-loop filters may apply to a reconstructed picture (comprising reconstructed blocks) and the reconstructed blocks, or the filtered reconstructed blocks form a reference picture that may be stored into a decoded picture buffer (DPB) as above discussed (Figure 5) .
[0110] In step 130 / 135 of Figure 5 or step 260 of Figure 6, an inter-predicted block is defined from inter-prediction data associated with the current block (CU or a PU of a CU) of a video picture. This inter-prediction data comprises motion information that may be represented (coded) according to the so-called whole-block-based AMVP mode (Adaptive Motion Vector Prediction) or the so-called whole-block-based merge mode.
[0111] ECM defines a DIMD mode for determining predicted samples of a current block for both luma and chroma samples. DIMD mode is not a linear-model based mode, in short non-LM mode, i.e. an intra prediction mode that does not refer to a linear model, and DIMD mode does not require explicit parameters signaling such as, for instance planar prediction mode, or Direct mode (DM) .
[0112] For luma sample prediction, the use of DIMD luma mode is signalled in a bitstream by a single flag and the intra predictor is not explicitly signalled in the bitstream but derived based on a gradient analysis of neighbouring reconstructed luma samples i.e. DIMD luma mode is implicit. Thus, intra predictor is derived based on a DIMD luma mode during reconstruction process identically at the encoder and decoder sides.
[0113] In DIMD luma mode, as illustrated on Figure 7, a L-shaped template around current luma block is defined. For example, the L-shaped template comprises a 3-samples wide (in width or height) template (composed of left, above and above-left reconstructed luma samples of reconstructed area R, e.g. a L-shaped template) is defined. Edge detection filters (3x3 horizontal and vertical Sobel filters) , adjacent to a current block B (current CU) , are applied on the samples of the L-shaped template in order to determine the amplitude and angle of luminance directions (orientation) for each middle line sample of the (available) template. A Histogram of Gradients (HoG) is computed where each entry corresponds to conventional intra angular modes and cumulated intensities (amplitudes) are stored: angle=arctan (Ghor / Gver) amplitude = |Ghor|+|Ghor|
[0114] with Ghor and Gver are the intensity of pure horizontal and vertical directions as calculated by Sobel filters. Conventional intra angular modes refer to the existing intra angular modes in for example HEVC and VVC specifications, where each angular mode specifically aims in capturing the direction of the texture in a certain direction or angle. Note the term non-angular modes refer to intra modes that do not capture the direction or angle of the texture. For example, DC (Direct coding) and Planar intra modes can be considered as non-angular modes.
[0115] In DIMD luma mode, a luma blending (mixing, fusing) is applied as a weighted average of multiple luma predictors of the current block: two HoG-based predictors and a planar predictor (M. Abdoli et al, “Non-CE3: Decoder-side Intra Mode Derivation with Prediction Fusion Using Planar” , JVET-O0449, Gothenburg, July 2019) .
[0116] A planar predictor is derived from a planar mode in which the predicted sample values are obtained as a weighted average of 4 reference samples of the L-shaped template. Here, the reference samples are located in the same row or column as the current sample of a current block and the reference samples on the bottom-left and on the top-right position with respect to the current block are used. In VVC, a reference sample smoothing filter is also applied for the Planar mode in the luma component depending on the block size.
[0117] As illustrated on Figure 8, the prediction block of the current block are derived from a weighted linear combination of six predictors for example: Five HoG-based predictors derived by using the five most represented angular intra modes Mi (i=1 to 5) and one predictor derived by using a planar mode. The number of predictors may be lower or greater than six. Figure 8 illustrates a variant wherein only the prediction block is derived as a weighted linear combination of three HoG-based predictors but the principle may be easily extended to six predictors: two HoG-based predictors M1 and M2 (having the largest histogram amplitude values) selected from the HoG and a third predictor derived by using a planar mode. The weighted linear combination of the three predictors uses weights w1, w2, w3 derived, for example as follows: the weight of planar is fixed to 21 / 64 (~1 / 3) . The remaining weight of 43 / 64 (~2 / 3) is then shared between the two other weights, proportionally to the amplitude of their HoG bars. As another example, the three weights may be derived as follows for a block of size W×H: a weight wDimdi is derived for each of the three predictors as discussed below in relation with DIMD merge mode. Next, the weight wi for each of the three derived modes is derived based on the weight wDimdi as follows: If the one the above or left HoG magnitudes is twice larger than the other one. In this case, the weights wi (x, y) are location dependent and computed as follows:
[0118] If the above HoG is twice the left, then:
[0119] If the left HoG is twice the above, then:
[0120] where Δi is pre-defined and may be set to 10.
[0121] For chroma sample prediction, the use of DIMD chroma mode is signalled in a bitstream by a single flag and the intra predictor is not explicitly signalled in the bitstream but derived by using a gradient analysis of neighbouring reconstructed luma and chroma samples i.e. DIMD chroma mode is implicit. Thus, intra predictor is derived from a DIMD chroma mode during reconstruction process identically at the encoder and decoder sides.
[0122] In DIMD chroma mode, an intra predictor of a chroma block may be derived based on a L-shaped template or a portion of a L-shaped template. For example, the L-shaped template may comprise co-located reconstructed luma samples of a luma block associated with the chroma block to be predicted. Specifically, a horizontal gradient and a vertical gradient are calculated for each co-located reconstructed luma sample of the template (grey circles on Figure 9, extracted from JVET-Y0092 (Hyeongmun Jang et al, “CE5-2.5 : De-block filtering with 4x4 chroma sample grid” , JVET-Y0092, Gothenburg, July 2019) to build a HoG. Then, the intra prediction mode with the largest histogram amplitude value is selected from the HoG.
[0123] In variant, in DIMD chroma mode, an intra predictor of a chroma block may be derived based on a template comprising neighboring reconstructed chroma samples of a L-shaped template around the chroma block.
[0124] In a variant, an intra predictor of a chroma block may be derived based on a template comprising co-located reconstructed luma samples of a luma block associated with the chroma block to be predicted and on neighboring reconstructed chroma samples.
[0125] Usually, when the intra prediction mode derived from the DIMD chroma mode is the same as the intra prediction mode derived from the DM (direct mode) mode, the intra prediction mode with the second largest histogram amplitude value is selected as the DIMD chroma mode.
[0126] A DIMD merge mode is also proposed in JVET-AF0120 (Saverio Blasi et al., “EE2-2.1 DIMD merge” , JVET-AF0120, Hannover Oct. 2023) . When using DIMD merge mode, a prediction block of the current block is derived from DIMD information that is extracted from neighbouring blocks of the current block. Firstly, a Merged Histogram of Gradients (MHoG) is computed for the current block based on the Histogram of Gradients HoG of at least one DIMD neighbouring block i.e. a neighboring block encoded / decoded based on a prediction block derived based on the DIMD or DIMD merge mode. When a single DIMD neighbouring block is available, then its histogram of gradients (HoG) is used to form the MHoG for the current block. If more than one DIMD neighbouring blocks are available, the corresponding HoGs are combined by means of amplitude averaging to derive the MHoG. Up to maximum 13 DIMD neighboring blocks are considered to extract DIMD information. Next, the MHoG is used to define DIMD merge modes and weights, as discussed above in relation with DIMD mode. The conventional intra angular modes and their weights corresponding to the five highest amplitudes in the MHoG are selected, and the corresponding predictors are blended as in DIMD merge mode.
