(en)An apparatus for coding video data according to certain aspects includes a memory and a processor in communication with the memory. The memory is configured to store video information, such as base layer video information and enhancement layer video information. The processor is configured to determine a value of a current video unit of enhancement layer video information based at least on a weighted inter-layer predictor and a weighted intra-layer predictor of at least one color component of the current video unit.

1.ApplicationNumber: US-201314019468-A
1.PublishNumber: US-9906786-B2
2.Date Publish: 20180227
3.Inventor: SEREGIN VADIM
CHEN JIANLE
LI XIANG
RAPAKA KRISHNAKANTH

4.Inventor Harmonized: SEREGIN VADIM(US)
CHEN JIANLE(US)
LI XIANG(US)
RAPAKA KRISHNAKANTH(US)

5.Country: US
6.Claims:
(en)An apparatus for coding video data according to certain aspects includes a memory and a processor in communication with the memory. The memory is configured to store video information, such as base layer video information and enhancement layer video information. The processor is configured to determine a value of a current video unit of enhancement layer video information based at least on a weighted inter-layer predictor and a weighted intra-layer predictor of at least one color component of the current video unit.

7.Description:
(en)CROSS REFERENCE TO RELATED APPLICATIONS
This application claims a priority benefit from U.S. Provisional Patent Application No. 61/698,526, filed Sep. 7, 2012, from U.S. Provisional Patent Application No. 61/707,843, filed Sep. 28, 2012, from U.S. Provisional Patent Application No. 61/754,490, filed Jan. 18, 2013, from U.S. Provisional Patent Application No. 61/807,271, filed Apr. 1, 2013, from U.S. Provisional Patent Application No. 61/809,800, filed Apr. 8, 2013, from U.S. Provisional Patent Application No. 61/814,288, filed Apr. 21, 2013, and from U.S. Provisional Patent Application No. 61/814,640, filed Apr. 22, 2013, all seven of which are incorporated by reference in their entireties.


TECHNICAL FIELD
This disclosure relates to video coding and compression and, in particular, to scalable video coding (SVC).
BACKGROUND
Digital video capabilities can be incorporated into a wide range of devices, including digital televisions, digital direct broadcast systems, wireless broadcast systems, personal digital assistants (PDAs), laptop or desktop computers, tablet computers, e-book readers, digital cameras, digital recording devices, digital media players, video gaming devices, video game consoles, cellular or satellite radio telephones, so-called “smart phones,” video teleconferencing devices, video streaming devices, and the like. Digital video devices implement video coding techniques, such as those described in the standards defined by MPEG-2, MPEG-4, ITU-T H.263, ITU-T H.264/MPEG-4, Part 10, Advanced Video Coding (AVC), the High Efficiency Video Coding (HEVC) standard presently under development, and extensions of such standards. The video devices may transmit, receive, encode, decode, and/or store digital video information more efficiently by implementing such video coding techniques.
Video coding techniques include spatial (intra-picture) prediction and/or temporal (inter-picture) prediction to reduce or remove redundancy inherent in video sequences. For block-based video coding, a video slice (e.g., a video frame or a portion of a video frame) may be partitioned into video blocks, which may also be referred to as treeblocks, coding units (CUs) and/or coding nodes. Video blocks in an intra-coded (I) slice of a picture are encoded using spatial prediction with respect to reference samples in neighboring blocks in the same picture. Video blocks in an inter-coded (P or B) slice of a picture may use spatial prediction with respect to reference samples in neighboring blocks in the same picture or temporal prediction with respect to reference samples in other reference pictures. Pictures may be referred to as frames, and reference pictures may be referred to a reference frames.
Spatial or temporal prediction results in a predictive block for a block to be coded. Residual data represents pixel differences between the original block to be coded and the predictive block. An inter-coded block is encoded according to a motion vector that points to a block of reference samples forming the predictive block, and the residual data indicating the difference between the coded block and the predictive block. An intra-coded block is encoded according to an intra-coding mode and the residual data. For further compression, the residual data may be transformed from the pixel domain to a transform domain, resulting in residual transform coefficients, which then may be quantized. The quantized transform coefficients, initially arranged in a two-dimensional array, may be scanned in order to produce a one-dimensional vector of transform coefficients, and entropy coding may be applied to achieve even more compression.
Some block-based video coding and compression schemes make use of scalable techniques, such as scalable video coding (SVC). Generally described, scalable video coding refers to video coding in which a base layer and one or more scalable enhancement layers are used. For SVC, a base layer typically carries video data with a base level of quality. One or more enhancement layers carry additional video data to support higher spatial, temporal, and/or quality (also referred to as signal-to-noise ratio or SNR) levels. In some cases, the base layer may be transmitted in a manner that is more reliable than the transmission of enhancement layers.
Each of these layers may include one or more video blocks, which may be coded in a particular coding order (e.g., coded sequentially left-to-right and line-by-line, top-to-bottom). Layers may have the same block size or varying block sizes, depending on the spatial resolution of the layer.
SUMMARY
The systems, methods, and devices of the invention each have several aspects, no single one of which is solely responsible for its desirable attributes. Without limiting the scope of this invention as expressed by the claims which follow, some features will now be discussed briefly. After considering this discussion, and particularly after reading the section entitled “Detailed Description” one will understand how the features of this invention provide advantages that include determining a predictor for a block of video information based on weighted predictor of a first layer of video data and a weighted predictor of a second layer of video data.
In one embodiment, an apparatus for coding digital video is provided. The apparatus includes a memory configured to store base layer video information and enhancement layer video information. The apparatus further includes a processor in communication with the memory. The processor is configured to determine a predictor of a current block of enhancement layer video information based at least on a weighted inter-layer predictor of at least one color component of the current block and a weighted intra-layer predictor of the at least one color component of the current block.
In another embodiment, a method of decoding video is provided. The decoding method includes obtaining a video bitstream defining base layer video information and enhancement layer video information. The method also includes determining a predictor of a current block of enhancement layer video information based at least on a weighted inter-layer predictor of at least one color component of the current block and a weighted intra-layer predictor of the at least one color component of the current block.
In an additional embodiment, a method of encoding video is provided. The encoding method includes generating a video bitstream defining base layer video information and enhancement layer video information. The encoding method also includes determining a predictor of a current block of enhancement layer video information based at least on a weighted inter-layer predictor of at least one color component of the current block and a weighted intra-layer predictor of the at least one color component of the current block.
In a further embodiment, a computer readable storage medium comprising executable instructions is provided. The instructions cause an apparatus to obtain base layer video information and enhancement layer video information. The instructions also cause an apparatus to determine a predictor of a current block of enhancement layer video information based at least on a weighted inter-layer predictor of at least one color component of the current block and a weighted intra-layer predictor of the at least one color component of the current block.
In a still further embodiment, an apparatus for coding digital video is provided. The apparatus includes a means for obtaining base layer video information and enhancement layer video information. The apparatus further includes a means for determining a predictor of a current block of enhancement layer video information based at least on a weighted inter-layer predictor of at least one color component of the current block and a weighted intra-layer predictor of the at least one color component of the current block.
The details of one or more examples are set forth in the accompanying drawings and the description below. Other features, objects, and advantages will be apparent from the description and drawings, and from the claims.



BRIEF DESCRIPTION OF DRAWINGS
FIG. 1 is a block diagram illustrating an example video encoding and decoding system that may utilize techniques in accordance with aspects described in this disclosure.
FIG. 2 is a block diagram illustrating an example of a video encoder that may implement techniques in accordance with aspects described in this disclosure.
FIG. 3 is a block diagram illustrating an example of a video decoder that may implement techniques in accordance with aspects described in this disclosure.
FIG. 4A is a block diagram illustrating example predictions of an enhancement layer video unit.
FIG. 4B is a block diagram illustrating an example weighted mode prediction of an enhancement layer video unit.
FIG. 5 is a flow diagram of an illustrative process for determining a value of a video unit using weighted predictions generated according to one or more prediction modes.
FIG. 6 is a flow diagram of an illustrative process for restricting the use of weighted mode base on block size.
FIG. 7 is a flow diagram of an illustrative process for restricting the use of motion compensation with weighted mode.
FIGS. 8A and 8B are flowcharts illustrating example methods for generalized residual prediction according to aspects of this disclosure.
FIG. 9 is a flowchart illustrating an example method for generalized residual prediction using single-loop decoding according to aspects of this disclosure.
FIGS. 10A and 10B are flowcharts illustrating example methods for generalized residual prediction using multi-loop decoding according to aspects of this disclosure.
FIG. 11 is a flowchart illustrating an example method for signaling generalized residual prediction parameters according to aspects of this disclosure.
FIG. 12 is a flowchart illustrating an example method for deriving generalized residual prediction parameters according to aspects of this disclosure.
FIG. 13 is a flowchart illustrating an example method for up-sampling or down-sampling a layer in generalized residual prediction according to aspects of this disclosure.