[0127] The DIMD information used to derive a predictor of a current block depends on the DIMD neighboring blocks. In JVET-AF0106 (Junyan Huo et al., “EE2- related: Non-adjacent spatial candidates for DIMD merge” , Hannover, oct. 2023) , DIMD neighboring blocks may be non-adjacent spatial neighboring blocks, i.e neighboring block of a current block that does not share a corner or a boundary with the current block.
[0128] For example, Figure 10 shows an example of definition of non-adjacent spatial neighboring blocks. The distances between non-adjacent spatial neighboring blocks of the current block and the current block may be defined based on the width and height of the current block.
[0129] A DIMD mode or DIMD merge mode (for luma or chroma) is associated with associated information such as weights as discussed above. Information associated with DIMD mode or DIMD merge mode may be stored in memory for a block of the video picture VP that is encoded / decoded based on a prediction block derived based on the DIMD mode or DIMD merge mode.
[0130] ECM further defines Template-based Intra Mode Derivation (TIMD) modes. A TIMD mode is used for predicting a block and leads to the derivation of an intra predictor of the block based on previously reconstructed samples of L-shaped template or a portion of a L-shaped template defined around the block. A TIMD mode associated with a block is also associated with a prediction cost evaluated based on samples of the L-shaped template or a portion of the L-shaped template.
[0131] For example, a TIMD mode may be derived for each intra prediction candidate mode of the MPM list, as well as for each of the wide-angle modes if the above-right and / or bottom-left samples of a first template T1 are available. Wide angle modes are described in JVET-Q2002 (Algorithm description for Versatile Video Coding and Test Model 8 (VTM 8) , JVET-Q2002, Brussels, 7-17 January 2020. In brief, 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 signaled using the original mode indexes, which are remapped to the indexes of wide angular modes after parsing. The number of replaced modes in wide-angular direction mode depends on the aspect ratio of a block. Two vertically adjacent predicted samples may use two non-adjacent reference samples in the case of wide-angle intra prediction. Hence, low-pass reference samples filter and side smoothing are applied to the wide-angle prediction to reduce the negative effect of the increased gap Δpα. If a wide-angle mode represents a non-fractional offset. There are 8 modes in the wide-angle modes satisfy this condition, which are [-14, -12, -10, -6, 72, 76, 78, 80] . When a block is predicted by these modes, the samples in the reference buffer are directly copied without applying any interpolation. With this modification, the number of samples needed to be smoothing is reduced. Besides, it aligns the design of non-fractional modes in the conventional prediction modes and wide-angle modes.
[0132] The prediction cost associated with a TIMD mode may be evaluated based on a Sum of Absolute Transformed Difference (SATD) or a Sum of Absolute Difference (SAD) or a Sum of Squared Errors (SSE) between samples of a second template T2 of the block and samples of a reconstructed template of the block. Samples of the reconstructed template of the block are samples located in the second template T2 of the block that are encoded and decoded based on the intra prediction candidate mode or on a wide-angle mode if exists. A list of TIMD candidate modes associated with prediction costs may then be built. Next, two intra prediction modes TIMD1 and TIMD2 with the minimum prediction cost are selected from the list of TIMD candidate modes. Next, a condition between the prediction costs of the two selected TIMD modes is checked to determine if a blending of these two selected TIMD modes is applied or not.
[0133] For example, the condition is fulfilled when the prediction cost of the second selected TIMD mode TIMD2 is lower than twice the prediction cost of the first selected TIMD mode TIMD1 (corresponding to the minimum prediction cost) : cost (TIMD2 < 2*cost (TIMD1) . If the condition is fulfilled, a fusion is applied, otherwise the TIMD mode added to the MPM list is the first TIMD mode TIMD1.
[0134] Figure 11 shows an example of the first template T1 of a block (current CU) used for template prediction, and an example of the second template used for calculating a prediction cost. PDPC weights are obtained by applying a PDPC (Position dependent intra prediction combination) filtering to the predictions derived based on the two selected TIMD modes. The PDPC weights are similar to what is used in VVC’s PDPC process where the weights depend on the block dimensions. The predictions derived based on the two selected TIMD modes may then be blended with the PDPC weights such weighted intra prediction is used to code the current block.
[0135] The weights may be computed from their prediction costs as follows: weight1 = costMode2 / (costMode1+ costMode2)
[0136] weight2 = 1 -weight1TIMD mode may then be associated with information that may indicate the two selected TIMD modes and the fusion (blending) flag indicating if the blending of the two selected TIMD modes is enabled. Said associated information may also indicate the weights weight1 and weight2 and / or the prediction cost. Information associated with TIMD mode may be stored in memory for a block of the video picture VP that is encoded / decoded based on a prediction block derived based on the TIMD mode.
[0137] ECM further defines an intra prediction fusion method that may be used to derive prediction block of a current block as a weighted combination of multiple intra-prediction blocks generated from different reference lines. Reference line refers to the samples in a certain row and / or column from the reconstructed area used for intra prediction. For example, the immediate row of samples from the above reference side of the current block is referred to as reference line 0. Figure 12 illustrates examples of reference lines. Basically, in the intra prediction fusion method, multiple intra-prediction blocks are first generated and then fused by weighted averaging. The intra prediction fusion method may be summarized as follows: For angular intra prediction modes including the single mode case of TIMD and DIMD, the intra prediction fusion method derives intra prediction block by weighting intra prediction blocks obtained from multiple reference lines represented as pfusion=w0pline+w1pline+1, where pline is the intra prediction block from a reference line 0 (immediate reference line from the above and / or left block boundaries) and pline+1 is the intra prediction block from the reference line above the default reference line. The weights are set as w0=3 / 4 and w1=1 / 4. The angular intra prediction modes may be obtained in various ways. For example, the same intra modes which are derived using TIMD and DIMD modes may be used. Or they may be decided based on rate-distortion optimizations in the encoder side and signaled in a bitstream to the decoder.
[0138] For TIMD mode with blending, pline is used for the first selected TIMD mode TIMD1 (w0=1, w1=0) and pline+1 is used for the second selected TIMD mode TIMD2 (w0=0, w1=1) .
[0139] For DIMD mode with blending, the number of intra prediction blocks selected for a weighted average is increased from 3 to 6.
[0140] Intra prediction fusion method is applied to luma blocks when angular intra mode has non-integer slope (required reference samples interpolation) and the block size is greater than 16, it is used with MRL (Multiple Reference Line) as illustrated in Figure 12 and not applied for ISP coded blocks. ISP is an intra sub-partition as defined in JVET-Q2002.
[0141] Deriving template-based intra prediction modes, such as DIMD and TIMD, for a current block requires memory access to reconstructed samples in a neighbourhood of the current block (called a template) . Memory access to samples of a template occurs when the reconstructing process of the current block begins and thus introduces additional steps (latency) in the reconstructing pipeline of a decoder.
[0142] Figure 13 shows an example of template of a current block used for deriving template-based intra prediction modes for predicting the current block.
[0143] As illustrated in Figure 13, a template is usually defined as a L-shaped template (or a portion of a L-shaped template) around the current block (hatched blocks) . Only the left or top portion of the L-shaped template can be used.
[0144] The template may comprise samples of at least one neighbouring block of the current block. For example, all the samples of a neighbouring block may be samples of the template as illustrated in Figure 13 by the block 131. For example, only portion of the samples of a neighbouring block as illustrated in Figure 13 by the block 132 may be samples of the template.
[0145] Template-based intra prediction of current block to be predicted is determined based on template-based intra prediction mode information.
[0146] Deriving template-based intra prediction mode information may comprise steps for retrieving lines of reconstructed samples in a template around the current block, possibly applying filtering, computing weights and / or computing histograms of gradients.