FIG. 14 is a flowchart illustrating an example method for remapping, up-sampling, or down-sampling motion information in generalized residual prediction according to aspects of this disclosure.
FIG. 15 is a flowchart illustrating an example method for determining a weighting factor in generalized residual prediction according to aspects of this disclosure.



DETAILED DESCRIPTION
Generally described, the present disclosure relates video coding. During the video coding process, predictions may be made regarding the components of individual video units (e.g., pixels or blocks of pixels). For example, luma and/or chroma values for certain pixels or blocks of pixels may be predicted from other information. This can conserve computing resources (e.g., bandwidth, storage space) because the predicted values may be excluded from a bitstream if the decoder is configured to properly predict the values. In single-layer video coding, the predictions are typically spatial (e.g., based on other video units in the same frame of video being coded) or temporal (e.g., based on corresponding video units of prior or subsequent frames). In scalable video coding (SVC), video information may be logically or physically coded as multiple layers, such as a base layer and any number of enhancement layers. An individual frame of SVC video may therefore encompass multiple layers of video information. In SVC, predictions for blocks in an enhancement layer frame may be made on a spatial or temporal basis, as described above, or on an inter-layer basis (e.g., based on corresponding video units of the base layer for the same frame). In some cases, however, a prediction made according to one of the aforementioned prediction methods (e.g., spatial, temporal, or inter-layer prediction) may not provide a substantial advantage in picture quality over the other prediction modes. In addition, each individual prediction mode has its own limitations. Combining predictions made according to multiple prediction modes may, in some cases, provide an advantage in picture quality over the use of a prediction made according to a single prediction mode.
Aspects of this disclosure relate to combining predictions made according to multiple prediction methods or modes to generate a final prediction for a video unit. In some embodiments, a new prediction mode, which may be referred to as combined mode or weighted mode, may be defined. In combined mode or weighted mode, predictions made according to other individual prediction modes may be weighted and combined to add diversity to the final prediction (e.g., a weighted inter-layer prediction may be combined with a weighted intra-layer prediction). By adding diversity to a final prediction, the result may be a better prediction in cases where, e.g., a single prediction mode does not provide a substantial advantage over other prediction modes or over a combination of predictions made according to multiple prediction modes. For example, a spatial prediction (e.g., an intra-prediction, as described in greater detail below) may provide only marginally better results, or results that are not better at all, than a temporal prediction or an inter-layer prediction (e.g., an inter-prediction or intraBL prediction, respectively, as described in greater detail below). The inter-prediction may be combined with one or more other predictions in a weighted fashion. The predictions may be weighted such that a prediction made according to one mode is weighed more heavily in the final prediction for a given video unit than a prediction made according to a different mode. The combined prediction may optionally be normalized. A final predictor generated in this fashion may provide substantially better results than any single predictor.
In some embodiments, the weights used in weighted mode may be different for one or more of the individual color components (e.g., luma and both chromas) of a video unit. For example, a first set of weights (e.g., 0.5 and 0.5) may be applied to two different predictors, respectively, for the luma component of the video unit. A second set of weights (e.g., 0 and 1) may be applied to two different predictors, respectively, for the chroma components. In some embodiments, weighted mode may be applied selectively to components of a video unit, such as only to the luma component.
The predictions that may be combined to generate a final predictor in the weighted mode or combined mode described herein are not limited to inter-predictions, intra-predictions, and/or inter-layer predictions, such as intraBL predictions. Rather, any prediction of a video unit may be used in the weighted mode. For example, a prediction for an enhancement layer block that is based on a base layer block reconstructed using residual prediction techniques may also be used. In this inter-layer residual prediction, the residue of the base layer may be used to predict the current block or other video unit in the enhancement layer. The residue may be defined as the difference between the temporal prediction for a video unit and a source video unit (e.g., an input video unit). In inter-layer residual prediction, also referred to herein as generalized residual prediction, the residue of the base layer is factored into the prediction of the current enhancement layer current block. Residual prediction techniques are described in greater detail below. Other prediction techniques may also be combined to generate a final predictor in weighted mode.
Additional aspects of the disclosure relate to reducing or eliminating any negative effect that implementation of the weighted mode may have on the performance of video coding, video information storage, or bitstream transmission. In some embodiments, weighted mode may only be used for certain larger block sizes. This can minimize the amount of additional weighted mode information that is included in a coded bitstream. In additional embodiments, certain prediction modes may be limited or restricted when used in weighted mode. For example, only urn-direction motion prediction may be used for enhancement layer blocks when a final predictor is being generated in weighted mode. This can reduce the amount of motion compensation information and associated computational complexity that is included in a coded bitstream. In further embodiments, certain prediction modes may be otherwise altered when used in weighted mode. For example, uni-direction prediction may be used for certain block sizes even if bi-directional motion compensation information is included in a bitstream, thereby reducing coding complexity. As another example, weighted mode may not be used when bi-directional motion compensation information is included, or when blocks to be predicted are below a certain size, even if a weighted mode flag is signaled in the bitstream or weighted mode would otherwise be used.
The techniques described in this disclosure generally relate to scalable video coding (SVC) and 3D video coding. For example, the techniques may be related to, and used with or within, a High Efficiency Video Coding (HEVC) scalable video coding (SVC) extension. SVC may be used to provide quality (also referred to as signal-to-noise (SNR)) scalability, spatial scalability and/or temporal scalability. In an SVC extension, there could be multiple layers of video information. The layer at the very bottom level may serve as a base layer (BL), and the layer at the very top may serve as an enhanced layer (EL). The “enhanced layer” is sometimes referred to as an “enhancement layer,” and these terms may be used interchangeably. All layers in the middle may serve as either or both ELs or BLs. For example, a layer in the middle may be an EL for the layers below it, such as the base layer or any intervening enhancement layers, and at the same time serve as a BL for the enhancement layers above it.
For purposes of illustration only, the techniques described in the disclosure are described with examples including only two layers (e.g., lower level layer such as the base layer, and a higher level layer such as the enhanced layer). It should be understood that the examples described in this disclosure can be extended to examples with multiple base layers and enhancement layers as well.
An enhancement layer may have different spatial resolution than base layer. For example, the spatial aspect ratio between EL and BL can be 1.0, 1.5, 2.0 or other different ratios. In other words, the spatial aspect of the EL may equal 1.0, 1.5, or 2.0 times the spatial aspect of the BL. In some examples, the scaling factor of the EL may be greater than the BL. For example, a size of pictures in the EL may be greater than a size of pictures in the BL. In this way, it may be possible, although not a limitation, that the spatial resolution of the EL is larger than the spatial resolution of the BL.
Video coding standards include ITU-T H.261, ISO/IEC MPEG-1 Visual, ITU-T H.262 or ISO/IEC MPEG-2 Visual, ITU-T H.263, ISO/IEC MPEG-4 Visual and ITU-T H.264 (also known as ISO/IEC MPEG-4 AVC), including its Scalable Video Coding (SVC) and Multiview Video Coding (MVC) extensions. In addition, a new video coding standard, namely High Efficiency Video Coding (HEVC), is being developed by the Joint Collaboration Team on Video Coding (JCT-VC) of ITU-T Video Coding Experts Group (VCEG) and ISO/IEC Motion Picture Experts Group (MPEG). A recent draft of HEVC is available from http://wg11.sc29.org/jct/doc_end_user/current_document.php?id=5885/JCTVC-I1003-v2, as of Jun. 7, 2012. Another recent draft of the HEVC standard, referred to as “HEVC Working Draft 7” is downloadable from http://phenix.it-sudparis.eu/jct/doc_end_user/documents/9_Geneva/wg11/JCTVC-I1003-v3.zip, as of Jun. 7, 2012. The full citation for the HEVC Working Draft 7 is document HCTVC-I1003, Bross et al., “High Efficiency Video Coding (HEVC) Text Specification Draft 7,” Joint Collaborative Team on Video Coding (JCT-VC) of ITU-T SG16 WP3 and ISO/IEC JTC1/SC29/WG11, 9 th Meeting: Geneva, Switzerland, Apr. 27, 2012 to May 7, 2012. Each of these references is incorporated by reference in its entirety.
Various aspects of the novel systems, apparatuses, and methods are described more fully hereinafter with reference to the accompanying drawings. This disclosure may, however, be embodied in many different forms and should not be construed as limited to any specific structure or function presented throughout this disclosure. Rather, these aspects are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art. Based on the teachings herein, one skilled in the art should appreciate that the scope of the disclosure is intended to cover any aspect of the novel systems, apparatuses, and methods disclosed herein, whether implemented independently of, or combined with, any other aspect of the disclosure. For example, an apparatus may be implemented or a method may be practiced using any number or subset of the aspects set forth herein. In addition, the scope of the disclosure is intended to cover such an apparatus or method which is practiced using other structure, functionality, or structure and functionality in addition to or other than the various aspects of the disclosure herein. It should be understood that any aspect disclosed herein may be embodied by one or more elements of a claim.