[0147] The above steps for deriving template-based intra prediction mode information of the current block introduce additional latency in reconstructing the current block because reconstructing the current block, based on a template-based intra prediction mode, requires waiting for reconstruction of all required samples of the template before deriving the template-based intra prediction mode information of the current block.
[0148] Figure 14 shows a block diagram of a method of reconstructing a current block based on template-based intra prediction mode information in accordance with prior art.
[0149] The grey shaded blocks indicate the steps of reconstructing the current block that introduce latency and memory access issues. The dashed blocks may be optional. For example, DIMD and / or TIMD may not require the samples of the template to be filtered, or they may not require calculating blending weights if for example the method only aims in deriving one template-based intra prediction mode for the current block. On the other hand, blocks with solid lines are mandatory and considered to be executed always for reconstructing a current block.
[0150] The problem is to reduce the latency and / or memory access for deriving the template-based intra prediction mode information of a current block.
[0151] At least one exemplary embodiment of the present application has been devised with the foregoing in mind.
[0152] Embodiments of the present disclosure relate to a method of determining template-based intra prediction of a block of a video picture based on template-based intra prediction mode information derived based on at least one reconstructed sample of a template. The template comprises at least one reconstructed sample of a current block neighboring the template-based intra predicted block and template-based intra prediction mode information is at least partially derived at the time of reconstructing the at least one sample of the current block.
[0153] Because the template-based intra prediction mode information of the template-based intra predicted block is at least partially derived at the time samples of the current block is reconstructed, the template-based intra prediction mode information of the template-based intra predicted block is already determined (at least partially) at the time of reconstructing the samples of the template-based intra predicted block. Hence, one of the advantages of embodiments of the present disclosure is reducing the latency. Moreover, embodiments of the present disclosure avoid fetching reconstructed samples from previous blocks. Instead, the template-based intra prediction mode information that has already been calculated is fetched. This means another advantage of embodiments of the present disclosure is reducing the memory access at the time of reconstructing the template-based intra predicted block.
[0154] In the following, embodiments of the present disclosure are described in the case of DIMD and TIMD modes. But the present disclosure extends to any other template-based intra prediction mode.
[0155] In the following, the term “template” refers to reconstructed samples of a neighbourhood of a template-based intra predicted block that is used for deriving the template-based intra prediction mode information. A template may comprise all reconstructed samples of a L-shaped form or may only comprise at least one reconstructed sample in the left portion of the L-shaped template, or it may only comprise at least one reconstructed sample in the above portion of the L-shaped template.
[0156] Figure 15 illustrates an example of template in accordance with at least one embodiment of the present disclosure. The size of the above and left portions of the template may be different or it may be the same. For example, the size of each portion of the template may depend on the current block size (height and width) .
[0157] In the following, the terms “reconstructing” means the process of building samples of a block by adding a prediction of the samples to residual. Reconstructing a block means reconstructing the samples of the block.
[0158] In some embodiments, the template-based intra mode information may comprise at least one of the following information:
[0159] - at least one intra prediction direction or mode;
[0160] - at least one gradient value;
[0161] - at least one histogram of gradients;
[0162] - at least one blending or fusion weight;
[0163] - at least one template cost.
[0164] At the time of reconstructing samples of a current block neighbouring a template-based intra predicted block to be reconstructed, a template-based intra mode information may be stored at least partially in memory. At the time the stored template-based intra mode information is required for reconstructing the template-based intra predicted block , at least a part of template-based intra mode information is fetched from memory.
[0165] For example, if DIMD information is required for reconstructing a DIMD-based predicted block neighbouring a current block then at the time of reconstructing samples of the current block, DIMD information may be obtained. DIMD information may comprise calculated HoG information based on reconstructed samples of the template (including at least one reconstructed sample of the current block) . For example, the HoG information may be calculated based on reconstructed samples located in the above and left portions of the template. The calculated HoG information may be stored. At the time of reconstructing the DIMD-based predicted block, the HoG information may be fetched and then the fetched HoG information are accumulated, and DIMD-based prediction of the DIMD-based predicted block may be determined using the accumulated HoG information as illustrated on Figure 16.
[0166] For example, if DIMD information is required for reconstructing a DIMD-based predicted block neighbouring a current block then at the time of reconstructing the current block, deriving DIMD information may comprise selecting at least one mode with highest HoG amplitude based on reconstructed samples of the template (including at least one reconstructed sample of the current block) and at the time of reconstructing the DIMD-based predicted block, the at least one selected mode may be used as intra prediction modes for the DIMD-based predicted block.
[0167] For example, if TIMD information is required for reconstructing a TIMD-based predicted block neighbouring a current block then at the time of reconstructing samples of the template (including at least one sample of the current block) , deriving TIMD information may comprise calculating intra modes and prediction costs of the intra modes based on at least one reconstructed sample of the template. At the time of reconstructing the TIMD-based predicted block, the intra modes and prediction costs may be fetched to derive TIMD information based on aggregation of intra modes and prediction costs.
[0168] In some embodiments, template-based intra prediction mode information is stored in memory at block level.
[0169] In some embodiments, the blocks of the template are reconstructed according to a decoding order.
[0170] Figure 15 illustrates an example of decoding order of blocks of the template in accordance with at least one of exemplary embodiment of the present disclosure.
[0171] In the example of Figure 15, block 7 refers to a template-based intra predicted block neighbouring of a current block (one of blocks 1 to 6) . Block 7 is going to be reconstructed based on template-based intra prediction mode information. Let’s assume block 1 is reconstructed followed by reconstructing block 2 followed by reconstructing block 3 followed by reconstructing block 4 followed by reconstructing block 5 followed by reconstructing block 6 and followed by reconstructing block 7. The template-based intra prediction mode information of the block 7 is derived and stored at the time of reconstructing the blocks 1 to 6 and the stored template-based intra prediction mode information is fetched from memory at the time of reconstructing block 7.
[0172] For example, if block 7 is going to be reconstructed using DIMD mode, then HoGs information is calculated (possible partially) for each blocks 1 to 6 based on at least one reconstructed sample in the blocks 1 to 6 at the time of reconstructing the at least one block 1 to 6. The HoG information, calculated for each block 1 to 6, is then stored at block level. At the time of reconstructing block 7, the stored HoG information related to blocks 1 to 6 may be fetched and used for determining template-based intra prediction mode information of block 7 at the time of reconstructing the block 7. Thus, at the time of reconstructing block 7 the reconstructed samples of blocks 1 to 6 required for deriving the template-based intra prediction mode information of the blocks 1 to 6 do not need to be re-accessed at the time of reconstructing the block 7.
[0173] Similarly, if block 7 is going to be reconstructed using TIMD mode, instead of re-accessing the reconstructed samples of blocks 1 to 6 to derive the template-based intra prediction mode information at the time of reconstructing block 7, the template-based intra prediction mode information is calculated and stored at the time of reconstructing the blocks 1 to 6 and the template-based intra prediction mode information is fetched from memory at the time of reconstructing block 7.
[0174] In some embodiments, the template-based intra prediction mode information may only contain selected information among a list of information related to the template-based intra prediction mode.
[0175] This is advantageous because storage requirement is reduced.
[0176] For example, when HoGs in DIMD mode are calculated based on reconstructed samples of the template, only M gradient values (with highest magnitudes) among N candidate gradient values may be stored. Accordingly, only the M modes associated with the selected HoGs for storage are stored.
[0177] For example, for TIMD, only M best prediction modes with lowest prediction cost values may be stored among N candidate prediction modes.
[0178] In some embodiments, TIMD costs may be normalized before storing in order to take into account the number of samples on which the cost is computed.
[0179] Precisely, if the cost metric is average-based, an actual TIMD block may normalize different costs coming from blocks of the template having different sizes.
[0180] In some embodiments, the HoG values in a DIMD mode may be normalized before storing them in the memory storing in order to take into account the number of samples on which the HoGs are computed.
[0181] In some embodiment, template-based intra mode information may be added to a history memory which stores template-based intra mode information for a group of blocks in a certain granularity such as CTU, slice, tile.
[0182] The history memory may then be used as potential prediction candidates for one or more of future blocks to be reconstructed.
[0183] In some embodiments, at least one block of the template is reconstructed based on a template-based intra prediction.
[0184] Figure 17 illustrates an example of DIMD-based predicted block of a template of a current block in accordance with at least one exemplary embodiment of the present disclosure.
[0185] Two DIMD-based predicted blocks B1 and B2 are illustrated where blocks B1 and B2 are immediate neighbours according to a decoding order. A template comprising 5 hatched blocks is defined around block B1. At least one reconstructed sample of the hatched blocks is used for deriving the template-based intra prediction information of block B1. The template of block B2 may comprise three hatched blocks neighbouring block B2 and block B1. Then, the template-based intra mode information for block B2 is derived at the time of reconstructing block B1, i.e. once the at least one sample of block B1 that is required for deriving the template-based intra prediction information of block B2 is reconstructed.
[0186] In some embodiments, first template-based intra prediction mode information may be derived for a first block based on at least one reconstructed sample of a third block of a template defined around the first block and second template-based intra prediction mode information may be derived for a second block based on at least one reconstructed sample of the third block of a template defined around the second block.
[0187] Figure 18 illustrates an example of deriving multiple template-based intra prediction information in accordance with at least one exemplary embodiment of the present disclosure.
[0188] Blocks B1 and B2 are intended to be reconstructed based on template-based intra prediction mode information. The template-based intra prediction mode information of block B1 is derived from reconstructed samples of a block B3 that belong to a template of block B1. For example, the reconstructed samples S31 and S33 of block B3 located at the right of block B3 are used. The template-based intra prediction information of block B2 is derived from reconstructed samples of a block B3 that belong to a template of block B2. For example, the reconstructed samples S32 and S33 of block B3 located on the bottom of block B3 are used. At the time of reconstructing block B3, the template-based intra prediction information of block B1 is derived and stored when the required samples at the right of block B3 are reconstructed. At the time of reconstructing block B3, the template-based intra prediction information of block B2 is derived and stored when the required samples on the bottom of block B3 are reconstructed.
[0189] In some embodiments, at least one reconstructed sample of the template may be used for deriving more than one template-based intra prediction mode information.
[0190] For example, as illustrated on Figure 18, the reconstructed samples S33 of block B3 are used for deriving the template-based intra prediction mode information of block B1 and the template-based intra prediction mode information of block B2.
[0191] In some embodiments, template-based intra prediction mode information is derived based on all reconstructed samples of the template defined around the current block.
[0192] In some embodiments, template-based intra prediction mode information may be derived based on reconstructed samples of at least one sub-block of the template and the template-based intra prediction mode information may be stored at sub-block level. The sub-block size is lower than the neighbouring block size.
[0193] For example, a MxN block of the template (or the whole template) may be split into multiple sub-blocks, for example mxn with m<=M and n<=N, where template-based intra prediction mode information is derived for each sub-block within the MxN block. Then the derived information is stored in the memory of the MxN block for every mxn sub-block.
[0194] For example, the template may be split into 8x8 or 4x4 sub-blocks and HoGs may be calculated for every 8x8 or 4x4 sub-block of the template. The HoGs may be stored at sub-block level i.e. each HoG calculated based on samples of a sub-block is stored in relation with said sub-block.
[0195] In some embodiments, the size of the sub-blocks may be fixed or may be determined based on the template size or the current block size.
[0196] Deriving template-based intra prediction mode information at a sub-block level improves the quality of the template-based intra prediction while keeping the latency and memory access requirements low.
[0197] Figure 19 shows an example of template-based intra prediction mode information derivation at sub-block level in accordance with at least one exemplary embodiment of the present disclosure.
[0198] Blocks B1, B2 and B3 require template-based intra prediction mode information at the time of their reconstruction. The template-based intra prediction mode information of each block B1, B2 and B3 is derived based on at least one reconstructed sample located at the bottom of a block B4 that belongs to the templates of block B1, block B2 and block B3. If a single template-based intra prediction mode information would be derived for blocks B1, B2 and B3 then the single template-based intra prediction mode information would not be optimal prediction as correlation of samples a block, for example B1, is higher to samples of template of block B4 (that are closer to samples of block B1) than samples of template of the other blocks B2 and B3.
[0199] In Figure 19, the bottom of block B4 is split into 10 sub-blocks SB0 to SB9. Template-based intra prediction mode information Ti (i=0 to 9) is derived respectively based at least one reconstructed sample of the sub-blocks SB0 to SB9. For example, the template-based intra prediction mode information for B1 is derived based on the template-based intra prediction mode information of T0 to T4 (i.e. template based intra prediction information derived for sub-blocks SB0 to SB4) , the template-based intra prediction mode information for B2 is derived based on the template-based intra prediction mode information of T4 to T7 (i.e. template based intra prediction information derived for sub-blocks SB4 to SB7) and the template-based intra prediction mode information for B3 is derived based on the template-based intra prediction mode information of T7 to T9 (i.e. template based intra prediction information derived for sub-blocks SB7 to SB9) .
[0200] Deriving template-based intra prediction information based on reconstructed samples of a block may require accessing samples outside of the block borders. However, those samples outside of the block borders may not be reconstructed yet because template-based intra prediction information is derived at the same time as reconstructing of the block.
[0201] In some embodiments, deriving template-based intra prediction information of a block may exclude samples outside of a border of the block.
[0202] This embodiment may have impact on accuracy of the template-based intra prediction of a block, especially for a block having small size because border samples are relatively numerous compared to the total number of samples of the block.
[0203] In some embodiments, deriving template-based intra prediction information of a block may be based on reconstructed samples outside of a border of the block.
[0204] This embodiment may suffer from latency issue because the template-based intra prediction information of the block has to wait the reconstruction of the samples outside of the border.
[0205] In some embodiments, samples outside of the block borders may be padded based on already reconstructed samples within the block.
[0206] For example, a replication padding may be applied from samples within the block.
[0207] Figure 20 shows an example of a 3x3 Sobel filter that is applied on reconstructed samples of a block of a template to derive template-based intra prediction mode information in accordance with at least one exemplary embodiment of the present disclosure.
[0208] In the example, the 3x3 Sobel filter has to be applied on reconstructed sample of the template of a block with template-based intra prediction mode located at the right border of the template. Three samples of the template may be padded to apply the Sobel filter on this sample.
[0209] In some embodiments, template-based intra prediction mode information may be derived only when all the samples of the template are reconstructed.
[0210] This embodiment is advantageous because it allows using reconstructed samples of the template without checking if the required samples of the block are reconstructed or not.
[0211] In some embodiments, template-based intra prediction mode information may be derived once at least one sample within a reference block is reconstructed, the reference block referring to a block that the template for template-based intra prediction mode , e.g. DIMD / TIMD is located.
[0212] Figure 21 illustrates an example of deriving template-based intra prediction mode information once at least one required sample of the template is reconstructed in accordance with at least one exemplary embodiment of the present disclosure.