Although particular aspects are described herein, many variations and permutations of these aspects fall within the scope of the disclosure. Although some benefits and advantages of the preferred aspects are mentioned, the scope of the disclosure is not intended to be limited to particular benefits, uses, or objectives. Rather, aspects of the disclosure are intended to be broadly applicable to different wireless technologies, system configurations, networks, and transmission protocols, some of which are illustrated by way of example in the figures and in the following description of the preferred aspects. The detailed description and drawings are merely illustrative of the disclosure rather than limiting, the scope of the disclosure being defined by the appended claims and equivalents thereof.
FIG. 1 is a block diagram illustrating an example video encoding and decoding system that may utilize techniques in accordance with aspects described in this disclosure. As shown in FIG. 1 , system 10 includes a source device 12 that provides encoded video data to be decoded at a later time by a destination device 14 . In particular, source device 12 provides the video data to destination device 14 via a computer-readable medium 16 . Source device 12 and destination device 14 may comprise any of a wide range of devices, including desktop computers, notebook (e.g., laptop) computers, tablet computers, set-top boxes, telephone handsets such as so-called “smart” phones, so-called “smart” pads, televisions, cameras, display devices, digital media players, video gaming consoles, video streaming device, or the like. In some cases, source device 12 and destination device 14 may be equipped for wireless communication.
Destination device 14 may receive the encoded video data to be decoded via computer-readable medium 16 . Computer-readable medium 16 may comprise any type of medium or device capable of moving the encoded video data from source device 12 to destination device 14 . In one example, computer-readable medium 16 may comprise a communication medium to enable source device 12 to transmit encoded video data directly to destination device 14 in real-time. The encoded video data may be modulated according to a communication standard, such as a wireless communication protocol, and transmitted to destination device 14 . The communication medium may comprise any wireless or wired communication medium, such as a radio frequency (RF) spectrum or one or more physical transmission lines. The communication medium may form part of a packet-based network, such as a local area network, a wide-area network, or a global network such as the Internet. The communication medium may include routers, switches, base stations, or any other equipment that may be useful to facilitate communication from source device 12 to destination device 14 .
In some examples, encoded data may be output from output interface 22 to a storage device. Similarly, encoded data may be accessed from the storage device by input interface. The storage device may include any of a variety of distributed or locally accessed data storage media such as a hard drive, Blu-ray discs, DVDs, CD-ROMs, flash memory, volatile or non-volatile memory, or any other suitable digital storage media for storing encoded video data. In a further example, the storage device may correspond to a file server or another intermediate storage device that may store the encoded video generated by source device 12 . Destination device 14 may access stored video data from the storage device via streaming or download. The file server may be any type of server capable of storing encoded video data and transmitting that encoded video data to the destination device 14 . Example file servers include a web server (e.g., for a website), an FTP server, network attached storage (NAS) devices, or a local disk drive. Destination device 14 may access the encoded video data through any standard data connection, including an Internet connection. This may include a wireless channel (e.g., a Wi-Fi connection), a wired connection (e.g., DSL, cable modem, etc.), or a combination of both that is suitable for accessing encoded video data stored on a file server. The transmission of encoded video data from the storage device may be a streaming transmission, a download transmission, or a combination thereof.
The techniques of this disclosure are not necessarily limited to wireless applications or settings. The techniques may be applied to video coding in support of any of a variety of multimedia applications, such as over-the-air television broadcasts, cable television transmissions, satellite television transmissions, Internet streaming video transmissions, such as dynamic adaptive streaming over HTTP (DASH), digital video that is encoded onto a data storage medium, decoding of digital video stored on a data storage medium, or other applications. In some examples, system 10 may be configured to support one-way or two-way video transmission to support applications such as video streaming, video playback, video broadcasting, and/or video telephony.
In the example of FIG. 1 , source device 12 includes video source 18 , video encoder 20 , and output interface 22 . Destination device 14 includes input interface 28 , video decoder 30 , and display device 32 . In accordance with this disclosure, video encoder 20 of source device 12 may be configured to apply the techniques for coding a bitstream including video data conforming to multiple standards or standard extensions. In other examples, a source device and a destination device may include other components or arrangements. For example, source device 12 may receive video data from an external video source 18 , such as an external camera. Likewise, destination device 14 may interface with an external display device, rather than including an integrated display device.
The illustrated system 10 of FIG. 1 is merely one example. Techniques for determining candidates for a candidate list for motion vector predictors for a current block may be performed by any digital video encoding and/or decoding device. Although generally the techniques of this disclosure are performed by a video encoding device, the techniques may also be performed by a video encoder/decoder, typically referred to as a “CODEC.” Moreover, the techniques of this disclosure may also be performed by a video preprocessor. Source device 12 and destination device 14 are merely examples of such coding devices in which source device 12 generates coded video data for transmission to destination device 14 . In some examples, devices 12 , 14 may operate in a substantially symmetrical manner such that each of devices 12 , 14 include video encoding and decoding components. Hence, system 10 may support one-way or two-way video transmission between video devices 12 , 14 , e.g., for video streaming, video playback, video broadcasting, or video telephony.
Video source 18 of source device 12 may include a video capture device, such as a video camera, a video archive containing previously captured video, and/or a video feed interface to receive video from a video content provider. As a further alternative, video source 18 may generate computer graphics-based data as the source video, or a combination of live video, archived video, and computer-generated video. In some cases, if video source 18 is a video camera, source device 12 and destination device 14 may form so-called camera phones or video phones. As mentioned above, however, the techniques described in this disclosure may be applicable to video coding in general, and may be applied to wireless and/or wired applications. In each case, the captured, pre-captured, or computer-generated video may be encoded by video encoder 20 . The encoded video information may then be output by output interface 22 onto a computer-readable medium 16 .
Computer-readable medium 16 may include transient media, such as a wireless broadcast or wired network transmission, or storage media (that is, non-transitory storage media), such as a hard disk, flash drive, compact disc, digital video disc, Blu-ray disc, or other computer-readable media. In some examples, a network server (not shown) may receive encoded video data from source device 12 and provide the encoded video data to destination device 14 , e.g., via network transmission, direct wired communication, etc. Similarly, a computing device of a medium production facility, such as a disc stamping facility, may receive encoded video data from source device 12 and produce a disc containing the encoded video data. Therefore, computer-readable medium 16 may be understood to include one or more computer-readable media of various forms, in various examples.
Input interface 28 of destination device 14 receives information from computer-readable medium 16 . The information of computer-readable medium 16 may include syntax information defined by video encoder 20 , which is also used by video decoder 30 , that includes syntax elements that describe characteristics and/or processing of blocks and other coded units, e.g., GOPs. Display device 32 displays the decoded video data to a user, and may comprise any of a variety of display devices such as a cathode ray tube (CRT), a liquid crystal display (LCD), a plasma display, an organic light emitting diode (OLED) display, or another type of display device.
Video encoder 20 and video decoder 30 may operate according to a video coding standard, such as the High Efficiency Video Coding (HEVC) standard presently under development, and may conform to the HEVC Test Model (HM). Alternatively, video encoder 20 and video decoder 30 may operate according to other proprietary or industry standards, such as the ITU-T H.264 standard, alternatively referred to as MPEG-4, Part 10, Advanced Video Coding (AVC), or extensions of such standards. The techniques of this disclosure, however, are not limited to any particular coding standard, including but not limited to any of the standards listed above. Other examples of video coding standards include MPEG-2 and ITU-T H.263. Although not shown in FIG. 1 , in some aspects, video encoder 20 and video decoder 30 may each be integrated with an audio encoder and decoder, and may include appropriate MUX-DEMUX units, or other hardware and software, to handle encoding of both audio and video in a common data stream or separate data streams. If applicable, MUX-DEMUX units may conform to the ITU H.223 multiplexer protocol, or other protocols such as the user datagram protocol (UDP).
Video encoder 20 and video decoder 30 each may be implemented as any of a variety of suitable encoder circuitry, such as one or more microprocessors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), discrete logic, software, hardware, firmware or any combinations thereof. When the techniques are implemented partially in software, a device may store instructions for the software in a suitable, non-transitory computer-readable medium and execute the instructions in hardware using one or more processors to perform the techniques of this disclosure. Each of video encoder 20 and video decoder 30 may be included in one or more encoders or decoders, either of which may be integrated as part of a combined encoder/decoder (CODEC) in a respective device. A device including video encoder 20 and/or video decoder 30 may comprise an integrated circuit, a microprocessor, and/or a wireless communication device, such as a cellular telephone.
The JCT-VC is working on development of the HEVC standard. The HEVC standardization efforts are based on an evolving model of a video coding device referred to as the HEVC Test Model (HM). The HM presumes several additional capabilities of video coding devices relative to existing devices according to, e.g., ITU-T H.264/AVC. For example, whereas H.264 provides nine intra-prediction encoding modes, the HM may provide as many as thirty-three intra-prediction encoding modes.
In general, the working model of the HM describes that a video frame or picture may be divided into a sequence of treeblocks or largest coding units (LCU) that include both luma and chroma samples. Syntax data within a bitstream may define a size for the LCU, which is a largest coding unit in terms of the number of pixels. A slice includes a number of consecutive treeblocks in coding order. A video frame or picture may be partitioned into one or more slices. Each treeblock may be split into coding units (CUs) according to a quadtree. In general, a quadtree data structure includes one node per CU, with a root node corresponding to the treeblock. If a CU is split into four sub-CUs, the node corresponding to the CU includes four leaf nodes, each of which corresponds to one of the sub-CUs.
Each node of the quadtree data structure may provide syntax data for the corresponding CU. For example, a node in the quadtree may include a split flag, indicating whether the CU corresponding to the node is split into sub-CUs. Syntax elements for a CU may be defined recursively, and may depend on whether the CU is split into sub-CUs. If a CU is not split further, it is referred as a leaf-CU. In this disclosure, four sub-CUs of a leaf-CU will also be referred to as leaf-CUs even if there is no explicit splitting of the original leaf-CU. For example, if a CU at 16×16 size is not split further, the four 8×8 sub-CUs will also be referred to as leaf-CUs although the 16×16 CU was never split.
A CU has a similar purpose as a macroblock of the H.264 standard, except that a CU does not have a size distinction. For example, a treeblock may be split into four child nodes (also referred to as sub-CUs), and each child node may in turn be a parent node and be split into another four child nodes. A final, unsplit child node, referred to as a leaf node of the quadtree, comprises a coding node, also referred to as a leaf-CU. Syntax data associated with a coded bitstream may define a maximum number of times a treeblock may be split, referred to as a maximum CU depth, and may also define a minimum size of the coding nodes. Accordingly, a bitstream may also define a smallest coding unit (SCU). This disclosure uses the term “block” to refer to any of a CU, PU, or TU, in the context of HEVC, or similar data structures in the context of other standards (e.g., macroblocks and sub-blocks thereof in H.264/AVC).
A CU includes a coding node and prediction units (PUs) and transform units (TUs) associated with the coding node. A size of the CU corresponds to a size of the coding node and must be square in shape. The size of the CU may range from 8×8 pixels up to the size of the treeblock with a maximum of 64×64 pixels or greater. Each CU may contain one or more PUs and one or more TUs. Syntax data associated with a CU may describe, for example, partitioning of the CU into one or more PUs. Partitioning modes may differ between whether the CU is skip or direct mode encoded, intra-prediction mode encoded, or inter-prediction mode encoded. PUs may be partitioned to be non-square in shape. Syntax data associated with a CU may also describe, for example, partitioning of the CU into one or more TUs according to a quadtree. A TU can be square or non-square (e.g., rectangular) in shape.
The HEVC standard allows for transformations according to TUs, which may be different for different CUs. The TUs are typically sized based on the size of PUs within a given CU defined for a partitioned LCU, although this may not always be the case. The TUs are typically the same size or smaller than the PUs. In some examples, residual samples corresponding to a CU may be subdivided into smaller units using a quadtree structure known as “residual quad tree” (RQT). The leaf nodes of the RQT may be referred to as transform units (TUs). Pixel difference values associated with the TUs may be transformed to produce transform coefficients, which may be quantized.
A leaf-CU may include one or more prediction units (PUs). In general, a PU represents a spatial area corresponding to all or a portion of the corresponding CU, and may include data for retrieving a reference sample for the PU. Moreover, a PU includes data related to prediction. For example, when the PU is intra-mode encoded, data for the PU may be included in a residual quadtree (RQT), which may include data describing an intra-prediction mode for a TU corresponding to the PU. As another example, when the PU is inter-mode encoded, the PU may include data defining one or more motion vectors for the PU. The data defining the motion vector for a PU may describe, for example, a horizontal component of the motion vector, a vertical component of the motion vector, a resolution for the motion vector (e.g., one-quarter pixel precision or one-eighth pixel precision), a reference picture to which the motion vector points, and/or a reference picture list (e.g., List 0, List 1, or List C) for the motion vector.
A leaf-CU having one or more PUs may also include one or more transform units (TUs). The transform units may be specified using an RQT (also referred to as a TU quadtree structure), as discussed above. For example, a split flag may indicate whether a leaf-CU is split into four transform units. Then, each transform unit may be split further into further sub-TUs. When a TU is not split further, it may be referred to as a leaf-TU. Generally, for intra coding, all the leaf-TUs belonging to a leaf-CU share the same intra prediction mode. That is, the same intra-prediction mode is generally applied to calculate predicted values for all TUs of a leaf-CU. For intra coding, a video encoder may calculate a residual value for each leaf-TU using the intra prediction mode, as a difference between the portion of the CU corresponding to the TU and the original block. A TU is not necessarily limited to the size of a PU. Thus, TUs may be larger or smaller than a PU. For intra coding, a PU may be collocated with a corresponding leaf-TU for the same CU. In some examples, the maximum size of a leaf-TU may correspond to the size of the corresponding leaf-CU.
Moreover, TUs of leaf-CUs may also be associated with respective quadtree data structures, referred to as residual quadtrees (RQTs). That is, a leaf-CU may include a quadtree indicating how the leaf-CU is partitioned into TUs. The root node of a TU quadtree generally corresponds to a leaf-CU, while the root node of a CU quadtree generally corresponds to a treeblock (or LCU). TUs of the RQT that are not split are referred to as leaf-TUs. In general, this disclosure uses the terms CU and TU to refer to leaf-CU and leaf-TU, respectively, unless noted otherwise.
A video sequence typically includes a series of video frames or pictures. A group of pictures (GOP) generally comprises a series of one or more of the video pictures. A GOP may include syntax data in a header of the GOP, a header of one or more of the pictures, or elsewhere, that describes a number of pictures included in the GOP. Each slice of a picture may include slice syntax data that describes an encoding mode for the respective slice. Video encoder 20 typically operates on video blocks within individual video slices in order to encode the video data. A video block may correspond to a coding node within a CU. The video blocks may have fixed or varying sizes, and may differ in size according to a specified coding standard.
As an example, the HM supports prediction in various PU sizes. Assuming that the size of a particular CU is 2N×2N, the HM supports intra-prediction in PU sizes of 2N×2N or N×N, and inter-prediction in symmetric PU sizes of 2N×2N, 2N×N, N×2N, or N×N. The HIM also supports asymmetric partitioning for inter-prediction in PU sizes of 2N×nU, 2N×nD, nL×2N, and nR×2N. In asymmetric partitioning, one direction of a CU is not partitioned, while the other direction is partitioned into 25% and 75%. The portion of the CU corresponding to the 25% partition is indicated by an “n” followed by an indication of “Up,” “Down,” “Left,” or “Right.” Thus, for example, “2N×nU” refers to a 2N×2N CU that is partitioned horizontally with a 2N×0.5N PU on top and a 2N×1.5N PU on bottom.
In this disclosure, “N×N” and “N by N” may be used interchangeably to refer to the pixel dimensions of a video block in terms of vertical and horizontal dimensions, e.g., 16×16 pixels or 16 by 16 pixels. In general, a 16×16 block will have 16 pixels in a vertical direction (y=16) and 16 pixels in a horizontal direction (x=16). Likewise, an N×N block generally has N pixels in a vertical direction and N pixels in a horizontal direction, where N represents a nonnegative integer value. The pixels in a block may be arranged in rows and columns. Moreover, blocks may not necessarily have the same number of pixels in the horizontal direction as in the vertical direction. For example, blocks may comprise N×M pixels, where M is not necessarily equal to N.
Following intra-predictive or inter-predictive coding using the PUs of a CU, video encoder 20 may calculate residual data for the TUs of the CU. The PUs may comprise syntax data describing a method or mode of generating predictive pixel data in the spatial domain (also referred to as the pixel domain) and the TUs may comprise coefficients in the transform domain following application of a transform, e.g., a discrete cosine transform (DCT), an integer transform, a wavelet transform, or a conceptually similar transform to residual video data. The residual data may correspond to pixel differences between pixels of the unencoded picture and prediction values corresponding to the PUs. Video encoder 20 may form the TUs including the residual data for the CU, and then transform the TUs to produce transform coefficients for the CU.
Following any transforms to produce transform coefficients, video encoder 20 may perform quantization of the transform coefficients. Quantization is a broad term intended to have its broadest ordinary meaning. In one embodiment, quantization refers to a process in which transform coefficients are quantized to possibly reduce the amount of data used to represent the coefficients, providing further compression. The quantization process may reduce the bit depth associated with some or all of the coefficients. For example, an n-bit value may be rounded down to an m-bit value during quantization, where n is greater than m.