[0213] Samples S1 are reconstructed samples in the template and samples S2 of the template are not yet reconstructed. The derivation is enabled to start once the samples S1 are reconstructed and before the reconstructing of samples S2 of the template. This enabled the parallel reconstruction of block samples and template-based intra mode derivation processes. The samples S2 will be used also in the template-based intra mode information derivation once some or all of them are reconstructed.
[0214] This embodiment further reduces latency for deriving template-based intra prediction mode information compared to the case all the samples of a block are reconstructed because the reconstruction of a block samples and the derivation of template-based intra prediction mode information are jointly executed. Moreover, the memory access for deriving template-based intra prediction mode information is also significantly reduced because the reconstructed samples do not need to be fetched from memory for deriving template-based intra prediction mode information.
[0215] In some embodiments, template-based intra prediction mode information is derived at the time of reconstructing at least one sample of the template only if a decoder is aware that template-based intra prediction mode information derived from the at least one reconstructed sample of the template may be used for predicting at least one other block to be reconstructed.
[0216] In some embodiments, template-based intra prediction mode information is derived at the time of reconstructing each block of a video picture.
[0217] Then, each time a block of the video picture is reconstructed, the template-based intra prediction information of at least one future block to be reconstructed must be derived. This may be a default behaviour of a decoder.
[0218] In some embodiments, template-based intra prediction mode derivation information, e.g. a flag or index, may be signalled in a bitstream to indicate that template-based intra prediction mode information is derived based on reconstructed samples of a block at the time of reconstructing said block.
[0219] In some embodiments, the template-based intra prediction mode derivation information is carried by at least one syntax element related to prediction modes for a group of CUs (blocks) .
[0220] In some embodiments, template-based intra prediction mode information is derived at the time of encoding each block of a video picture. This may be a default behaviour of an encoder. Thus, regardless of whether the template-based intra mode information will be used in future blocks or not, the template-based intra mode information is calculated and stored.
[0221] In some embodiments, in the encoder side, template-based intra prediction mode information is derived, according to the prior art where the template based intra mode information is derived for a block if the encoding mode of the block is a template-based intra mode. This is done by fetching the template samples from the neighbourhood and calculating the template-based intra mode information.
[0222] The encoder and decoder, in both scenarios described in above, generate identical prediction results, but the decoder has less latency and memory access requirements than encoder.
[0223] The blocks in some CTU borders may not be aware of the prediction information for blocks of next CTUs at the sample reconstructing stage.
[0224] In some embodiments, at the time of reconstructing or reconstructing of samples of a block located at a boundary between a current CTU and a next CTU, template-based intra prediction mode information is enforced to be derived at the time of reconstructing a block of the next CTU located at the boundary.
[0225] Figure 22 illustrates an example of reconstructing a current block located at a boundary of a current CTU in accordance with at least one exemplary embodiment of the present disclosure.
[0226] At the time of the reconstructing of the current block B1, located in the left side of a boundary B separating a current CTU with a next CTU, the decoder is not aware that the block B2 of the next CTU located in the right side of the boundary B requires template-based intra prediction mode information at the time of reconstructing the block B2 of the next CTU. Thus, the decoder may enforce deriving template-based intra prediction information at the time of reconstructing the block B1 of the current CTU.
[0227] In some embodiments, template-based intra prediction mode information is not derived at the time of reconstructing a block located at a border of a CTU.
[0228] Then, the template-based intra prediction mode information is derived at the time of reconstructing the block located at a border of a CTU if said block requires to be predicted based on template-based intra prediction mode information.
[0229] Signaling the template-based intra prediction mode derivation information for blocks may be done in certain granularity.
[0230] In some embodiments, template-based intra prediction mode derivation information may be signaled in sequence parameter set, picture parameter set or in picture header, or at slice, tile, CTU, CU, PU, or TU level..
[0231] Figure 23 illustrates an example signalling template-based intra prediction mode derivation information at sequence parameter set (SPS) in accordance with at least one exemplary embodiments of the present disclosure.
[0232] For example, a flag sps_proposed_enabled_flag may be added to the well-known SPS to indicate that if DIMD prediction mode is enabled then DIMD information may be derived at the time of reconstructing blocks of the video (sequence) .
[0233] In some embodiments, the derived template-based intra mode derivation information may be used for template-based intra prediction mode in a different channel. For example, if a chroma block is coded in DIMD chroma mode it requires HoG derivation from both chroma and co-located luma samples (an example is shown in Figure 9) . In order to reduce the latency and memory access, instead of fetching the co-located luma samples for HoG calculation, such information may be calculated and stored at the time of the corresponding luma samples are being reconstructed. When a reconstructing a luma block, if the decoder is aware that some of the samples of the luma block will be used for template-based intra mode of a chroma block, then those corresponding information is derived and stored already in the luma blocks CU information.
[0234] Figure 24 illustrates an example signalling template-based intra prediction mode derivation information at sequence parameter set (SPS) in accordance with at least one exemplary embodiments of the present disclosure.
[0235] For example, a flag sps_proposed_enabled_flag may be added to the well-known SPS to indicate that if TIMD prediction mode is enabled then TIMD information may be derived at the time of reconstructing blocks of the video (sequence) .
[0236] In some embodiments, deriving template-based prediction mode information at the time of reconstructing blocks of a video picture is not enabled for blocks having sizes lower than a threshold.
[0237] Typically, the threshold equals 4 (4x4 blocks) since the majority of samples are located on the block borders, therefore, they will need padding or exclusion.
[0238] This embodiment is advantageous because it avoids performance loss on the derived template-based prediction mode information based on these samples.
[0239] In some embodiments, for small blocks having sizes lower than the threshold, a neighbouring block of a small block that tries to fetch template-based intra prediction mode information that should have been derived at the time of the reconstructing of the small block, may check the threshold condition and decide to derive template-based intra prediction mode information at the time of reconstructing the neighbouring block as in prior art.
[0240] Figure 25 shows a schematic block diagram illustrating an example of a system 600 in which various aspects and exemplary embodiments are implemented.
[0241] System 600 may be embedded as one or more devices including the various components described below. In various exemplary embodiments, system 600 may be configured to implement one or more of the aspects described in the present application.
[0242] Examples of equipment that may form all or part of the system 600 include personal computers, laptops, smartphones, tablet computers, digital multimedia set top boxes, digital television receivers, personal video recording systems, connected home appliances, connected vehicles and their associated processing systems, head mounted display devices (HMD, see-through glasses) , projectors (beamers) , “caves” (system including multiple displays) , servers, video encoders, video decoders, post-processors processing output from a video decoder, pre-processors providing input to a video encoder, web servers, video servers (e.g. a broadcast server, a video-on-demand server or a web server) , still or video camera, encoding or decoding chip or any other communication devices. Elements of system 600, singly or in combination, may be embodied in a single integrated circuit (IC) , multiple ICs, and / or discrete components. For example, in at least one exemplary embodiment, the processing and encoder / decoder elements of system 600 may be distributed across multiple ICs and / or discrete components. In various exemplary embodiments, system 600 may be communicatively coupled to other similar systems, or to other electronic devices, via, for example, a communications bus or through dedicated input and / or output ports.
[0243] System 600 may include at least one processor 610 configured to execute instructions loaded therein for implementing, for example, the various aspects described in the present application. Processor 610 may include embedded memory, input output interface, and various other circuitries as known in the art. System 600 may include at least one memory 620 (for example a volatile memory device and / or a non-volatile memory device) . System 600 may include a storage device 640, which may include non-volatile memory and / or volatile memory, including, but not limited to, Electrically Erasable Programmable Read-Only Memory (EEPROM) , Read-Only Memory (ROM) , Programmable Read-Only Memory (PROM) , Random Access Memory (RAM) , Dynamic Random-Access Memory (DRAM) , Static Random-Access Memory (SRAM) , flash, magnetic disk drive, and / or optical disk drive. The storage device 640 may include an internal storage device, an attached storage device, and / or a network accessible storage device, as non-limiting examples.