Following quantization, the video encoder may scan the transform coefficients, producing a one-dimensional vector from the two-dimensional matrix including the quantized transform coefficients. The scan may be designed to place higher energy (and therefore lower frequency) coefficients at the front of the array and to place lower energy (and therefore higher frequency) coefficients at the back of the array. In some examples, video encoder 20 may utilize a predefined scan order to scan the quantized transform coefficients to produce a serialized vector that can be entropy encoded. In other examples, video encoder 20 may perform an adaptive scan. After scanning the quantized transform coefficients to form a one-dimensional vector, video encoder 20 may entropy encode the one-dimensional vector, e.g., according to context-adaptive variable length coding (CAVLC), context-adaptive binary arithmetic coding (CABAC), syntax-based context-adaptive binary arithmetic coding (SBAC), Probability Interval Partitioning Entropy (PIPE) coding or another entropy encoding methodology. Video encoder 20 may also entropy encode syntax elements associated with the encoded video data for use by video decoder 30 in decoding the video data.
To perform CABAC, video encoder 20 may assign a context within a context model to a symbol to be transmitted. The context may relate to, for example, whether neighboring values of the symbol are non-zero or not. To perform CAVLC, video encoder 20 may select a variable length code for a symbol to be transmitted. Codewords in VLC may be constructed such that relatively shorter codes correspond to more probable symbols, while longer codes correspond to less probable symbols. In this way, the use of VLC may achieve a bit savings over, for example, using equal-length codewords for each symbol to be transmitted. The probability determination may be based on a context assigned to the symbol.
Video encoder 20 may further send syntax data, such as block-based syntax data, frame-based syntax data, and GOP-based syntax data, to video decoder 30 , e.g., in a frame header, a block header, a slice header, or a GOP header. The GOP syntax data may describe a number of frames in the respective GOP, and the frame syntax data may indicate an encoding/prediction mode used to encode the corresponding frame.
FIG. 2 is a block diagram illustrating an example of a video encoder that may implement techniques in accordance with aspects described in this disclosure. Video encoder 20 may be configured to perform any or all of the techniques of this disclosure. As one example, mode select unit 40 may be configured to perform any or all of the techniques described in this disclosure. However, aspects of this disclosure are not so limited. In some examples, the techniques described in this disclosure may be shared among the various components of video encoder 20 . In some examples, in addition to or instead of, a processor (not shown) may be configured to perform any or all of the techniques described in this disclosure.
In some embodiments, the mode select unit 40 , the motion estimation unit 42 , the motion compensation unit 44 , the inter-layer prediction unit 45 , the intra prediction unit 46 , the weighted prediction unit 47 , or another component of the mode select unit 40 (shown or not shown), or another component of the encoder 20 (shown or not shown) may perform the techniques of this disclosure. For example, the mode select unit 40 may receive video data for encoding, which may be encoded into a base layer and corresponding one or more enhancement layers. The mode select unit 40 , the motion estimation unit 42 , the motion compensation unit 44 , the inter-layer prediction unit 45 , the intra prediction unit 46 , the weighted prediction unit 47 , or another appropriate unit of the encoder 20 may determine a value of a current video unit based on, at least in part, a combination of two or more weighted predictions, including intra predictions, inter predictions, inter-layer predictions such as intraBL predictions, generalized residual predictions, etc. The encoder 20 can encode the video unit and signal the weighting factor(s) or weighting information in a bitstream.
Video encoder 20 may perform intra- and inter-coding of video blocks within video slices. Intra-coding relies on spatial prediction to reduce or remove spatial redundancy in video within a given video frame or picture. Inter-coding relies on temporal prediction to reduce or remove temporal redundancy in video within adjacent frames or pictures of a video sequence. Intra-mode (I mode) may refer to any of several spatial based coding modes. Inter-modes, such as uni-directional prediction (P mode) or bi-prediction (B mode), may refer to any of several temporal-based coding modes.
As shown in FIG. 2 , video encoder 20 receives a current video block within a video frame to be encoded. In the example of FIG. 1 , video encoder 20 includes mode select unit 40 , reference frame memory 64 , summer 50 , transform processing unit 52 , quantization unit 54 , and entropy coding unit 56 . Mode select unit 40 , in turn, includes motion compensation unit 44 , motion estimation unit 42 , inter-layer prediction unit 45 , intra-prediction unit 46 , weighted prediction unit 47 , and partition unit 48 . For video block reconstruction, video encoder 20 also includes inverse quantization unit 58 , inverse transform unit 60 , and summer 62 . A deblocking filter (not shown in FIG. 2 ) may also be included to filter block boundaries to remove blockiness artifacts from reconstructed video. If desired, the deblocking filter would typically filter the output of summer 62 . Additional filters (in loop or post loop) may also be used in addition to the deblocking filter. Such filters are not shown for brevity, but if desired, may filter the output of summer 62 (as an in-loop filter).
During the encoding process, video encoder 20 receives a video frame or slice to be coded. The frame or slice may be divided into multiple video blocks. Motion estimation unit 42 and motion compensation unit 44 perform inter-predictive coding of the received video block relative to one or more blocks in one or more reference frames to provide temporal prediction. Inter-layer prediction unit 45 may alternatively perform inter-layer predictive coding of the received video block relative to one or more blocks of a different layer of video information, such as a co-located block of a base layer, as described in greater detail below. Intra-prediction unit 46 may alternatively perform intra-predictive coding of the received video block relative to one or more neighboring blocks in the same frame or slice as the block to be coded to provide spatial prediction. Video encoder 20 may perform multiple coding passes, e.g., to select an appropriate coding mode for each block of video data. Weighted prediction unit 47 may alternatively perform weighted prediction using a combination of predictions made according to various prediction modes, as described in greater detail below.
Moreover, partition unit 48 may partition blocks of video data into sub-blocks, based on evaluation of previous partitioning schemes in previous coding passes. For example, partition unit 48 may initially partition a frame or slice into LCUs, and partition each of the LCUs into sub-CUs based on rate-distortion analysis (e.g., rate-distortion optimization). Mode select unit 40 may further produce a quadtree data structure indicative of partitioning of an LCU into sub-CUs. Leaf-node CUs of the quadtree may include one or more PUs and one or more TUs.
Mode select unit 40 may select one of the coding modes, intra or inter, e.g., based on error results, and provides the resulting intra- or inter-coded block to summer 50 to generate residual block data and to summer 62 to reconstruct the encoded block for use as a reference frame. Mode select unit 40 also provides syntax elements, such as motion vectors, intra-mode indicators, partition information, and other such syntax information, to entropy coding unit 56 .
Motion estimation unit 42 and motion compensation unit 44 may be highly integrated, but are illustrated separately for conceptual purposes. Motion estimation, performed by motion estimation unit 42 , is the process of generating motion vectors, which estimate motion for video blocks. A motion vector, for example, may indicate the displacement of a PU of a video block within a current video frame or picture relative to a predictive block within a reference frame (or other coded unit) relative to the current block being coded within the current frame (or other coded unit). A predictive block is a block that is found to closely match the block to be coded, in terms of pixel difference, which may be determined by sum of absolute difference (SAD), sum of square difference (SSD), or other difference metrics. In some examples, video encoder 20 may calculate values for sub-integer pixel positions of reference pictures stored in reference frame memory 64 . For example, video encoder 20 may interpolate values of one-quarter pixel positions, one-eighth pixel positions, or other fractional pixel positions of the reference picture. Therefore, motion estimation unit 42 may perform a motion search relative to the full pixel positions and fractional pixel positions and output a motion vector with fractional pixel precision.
Motion estimation unit 42 calculates a motion vector for a PU of a video block in an inter-coded slice by comparing the position of the PU to the position of a predictive block of a reference picture. The reference picture may be selected from a first reference picture list (List 0) or a second reference picture list (List 1), each of which identify one or more reference pictures stored in reference frame memory 64 . Motion estimation unit 42 sends the calculated motion vector to entropy encoding unit 56 and motion compensation unit 44 .
Motion compensation, performed by motion compensation unit 44 , may involve fetching or generating the predictive block based on the motion vector determined by motion estimation unit 42 . Again, motion estimation unit 42 and motion compensation unit 44 may be functionally integrated, in some examples. Upon receiving the motion vector for the PU of the current video block, motion compensation unit 44 may locate the predictive block to which the motion vector points in one of the reference picture lists. Summer 50 forms a residual video block by subtracting pixel values of the predictive block from the pixel values of the current video block being coded, forming pixel difference values, as discussed below. In general, motion estimation unit 42 performs motion estimation relative to luma components, and motion compensation unit 44 uses motion vectors calculated based on the luma components for both chroma components and luma components. Mode select unit 40 may also generate syntax elements associated with the video blocks and the video slice for use by video decoder 30 in decoding the video blocks of the video slice.
Intra-prediction unit 46 may intra-predict or calculate a current block, as an alternative to the inter-prediction performed by motion estimation unit 42 and motion compensation unit 44 , as described above. In particular, intra-prediction unit 46 may determine an intra-prediction mode to use to encode a current block. In some examples, intra-prediction unit 46 may encode a current block using various intra-prediction modes, e.g., during separate encoding passes, and intra-prediction unit 46 (or mode select unit 40 , in some examples) may select an appropriate intra-prediction mode to use from the tested modes.