[0244] System 600 may include an encoder / decoder module 630 configured, for example, to process data to provide encoded / decoded video picture data, and the encoder / decoder module 630 may include its own processor and memory. The encoder / decoder module 630 may represent module (s) that may be included in a device to perform the encoding and / or decoding functions. As is known, a device may include one or both encoding and decoding modules. Additionally, encoder / decoder module 630 may be implemented as a separate element of system 600 or may be incorporated within processor 610 as a combination of hardware and software as known to those skilled in the art.
[0245] Program code to be loaded onto processor 610 or encoder / decoder 630 to perform the various aspects described in the present application may be stored in storage device 640 and subsequently loaded onto memory 620 for execution by processor 610. In accordance with various exemplary embodiments, one or more of processor 610, memory 620, storage device 640, and encoder / decoder module 630 may store one or more of various items during the performance of the processes described in the present application. Such stored items may include, but are not limited to video picture data, information data used for encoding video picture data, a bitstream, matrices, variables, and intermediate or final results from the processing of equations, formulas, operations, and operational logic.
[0246] In several exemplary embodiments, memory inside of the processor 610 and / or the encoder / decoder module 630 may be used to store instructions and to provide working memory for processing that may be performed during encoding or decoding.
[0247] In other exemplary embodiments, however, a memory external to the processing device (for example, the processing device may be either the processor 610 or the encoder / decoder module 630) may be used for one or more of these functions. The external memory may be the memory 620 and / or the storage device 640, for example, a dynamic volatile memory and / or a non-volatile flash memory. In several exemplary embodiments, an external non-volatile flash memory may be used to store the operating system of a television. In at least one exemplary embodiment, a fast external dynamic volatile memory such as a RAM may be used as working memory for video coding and decoding operations, such as for MPEG-2 part 2 (also known as ITU-T Recommendation H. 262 and ISO / IEC 13818-2, also known as MPEG-2 Video) , AVC, HEVC, EVC, VVC, AV1, etc.
[0248] The input to the elements of system 600 may be provided through various input devices as indicated in block 690. Such input devices include, but are not limited to, (i) an RF portion that may receive an RF signal transmitted, for example, over the air by a broadcaster, (ii) a Composite input terminal, (iii) a USB input terminal, (iv) an HDMI input terminal, (v) a bus such as CAN (Controller Area Network) , CAN FD (Controller Area Network Flexible Data-Rate) , FlexRay (ISO 17458) or Ethernet (ISO / IEC 802-3) bus when the present invention is implemented in the automotive domain.
[0249] In various exemplary embodiments, the input devices of block 690 may have associated respective input processing elements as known in the art. For example, the RF portion may be associated with elements necessary for (i) selecting a desired frequency (also referred to as selecting a signal, or band-limiting a signal to a band of frequencies) , (ii) down-converting the selected signal, (iii) band-limiting again to a narrower band of frequencies to select (for example) a signal frequency band which may be referred to as a channel in certain exemplary embodiments, (iv) demodulating the down-converted and band-limited signal, (v) performing error correction, and (vi) demultiplexing to select the desired stream of data packets. The RF portion of various exemplary embodiments may include one or more elements to perform these functions, for example, frequency selectors, signal selectors, band-limiters, channel selectors, filters, downconverters, demodulators, error correctors, and de-multiplexers. The RF portion may include a tuner that performs various of these functions, including, for example, down-converting the received signal to a lower frequency (for example, an intermediate frequency or a near-baseband frequency) or to baseband.
[0250] In one set-top box embodiment, the RF portion and its associated input processing element may receive an RF signal transmitted over a wired (for example, cable) medium. Then, the RF portion may perform frequency selection by filtering, down-converting, and filtering again to a desired frequency band.
[0251] Various exemplary embodiments rearrange the order of the above-described (and other) elements, remove some of these elements, and / or add other elements performing similar or different functions.
[0252] Adding elements may include inserting elements in between existing elements, such as, for example, inserting amplifiers and an analog-to-digital converter. In various exemplary embodiments, the RF portion may include an antenna.
[0253] Additionally, the USB and / or HDMI terminals may include respective interface processors for connecting system 600 to other electronic devices across USB and / or HDMI connections. It is to be understood that various aspects of input processing, for example, Reed-Solomon error correction, may be implemented, for example, within a separate input processing IC or within processor 610 as necessary. Similarly, aspects of USB or HDMI interface processing may be implemented within separate interface ICs or within processor 610 as necessary. The demodulated, error corrected, and demultiplexed stream may be provided to various processing elements, including, for example, processor 610, and encoder / decoder 630 operating in combination with the memory and storage elements to process the data stream as necessary for presentation on an output device.
[0254] Various elements of system 600 may be provided within an integrated housing. Within the integrated housing, the various elements may be interconnected and transmit data therebetween using suitable connection arrangement 690, for example, an internal bus as known in the art, including the I2C bus, wiring, and printed circuit boards.
[0255] The system 600 may include communication interface 650 that enables communication with other devices via communication channel 651. The communication interface 650 may include, but is not limited to, a transceiver configured to transmit and to receive data over communication channel 651. The communication interface 650 may include, but is not limited to, a modem or network card and the communication channel 651 may be implemented, for example, within a wired and / or a wireless medium.
[0256] Data may be streamed to system 600, in various exemplary embodiments, using a Wi-Fi network such as IEEE 802.11. The Wi-Fi signal of these exemplary embodiments may be received over the communications channel 651 and the communications interface 650 which are adapted for Wi-Fi communications. The communications channel 651 of these exemplary embodiments may be typically connected to an access point or router that provides access to outside networks including the Internet for allowing streaming applications and other over-the-top communications.
[0257] Other exemplary embodiments may provide streamed data to the system 600 using a set-top box that delivers the data over the HDMI connection of the input block 690.
[0258] Still other exemplary embodiments may provide streamed data to the system 600 using the RF connection of the input block 690.
[0259] The streamed data may be used as a way for signaling information used by the system 600. The signaling information may comprise the bitstream B and / or information such a number of pixels of a video picture and / or any coding / decoding setup parameters.
[0260] It is to be appreciated that signaling may be accomplished in a variety of ways. For example, one or more syntax elements, flags, and so forth may be used to signal information to a corresponding decoder in various exemplary embodiments.
[0261] System 600 may provide an output signal to various output devices, including a display 661, speakers 671, and other peripheral devices 681. The other peripheral devices 681 may include, in various examples of exemplary embodiments, one or more of a stand-alone DVR, a disk player, a stereo system, a lighting system, and other devices that provide a function based on the output of system 600.
[0262] In various exemplary embodiments, control signals may be communicated between the system 600 and the display 661, speakers 671, or other peripheral devices 681 using signaling such as AV. Link (Audio / Video Link) , CEC (Consumer Electronics Control) , or other communications protocols that enable device-to-device control with or without user intervention.
[0263] The output devices may be communicatively coupled to system 600 via dedicated connections through respective interfaces 660, 670, and 680.
[0264] Alternatively, the output devices may be connected to system 600 using the communications channel 651 via the communications interface 650. The display 661 and speakers 671 may be integrated in a single unit with the other components of system 600 in an electronic device such as, for example, a television.
[0265] In various exemplary embodiments, the display interface 660 may include a display driver, such as, for example, a timing controller (T Con) chip.
[0266] The display 661 and speaker 671 may alternatively be separate from one or more of the other components, for example, if the RF portion of input 690 is part of a separate set-top box. In various exemplary embodiments in which the display 661 and speakers 671 may be external components, the output signal may be provided via dedicated output connections, including, for example, HDMI ports, USB ports, or COMP outputs.
[0267] In Figures 1-24, various methods are described herein, and each of the methods includes one or more blocks or actions for achieving the described method. Unless a specific order of blocks or actions is required for proper operation of the method, the order and / or use of specific blocks and / or actions may be modified or combined.
[0268] Some examples are described with regard to block diagrams and / or operational flowcharts. Each block represents a circuit element, module, or portion of code which includes one or more executable instructions for implementing the specified logical function (s) . It should also be noted that in other implementations, the function (s) noted in the blocks may occur out of the indicated order. For example, two blocks shown in succession may, in fact, be executed substantially concurrently or the blocks may sometimes be executed in the reverse order, depending on the functionality involved.
[0269] The implementations and aspects described herein may be implemented in, for example, a method or a process, an apparatus, a computer program, a data stream, a bitstream, or a signal. Even if only discussed in the context of a single form of implementation (for example, discussed only as a method) , the implementation of features discussed may also be implemented in other forms (for example, an apparatus or computer program) .
[0270] The methods may be implemented in, for example, a processor, which refers to processing devices in general, including, for example, a computer, a microprocessor, an integrated circuit, or a programmable logic device. Processors also include communication devices.
[0271] Additionally, the methods may be implemented by instructions being performed by a processor, and such instructions (and / or data values produced by an implementation) may be stored on a computer readable storage medium. A computer readable storage medium may take the form of a computer readable program product embodied in one or more computer readable medium (s) and having computer readable program code embodied thereon that is executable by a computer. A computer readable storage medium as used herein may be considered a non-transitory storage medium given the inherent capability to store the information therein as well as the inherent capability to provide retrieval of the information therefrom. A computer readable storage medium may be, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. It is to be appreciated that the following, while providing more specific examples of computer readable storage mediums to which the present exemplary embodiments may be applied, is merely an illustrative and not an exhaustive listing as is readily appreciated by one of ordinary skill in the art: a portable computer diskette; a hard disk; a read-only memory (ROM) ; an erasable programmable read-only memory (EPROM or Flash memory) ; a portable compact disc read-only memory (CD-ROM) ; an optical storage device; a magnetic storage device; or any suitable combination of the foregoing.
[0272] The instructions may form an application program tangibly embodied on a processor-readable medium.
[0273] Instructions may be, for example, in hardware, firmware, software, or a combination. Instructions may be found in, for example, an operating system, a separate application, or a combination of the two. A processor may be characterized, therefore, as, for example, both a device configured to carry out a process and a device that includes a processor-readable medium (such as a storage device) having instructions for carrying out a process. Further, a processor-readable medium may store, in addition to or in lieu of instructions, data values produced by an implementation.
[0274] An apparatus may be implemented in, for example, appropriate hardware, software, and firmware. Examples of such apparatus include personal computers, laptops, smartphones, tablet computers, digital multimedia set top boxes, digital television receivers, personal video recording systems, connected home appliances, head mounted display devices (HMD, see-through glasses) , projectors (beamers) , “caves” (system including multiple displays) , servers, video encoders, video decoders, post-processors processing output from a video decoder, pre-processors providing input to a video encoder, web servers, set-top boxes, and any other device for processing video pictures or other communication devices. As should be clear, the equipment may be mobile and even installed in a mobile vehicle.
[0275] Computer software may be implemented by the processor 610 or by hardware, or by a combination of hardware and software. As a non-limiting example, the exemplary embodiments may be also implemented by one or more integrated circuits. The memory 620 may be of any type appropriate to the technical environment and may be implemented using any appropriate data storage technology, such as optical memory devices, magnetic memory devices, semiconductor-based memory devices, fixed memory, and removable memory, as non-limiting examples. The processor 610 may be of any type appropriate to the technical environment, and may encompass one or more of microprocessors, general purpose computers, special purpose computers, and processors based on a multi-core architecture, as non-limiting examples.
[0276] As will be evident to one of ordinary skill in the art, implementations may produce a variety of signals formatted to carry information that may be, for example, stored or transmitted. The information may include, for example, instructions for performing a method, or data produced by one of the described implementations. -For example, a signal may be formatted to carry the bitstream of a described exemplary embodiment. Such a signal may be formatted, for example, as an electromagnetic wave (for example, using a radio frequency portion of spectrum) or as a baseband signal. The formatting may include, for example, encoding a data stream and modulating a carrier with the encoded data stream. The information that the signal carries may be, for example, analog or digital information. The signal may be transmitted over a variety of different wired or wireless links, as is known. The signal may be stored on a processor-readable medium.
[0277] The terminology used herein is for the purpose of describing particular exemplary embodiments only and is not intended to be limiting. As used herein, the singular forms "a" , "an" , and "the" may be intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "includes / comprises" and / or "including / comprising" when used in this specification, may specify the presence of stated, for example, features, integers, blocks, operations, elements, and / or components but do not preclude the presence or addition of one or more other features, integers, blocks, operations, elements, components, and / or groups thereof. Moreover, when an element is referred to as being "responsive" or "connected" or “associated with” to another element, it may be directly responsive or connected to or associated with the other element, or intervening elements may be present. In contrast, when an element is referred to as being "directly responsive" or "directly connected" to or “directly associated with” other element, there are no intervening elements present.
[0278] It is to be appreciated that the use of any of the symbol / term “ / ” , “and / or” , and “at least one of” , for example, in the cases of “A / B” , “A and / or B” and “at least one of A and B” , may be intended to encompass the selection of the first listed option (A) only, or the selection of the second listed option (B) only, or the selection of both options (A and B) . As a further example, in the cases of “A, B, and / or C” and “at least one of A, B, and C” , such phrasing is intended to encompass the selection of the first listed option (A) only, or the selection of the second listed option (B) only, or the selection of the third listed option (C) only, or the selection of the first and the second listed options (A and B) only, or the selection of the first and third listed options (A and C) only, or the selection of the second and third listed options (B and C) only, or the selection of all three options (A and B and C) . This may be extended, as is clear to one of ordinary skill in this and related arts, for as many items as are listed.
[0279] Various numeric values may be used in the present application. The specific values may be for example purposes and the aspects described are not limited to these specific values.
[0280] It will be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements are not limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element without departing from the teachings of the present application. No ordering is implied between a first element and a second element.
[0281] Reference to “one exemplary embodiment” or “an exemplary embodiment” or “one implementation” or “an implementation” , as well as other variations thereof, is frequently used to convey that a particular feature, structure, characteristic, and so forth (described in connection with the exemplary embodiment / implementation) is included in at least one exemplary embodiment / implementation. Thus, the appearances of the phrase “in one exemplary embodiment” or “in an exemplary embodiment” or “in one implementation” or “in an implementation” , as well any other variations, appearing in various places throughout the present application are not necessarily all referring to the same exemplary embodiment.
[0282] Similarly, reference herein to “in accordance with an exemplary embodiment / example / implementation” or “in an exemplary embodiment / example / implementation” , as well as other variations thereof, is frequently used to convey that a particular feature, structure, or characteristic (described in connection with the exemplary embodiment / example / implementation) may be included in at least one exemplary embodiment / example / implementation. Thus, the appearances of the expression “in accordance with an exemplary embodiment / example / implementation” or “in an exemplary embodiment / example / implementation” in various places in the present application are not necessarily all referring to the same exemplary embodiment / example / implementation, nor are separate or alternative exemplary embodiment / examples / implementation necessarily mutually exclusive of other exemplary embodiments / examples / implementation.