For example, intra-prediction unit 46 may calculate rate-distortion values using a rate-distortion analysis for the various tested intra-prediction modes, and select the intra-prediction mode having the best rate-distortion characteristics among the tested modes. Rate-distortion analysis generally determines an amount of distortion (or error) between an encoded block and an original, unencoded block that was encoded to produce the encoded block, as well as a bitrate (that is, a number of bits) used to produce the encoded block. Intra-prediction unit 46 may calculate ratios from the distortions and rates for the various encoded blocks to determine which intra-prediction mode exhibits the best rate-distortion value for the block.
After selecting an intra-prediction mode for a block, intra-prediction unit 46 may provide information indicative of the selected intra-prediction mode for the block to entropy coding unit 56 . Entropy coding unit 56 may encode the information indicating the selected intra-prediction mode. Video encoder 20 may include in the transmitted bitstream configuration data, which may include a plurality of intra-prediction mode index tables and a plurality of modified intra-prediction mode index tables (also referred to as codeword mapping tables), definitions of encoding contexts for various blocks, and indications of a most probable intra-prediction mode, an intra-prediction mode index table, and a modified intra-prediction mode index table to use for each of the contexts.
Video encoder 20 forms a residual video block by subtracting the prediction data from mode select unit 40 from the original video block being coded. Summer 50 represents the component or components that perform this subtraction operation. Transform processing unit 52 applies a transform, such as a discrete cosine transform (DCT) or a conceptually similar transform, to the residual block, producing a video block comprising residual transform coefficient values. Transform processing unit 52 may perform other transforms which are conceptually similar to DCT. Wavelet transforms, integer transforms, sub-band transforms or other types of transforms could also be used. In any case, transform processing unit 52 applies the transform to the residual block, producing a block of residual transform coefficients. The transform may convert the residual information from a pixel value domain to a transform domain, such as a frequency domain. Transform processing unit 52 may send the resulting transform coefficients to quantization unit 54 . Quantization unit 54 quantizes the transform coefficients to further reduce bit rate. The quantization process may reduce the bit depth associated with some or all of the coefficients. The degree of quantization may be modified by adjusting a quantization parameter. In some examples, quantization unit 54 may then perform a scan of the matrix including the quantized transform coefficients. Alternatively, entropy encoding unit 56 may perform the scan.
Following quantization, entropy coding unit 56 entropy codes the quantized transform coefficients. For example, entropy coding unit 56 may perform context adaptive variable length coding (CAVLC), context adaptive binary arithmetic coding (CABAC), syntax-based context-adaptive binary arithmetic coding (SBAC), probability interval partitioning entropy (PIPE) coding or another entropy coding technique. In the case of context-based entropy coding, context may be based on neighboring blocks. Following the entropy coding by entropy coding unit 56 , the encoded bitstream may be transmitted to another device (e.g., video decoder 30 ) or archived for later transmission or retrieval.
Inverse quantization unit 58 and inverse transform unit 60 apply inverse quantization and inverse transformation, respectively, to reconstruct the residual block in the pixel domain, e.g., for later use as a reference block. Motion compensation unit 44 may calculate a reference block by adding the residual block to a predictive block of one of the frames of reference frame memory 64 . Motion compensation unit 44 may also apply one or more interpolation filters to the reconstructed residual block to calculate sub-integer pixel values for use in motion estimation. Summer 62 adds the reconstructed residual block to the motion compensated prediction block produced by motion compensation unit 44 to produce a reconstructed video block for storage in reference frame memory 64 . The reconstructed video block may be used by motion estimation unit 42 and motion compensation unit 44 as a reference block to inter-code a block in a subsequent video frame.
FIG. 3 is a block diagram illustrating an example of a video decoder that may implement techniques in accordance with aspects described in this disclosure. Video decoder 30 may be configured to perform any or all of the techniques of this disclosure. As one example, motion compensation unit 72 , inter-layer prediction unit 73 , intra prediction unit 74 , and/or weighted prediction unit 75 may be configured to perform any or all of the techniques described in this disclosure. However, aspects of this disclosure are not so limited. In some examples, the techniques described in this disclosure may be shared among the various components of video decoder 30 . In some examples, in addition to or instead of, a processor (not shown) may be configured to perform any or all of the techniques described in this disclosure.
In some embodiments, the entropy decoding unit 70 , the motion compensation unit 72 , the inter-layer prediction unit 73 , the intra prediction unit 74 , the weighted prediction unit 75 , or another component of the decoder 30 (shown or not shown) may perform the techniques of this disclosure. For example, the entropy decoding unit 70 may receive an encoded video bitstream, which may encode data relating to a base layer and corresponding one or more enhancement layers. The motion compensation unit 72 , the inter-layer prediction unit 73 , the intra prediction unit 74 , the weighted prediction unit 75 , or another appropriate unit of the decoder 30 may determine a value of a current video unit based on, at least in part, a combination of two or more weighted predictions, including intra predictions, inter predictions, intraBL predictions, generalized residual predictions, etc. The decoder 30 can decode the video unit and receive the weighting factor(s) or weighting information in a bitstream.
In the example of FIG. 3 , video decoder 30 includes an entropy decoding unit 70 , motion compensation unit 72 , inter-layer prediction unit 73 , intra prediction unit 74 , weighted prediction unit 75 , inverse quantization unit 76 , inverse transformation unit 78 , reference frame memory 82 and summer 80 . Video decoder 30 may, in some examples, perform a decoding pass generally reciprocal to the encoding pass described with respect to video encoder 20 ( FIG. 2 ). Motion compensation unit 72 may generate prediction data based on motion vectors received from entropy decoding unit 70 , while intra-prediction unit 74 may generate prediction data based on intra-prediction mode indicators received from entropy decoding unit 70 .
During the decoding process, video decoder 30 receives an encoded video bitstream that represents video blocks of an encoded video slice and associated syntax elements from video encoder 20 . Entropy decoding unit 70 of video decoder 30 entropy decodes the bitstream to generate quantized coefficients, motion vectors or intra-prediction mode indicators, and other syntax elements. Entropy decoding unit 70 forwards the motion vectors to and other syntax elements to motion compensation unit 72 . Video decoder 30 may receive the syntax elements at the video slice level and/or the video block level.
When the video slice is coded as an intra-coded (I) slice, intra prediction unit 74 may generate prediction data for a video block of the current video slice based on a signaled intra prediction mode and data from previously decoded blocks of the current frame or picture. When the video frame is coded as an inter-coded (e.g., B, P or GPB) slice, motion compensation unit 72 produces predictive blocks for a video block of the current video slice based on the motion vectors and other syntax elements received from entropy decoding unit 70 . The predictive blocks may be produced from one of the reference pictures within one of the reference picture lists. Video decoder 30 may construct the reference frame lists, List 0 and List 1, using default construction techniques based on reference pictures stored in reference frame memory 92 . Motion compensation unit 72 determines prediction information for a video block of the current video slice by parsing the motion vectors and other syntax elements, and uses the prediction information to produce the predictive blocks for the current video block being decoded. For example, motion compensation unit 72 uses some of the received syntax elements to determine a prediction mode (e.g., intra- or inter-prediction) used to code the video blocks of the video slice, an inter-prediction slice type (e.g., B slice, P slice, or GPB slice), construction information for one or more of the reference picture lists for the slice, motion vectors for each inter-encoded video block of the slice, inter-prediction status for each inter-coded video block of the slice, and other information to decode the video blocks in the current video slice.
Motion compensation unit 72 may also perform interpolation based on interpolation filters. Motion compensation unit 72 may use interpolation filters as used by video encoder 20 during encoding of the video blocks to calculate interpolated values for sub-integer pixels of reference blocks. In this case, motion compensation unit 72 may determine the interpolation filters used by video encoder 20 from the received syntax elements and use the interpolation filters to produce predictive blocks.
Inverse quantization unit 76 inverse quantizes, e.g., de-quantizes, the quantized transform coefficients provided in the bitstream and decoded by entropy decoding unit 80 . The inverse quantization process may include use of a quantization parameter QPy calculated by video decoder 30 for each video block in the video slice to determine a degree of quantization and, likewise, a degree of inverse quantization that should be applied.
Inverse transform unit 78 applies an inverse transform, e.g., an inverse DCT, an inverse integer transform, or a conceptually similar inverse transform process, to the transform coefficients in order to produce residual blocks in the pixel domain.
After motion compensation unit 82 generates the predictive block for the current video block based on the motion vectors and other syntax elements, video decoder 30 forms a decoded video block by summing the residual blocks from inverse transform unit 78 with the corresponding predictive blocks generated by motion compensation unit 72 . Summer 90 represents the component or components that perform this summation operation. If desired, a deblocking filter may also be applied to filter the decoded blocks in order to remove blockiness artifacts. Other loop filters (either in the coding loop or after the coding loop) may also be used to smooth pixel transitions, or otherwise improve the video quality. The decoded video blocks in a given frame or picture are then stored in reference picture memory 92 , which stores reference pictures used for subsequent motion compensation. Reference frame memory 82 also stores decoded video for later presentation on a display device, such as display device 32 of FIG. 1 .
The example encoder 20 and decoder 30 described above are merely illustrative, and are not intended to be limiting. In some embodiments, an encoder or decoder configured to perform the techniques and provide the functionality described herein may include additional or fewer components. In some embodiments, other components or devices may provide the functionality described herein, such as a general purpose computer processor configured to execute specific instructions stored in a computer-readable medium.
As described above, weighted mode predictions may include combinations of individual predictions generated according to various prediction modes. For example, a single weighted mode prediction may include intra predictions and/or inter predictions. SVC-specific prediction modes may also be used to provide individual predictions for weighted mode. For example, predictions made according to intraBL mode, difference domain intra mode, generalized residual prediction, and the like may be combined with each other and/or with inter predictions, intra predictions, etc. These and other prediction modes are described in greater detail below. The aforementioned prediction modes are illustrative only, and are not intended to be limiting. In general, predictions made according to any prediction mode or technique may be combined in the weighted mode described herein.
FIG. 4A shows illustrative predictions for a particular video unit, including both intra-layer and inter-layer predictions. A block or other video unit of an enhancement layer (EL) frame 140 may be predicted from various sources, both in the same frame and in different frames, and both in the same layer or in a different layer. As shown in FIG. 4A , current block 142 may be predicted at (A) from another block 144 in the same frame 140 and in the same layer (EL). Current block 142 may also be predicted at (B) from a corresponding block 132 in a reference frame 130 in the same layer. Current block 142 may also be predicted at (C) from a corresponding block 122 in a frame 120 of a different layer—the base layer (BL). In cases where the EL has a different, for example larger resolution, that the BL (e.g., the EL provides spatial enchantment), upsampling of a BL reconstructed block can be applied to match the resolution of the EL prior to predicting a corresponding EL block.
Some examples of inter-layer prediction may include inter-layer intra prediction, inter-layer motion prediction, and inter-layer residual prediction. Inter-layer intra prediction, or IntraBL, uses the reconstruction of co-located blocks in the base layer to predict the current block in the enhancement layer. Inter-layer motion prediction uses motion of the base layer to predict motion in the enhancement layer. Inter-layer residual prediction uses the residue of the base layer to predict the residue of the enhancement layer.
In inter-layer residual prediction, the residue of the base layer may be used to predict the current block in the enhancement layer. The residue may be defined as the difference between the temporal prediction for a video unit and the source video unit. In some embodiments, the current EL block may be reconstructed using the residue from the enhancement layer, the temporal prediction from the enhancement layer, and the residue from the base layer. The current block may be reconstructed according to the following equation:

Î e =r e +P e +r b (1)

where Î e (or Î c ) denotes the reconstruction of the current block, r e denotes the residue from the enhancement layer, P e (or P c ) denotes the temporal prediction from the enhancement layer, and r b denotes the residue prediction from the base layer.

In order to use inter-layer residual prediction for a block in the enhancement layer, the co-located block in the base layer should be an inter-coded block, and the residue of the co-located base layer block may be upsampled according to the spatial resolution ratio of the enhancement layer (e.g., because the layers in SVC may have different spatial resolutions). In inter-layer residual prediction, the difference between the residue of the enhancement layer and the residue of the upsampled base layer may be coded in the bitstream. The residue of the base layer may be normalized based on the ratio between quantization steps of base and enhancement layers. Inter-layer residual prediction is described in greater detail below.
To add more diversity to a final prediction, predictors from prediction modes described above, or other prediction modes, can be weighted and combined to generate a weighted mode predictor. FIG. 4B shows an illustrative prediction for the current block 142 generated in weighted mode. The intra prediction from block 144 in the same layer may be weighted and combined with the weighted intraBL prediction from block 122 in the base layer. In another example, the inter-prediction from block 132 in the same layer may be weighted and combined with the weighted intraBL prediction from block 122 in the base layer. In one more example, all three mentioned predictions (intra-, inter-, and inter-layer) can be weighted and combined. As a result, the reconstruction of block 142 may be based on a diverse prediction determined from several individual predictions. This can help avoid the errors and other limitations inherent in individual prediction modes.
Generally described, a weighted mode predictor may be determined as follows:
WMpredictor = 1 Norm · ∑ i ⁢ w i · Predictor i ( 2 )
where Predictor i is a predictor from some prediction mode, w i is a weight corresponding to the prediction mode, and Norm is a normalization factor. Summation is performed over a number i of prediction modes. In some embodiments, the number i of prediction modes may be more than 2.

The weights and normalization factor in equation (2) can be any number (e.g., floating point numbers or integers), and may be different for different color components. In one specific, non-limiting embodiment, w 1 =1, w 2 =1 and Norm=2 for the luma component, while w 1 =0, w 2 =1 and Norm=1 for the chroma components. These weights are illustrative only. In some embodiments, the weights may be the same for each color component of the video unit being predicted.
In some embodiments, an additional constraint may be imposed on equation (2) such that
Norm = ∑ i ⁢ w i ( 3 )
As seen in equation (3), the normalization factor Norm may equal the sum of the weighting factors. When such a constraint is imposed on Norm, equation (2) can produce a weighted average of multiple predictors.

FIG. 5 shows an example method or process 500 for determining a value of a current video unit, such as a pixel or block of pixels, using weighted mode. For example, a digital video coder, such as the encoder 20 or decoder 30 described above, may use the process 500 to generate predictions in weighted mode. In some embodiments, the process 500 may be performed by some other component or apparatus. Illustratively, the description which follows describes the steps of the process 500 with respect to the weighted prediction unit 75 and other components of the decoder 30 in FIG. 3 .
The process 500 begins at block 502 . At block 504 , video information is stored. For example, video information associated with a base layer, an enhancement layer, an upsampled layer, a downsampled layer, a reference layer, or any combination, may be stored at block 504 .
At block 506 , a value of a current video unit (such as a pixel or block of pixels) is determined (or predicted) using a weighted prediction generated according to an individual prediction mode. Weighted mode can be applied to all color components of a video unit, or some subset thereof. For example, the weighted mode can be applied to luma and both chromas, or alternatively it can be applied selectively to one or more components, such as only to the luma component. In weighted mode, the value of the video unit may be determined using one or more weighted predictions of at least one color component of the video unit. For example, the value may be determined by selectively applying a weighting factor to one or more color components of the video information, depending upon the individual prediction mode being utilized.
Returning to the example in FIG. 4B , the weighted mode prediction for EL block 142 is to be based on predictions made using both intra prediction and intraBL prediction. First, an intra prediction may be made for a video unit in EL block 142 using a value from EL block 144 and a weighting factor. The weighting factor may be determined based on any number of factors. In some embodiments, the weighting factors for each prediction used in the final weighted mode prediction may be determined such that each prediction (e.g., the intra prediction and intraBL prediction in this case) are weighted evenly. In some embodiments, the weighting factors may be based at least partly on the location of the video unit to be predicted. For video units such as sub-blocks or pixels that are close to previously reconstructed video units or blocks in the same layer and frame, a higher weighting factor may be used for the intra prediction than the intraBL prediction. In the example of FIG. 4B , intra predictions for pixels or other video units that are close to the top of EL block 142 , and therefore close to the previously-reconstructed EL block 144 , may be given a weight close to or equal to 1. Intra predictions for video units that are farther from the top of EL block 142 may be given a lower weight, such as a weight that approaches or equals zero for video units at the bottom of EL block 142 . This example is illustrative only, and assumes many details, such as a vertical prediction direction, etc. Other factors and techniques for determining weighting factors may be used, as will be appreciated by one of skill in the art.
At 508 , the process 500 determines whether there are any additional weighting factors to apply to predictions made according to individual prediction modes. If so, the process 500 returns to 506 . Returning to the example of FIG. 4B , an intraBL prediction for a given video unit in EL block 142 may be a weighted prediction from a collocated or otherwise corresponding block 122 in the base layer. For video units that are close the top of EL block 142 , the weighting factor may be close to or equal to zero for the intraBL prediction. The weighting factor for the intraBL prediction may approach or equal 1 for video units that are closer or at the bottom of EL block 142 . The value of the current video unit may be determined as the sum of all the weighted predictions, or as the sum of a selected group of the weighted predictors.
If there are no additional weighting factors at block 508 , the process may terminate at block 510 .