[0283] Reference numerals appearing in the claims are by way of illustration only and shall have no limiting effect on the scope of the claims. Although not explicitly described, the present exemplary embodiments / examples and variants may be employed in any combination or sub-combination.
[0284] When a figure. is presented as a flow diagram, it should be understood that it also provides a block diagram of a corresponding apparatus. Similarly, when a figure is presented as a block diagram, it should be understood that it also provides a flow diagram of a corresponding method / process.
[0285] Although some of the diagrams include arrows on communication paths to show a primary direction of communication, it is to be understood that communication may occur in the opposite direction to the depicted arrows.
[0286] Various implementations involve decoding. “Decoding” , as used in this application, may encompass all or part of the processes performed, for example, on a received video picture (including possibly a received bitstream which encodes one or more video picture) in order to produce a final output suitable for display or for further processing in the reconstructed video domain. In various exemplary embodiments, such processes include one or more of the processes typically performed by a decoder. In various exemplary embodiments, such processes also, or alternatively, include processes performed by a decoder of various implementations described in the present application, for example,
[0287] As further examples, in one exemplary embodiment “decoding” may refer only to de-quantizing, in one exemplary embodiment “decoding” may refer to entropy decoding, in another exemplary embodiment “decoding” may refer only to differential decoding, and in another exemplary embodiment “decoding” may refer to combinations of de-quantizing, entropy decoding and differential decoding. Whether the phrase “decoding process” may be intended to refer specifically to a subset of operations or generally to the broader decoding process will be clear based on the context of the specific description and is believed to be well understood by those skilled in the art.
[0288] Various implementations involve encoding. In an analogous way to the above discussion about “decoding” , “encoding” as used in the present application may encompass all or part of the processes performed, for example, on an input video picture in order to produce an output bitstream. In various exemplary embodiments, such processes include one or more of the processes typically performed by an encoder. In various exemplary embodiments, such processes also, or alternatively, include processes performed by an encoder of various implementations described in this application.
[0289] As further examples, in one exemplary embodiment “encoding” may refer only to quantizing, in one exemplary embodiment “encoding” may refer only to entropy encoding, in another exemplary embodiment “encoding” may refer only to differential encoding, and in another exemplary embodiment “encoding” may refer to combinations of quantizing, differential encoding and entropy encoding. Whether the phrase “encoding process” may be intended to refer specifically to a subset of operations or generally to the broader encoding process will be clear based on the context of the specific descriptions and is believed to be well understood by those skilled in the art.
[0290] Additionally, the present application may refer to “obtaining” various pieces of information. Obtaining the information may include one or more of, for example, estimating the information, calculating the information, predicting the information, or retrieving the information from memory, processing the information, moving the information, copying the information, erasing the information, calculating the information, determining the information, predicting the information, or estimating the information.
[0291] Additionally, this application may refer to “receiving” various pieces of information. Receiving the information may include one or more of, for example, accessing the information, or receiving information from a communication network.
[0292] Also, as used herein, the word “signal” refers to, among other things, indicating something to a corresponding decoder. For example, in certain exemplary embodiments the encoder signals a particular information such as coding parameter or encoded video picture data. In this way, in an exemplary embodiment the same parameter may be used at both the encoder side and the decoder side. Thus, for example, an encoder may transmit (explicit signaling) a particular parameter to the decoder so that the decoder may use the same particular parameter. Conversely, if the decoder already has the particular parameter as well as others, then signaling may be used without transmitting (implicit signaling) to simply allow the decoder to know and select the particular parameter. By avoiding transmission of any actual functions, a bit savings is realized in various exemplary embodiments. It is to be appreciated that signaling may be accomplished in a variety of ways. For example, one or more syntax elements, flags, and so forth are used to signal information to a corresponding decoder in various exemplary embodiments. While the preceding relates to the verb form of the word “signal” , the word “signal” may also be used herein as a noun.
[0293] A number of implementations have been described. Nevertheless, it will be understood that various modifications may be made. For example, elements of different implementations may be combined, supplemented, modified, or removed to produce other implementations. Additionally, one of ordinary skill will understand that other structures and processes may be substituted for those disclosed and the resulting implementations will perform at least substantially the same function (s) , in at least substantially the same way (s) , to achieve at least substantially the same result (s) as the implementations disclosed. Accordingly, these and other implementations are contemplated by this application.
Claims
1.A method of determining template-based intra prediction of a block of a video picture based on template-based intra prediction mode information derived based on at least one reconstructed sample of a template, the template comprising at least one reconstructed sample of the current block neighboring the template-based intra predicted block, wherein template-based intra prediction mode information is at least partially derived at the time of reconstructing the at least one sample of the current block.2.The method of claim 1, wherein template-based intra prediction mode information is stored at block level.3.The method of claim 1, wherein template-based intra prediction mode information is derived based on reconstructed samples of at least one sub-block of the template and template-based intra prediction mode information is stored at sub-block level, the sub-block size being lower than the neighboring block size.4.The method of claim 3, wherein the size of the sub-blocks is fixed or is determined based on the template size or the current block size.5.The method of one of claims 1 to 4, wherein at least one reconstructed sample of the template is used for deriving more than one template-based intra prediction mode information.6.The method of one of claims 1 to 5, wherein template-based intra prediction mode information is derived based on all reconstructed samples of the template.7.The method of one of claims 1 to 6, wherein deriving template-based intra prediction information excludes samples outside of a border of the current block or is based on reconstructed samples outside of a border of the current block or is based on padded samples outside of the current block borders based already reconstructing samples within the current block.8.The method of one of claims 1 to 7, wherein template-based intra prediction mode information is derived only when all the samples of the template are reconstructed.9.The method of one of claims 1 to 7, wherein template-based intra prediction mode information is derived once at least one sample within a reference block is reconstructed, the reference block referring to a block that the template for template-based intra prediction mode is located.10.The method of one of claims 1 to 9, wherein template-based intra prediction mode information is derived at the time of reconstructing each block of the video picture.11.The method of one of claims 1 to 10, wherein a flag or an index is signaled in a bitstream to indicate that template-based intra prediction mode information is derived based on reconstructed samples of a block at the time of reconstructing said block.12.The method of one of claims 1 to 11, wherein the video picture being split into multiple adjacent coding tree units, at the time of reconstructing samples of a block located at a boundary between a current coding tree unit and a next coding tree unit, template-based intra prediction mode information is enforced to be derived at the time of reconstructing a block located at the boundary of the next coding tree unit.13.The method of one of claims 1 to 12, wherein the video picture being split into multiple adjacent coding tree units, template-based intra prediction mode information is not derived at the time of reconstructing a block located at a border of a coding tree unit.14.The method of one of claims 12 to 13, wherein the flag or index is signaled in sequence parameter set, picture parameter set or in picture header, or at slice, tile, coding tree unit, block, prediction unit, or transformation unit level.15.The method of one of claims 1 to 14, wherein deriving template-based prediction mode information at the time of reconstructing blocks of a video picture is not enabled for blocks having sizes lower than a threshold.16.An apparatus comprising means for performing one of the methods claimed in any one of claims 1 to 15.17.A computer program product including instructions which, when the program is executed by one or more processors, causes the one or more processors to carry out a method claimed in any one of claims 1 to 15.18.A non-transitory storage medium carrying instructions of program code for executing a method claimed in any one of claims 1 to 15.