Weighted mode may be expressly signaled in a coded bitstream as an additional flag (e.g., weighted_mode_flag=0 or 1) at least at the block, PU, CU, LCU, slice, frame, sequence level or elsewhere. For example, an encoder 20 implementing the process 500 above may first determine that weighted mode is to be used, and then signal the weighted mode flag in the bitstream so that a decoder may know that the video information for the particular block, PU, CU, LCU, slice, frame or sequence level which follows is encoded in weighted mode. A decoder 30 implementing the process 500 may detect the weighted mode flag and then use the process 500 to decode at least a portion of the bitstream. Alternatively, a determination of whether to use weighted mode can be derived from BL information, EL information, some combination thereof, etc.
In some embodiments, if a weighted mode flag is signaled, the weights can also be signaled for each block, PU, CU, LCU, slice, frame, sequence level or elsewhere. Alternatively, as described above, the weights used in weighted mode can be fixed (e.g. w i =1, or 0.5 when two predictions are combined), derived from the previously-reconstructed neighbors, or determined based upon the distance from the block boundary. For example, for the intra weighted mode, bigger weights can be given to the intra prediction mode for the pixels closed to the block boundary, since intra prediction may have better correlation to the immediate neighbor pixels.
In some embodiments, weights can be organized in a list with one or more entries, where one or more weights are assigned to each weight index of the list. The weight index can be signaled in the bitstream. This list may be fixed or otherwise known to both the encoder and decoder. In other embodiments, the list is implicitly or explicitly derived. For example, the weight list can be determined based on the weight's hit ratio obtained from previous coded blocks. Alternatively, or additionally, the list can be different for different prediction modes of the weighted mode. For example, one list can be used for intra predicted blocks, and another list can be used for inter predicted blocks. These lists can be updated separately, e.g., based on statistics from previously-processed blocks. The usage of this mode can be indicated by a syntax element in one or more headers, such as slice, picture, sequence headers or video parameter set.
In one specific non-limiting embodiment, Predictor 1 in equation (2) may be a reconstructed (e.g., upsampled in spatial scalability case) BL block or an intraBL predictor. A conventional intra or inter mode predictor may be used as a Predictor 2 . The weights w 1 =1 and w 2 =1, and Norm=2, may be used for the luma component, and w 1 =0, w 2 =1, and Norm=1 may be used for both chroma components.
When Predictor 2 in the example above is a predictor generated according to intra prediction mode, all intra directions, or some subset thereof, may be used to predict the luma component and/or a chroma component. In some embodiments, in order to reduce intra direction overhead, intra DC or DM prediction can be used for both chroma components. As a result, the chroma intra direction(s) do not need to be signaled. This may reduce either or both of overhead signaling and encoder complexity, because searching for a best intra direction is not necessary.
Different syntax structures can be used depending on the weighted mode application. For example, a “weighted_mode” flag can be signaled per CU, PU, etc. The flag may be signaled before luma and chroma intra prediction directions if the weighting is applied for various upsampling of the base block. In such cases, signaling of luma and/or chroma inter directions may be suppressed.
As yet another example, different upsampling filters can be applied to the BL block, and the weighted mode can be applied to the result of difference upsamplers. Other mode combinations, weights, and normalization factors can be used, as will be appreciated by one of skill in the art.
In some embodiments, one context model is used to code a weighted mode flag for all blocks. In other embodiments, the context model can be dependent on which individual prediction modes are used with weighted mode. For example, one context model can be assigned to the intra predicted blocks and another context model can be assigned to the inter predicted blocks.
In additional embodiments, the context model can be separated according to which inter prediction mode (e.g., uni-prediction or bi-prediction) is used. For example, uni-predicted blocks can use one context model, and bi-predicted blocks can use another context model. In further embodiments, the context model may be dependent upon other factors, such as intra directions, merge flag, skip flag, CU depth or slice types. The weighted mode context model cost can be used to determine the best CABAC initialization table from existing initialization tables used for EL context model initialization.
In certain situations, the bandwidth requirements and/or coding complexity introduced by the use of weighted mode may degrade performance or otherwise produce unsatisfactory results. For example, weighted mode using bi-directional sub-pixel (sub-pel) motion compensation for the enhancement layer and upsampling for the base layer can result in a new “worst case scenario” when compared to the HEVC single layer coding techniques. In HEVC single layer coding, the “worst case scenario” involves only bi-directional prediction. To reduce bandwidth requirements and/or coding complexity for SVC using weighted mode, the techniques and processes described below with respect to FIGS. 6 and 7 may be applied.
FIG. 6 shows an example process 600 for limiting the use of weighted mode to certain block sizes. Advantageously, this can reduce the bandwidth required to signal weighted-mode-specific information. The process 600 may be implemented by any number of components or devices, such as the weighted prediction unit 47 of the encoder 20 , the weighted prediction unit 75 of the decoder 30 , or some other component of the encoder 20 and/or decoder 30 described above.
The process 600 begins at block 602 . At 604 , the size of a current EL block to be reconstructed is determined. For example, block sizes may range from 4×4 to 32×32, including various sizes in between.
At decision block 606 , the component or device implementing the process 600 can determine whether the size, determined above, exceeds a threshold. If so, the process 600 proceeds to 608 , wherein the value of at least one color component of a video unit in the current EL block is determined by selectively applying a weighting factor to a prediction, as described above with respect to FIG. 5 . For example, blocks larger then 8×8 or larger than 16×16 may be coded (e.g., encoded or decoded) using the weighted prediction mode.
For blocks that fail to meet or exceed the threshold, as determined above at 606 , the process 600 proceeds to 610 . At 610 , the values of the video units in the current block are generated using individual prediction modes (e.g., without summing weighted predictions made according to multiple prediction modes).
In some embodiments, a threshold size may not necessarily be used. Instead, a list or group of block sizes with which weighted mode may be used can be maintained. In some embodiments, a minimum block size and partition mode can be defined for which a weighted mode may be used. In one specific non-limiting example, weighted mode can be used only for CU sizes greater than or equal to 8×8 and having a partition mode equal to 2N×2N. For such blocks, the process 600 proceeds to 608 , described above. For an 8×8 CU having a partition mode other than 2N×2N, weighted mode can be prohibited, restricted, or suppressed (e.g., a weighted mode flag is not signaled for the CU by the encoder 20 , or a weighted mode flag is ignored by the decoder 30 ), and the process 600 proceeds to 610 .
As yet another embodiment, weighted mode may be used only with certain individual prediction modes. For example weighted mode may be suppressed (and no weighted mode flag signaled, or a signaled weighted mode flag is ignored) when using intra residual prediction mode, inter skip mode, etc. As another example, weighted mode may be used only with intra predictions made using certain intra directions. As a further example, certain types of motion compensation may be suppressed, restricted, or limited when weighted mode is used, or weighted mode may be suppressed when certain types of motion compensation are used.
FIG. 7 shows an example process 700 for limiting, restricting, or suppressing the use of weighted mode with certain types of motion compensation prediction modes. Advantageously, this can reduce the bandwidth required to signal weighted-mode-specific information when such modes are used. This can also reduce the complexity of determining video unit values using weighted mode. The process 700 may be implemented by any number of components or devices, such as the weighted prediction unit 47 of the encoder 20 , the weighted prediction unit 75 of the decoder 30 , or some other component of the encoder 20 and/or decoder 30 described above.
The process 700 begins at block 702 . At block 704 , the component or device executing the process 700 may determine that motion compensation is to be used for the current block. At decision block 706 , the component or device executing to the process 700 may determine whether the motion compensation is bi-directional, or if bi-prediction is otherwise to be used. For example, in merge mode, the motion vector candidate used for inter-prediction may be bi-directional. If so, the process 700 proceeds to 708 . Otherwise, the process 700 proceeds to 710 .
At 708 , the motion compensation for the current video unit or block may be converted to uni-directional prediction in order to conserve bandwidth (e.g., at the encoder 20 ) or to reduce coding complexity (e.g., at the decoder 30 ). In one specific, non-limiting embodiment, an inter prediction using either a uni-L0-direction or uni-L1-direction, but not both, is used in combination with an intraBL prediction for weighted mode. The bi-directional motion vector is converted to either a uni-L0 or urn-L1 motion vector before or during motion compensation (e.g., a single motion vector is used from either the L0 or L1 list). Alternatively, the bi-directional motion vector candidate can be converted to uni-directional (either uni-L0 or uni-L1) motion vector during merge list construction.
In some embodiments, bi-prediction can be suppressed or converted to urn-direction prediction only for luma components, while a chroma component may be bi-predicted. In some embodiments (e.g., for weighted mode based on AMVP mode), the bi-directional motion vector is suppressed at the encoder and only a uni-directional motion vector is coded in the bitstream. In additional embodiments, inter direction signaling can be restricted such that the bitstream contains only uni-L0 or uni-L1 directions. As a result, any redundant bit associated with bi-direction is not included in the bitstream. In further embodiments, only integer pel motion vectors are used. As a result, bandwidth can be reduced because there is no need to do an interpolation, even for bi-directional prediction.
In some embodiments, bi-prediction is restricted in weighted mode by applying a normative bitstream constraint to avoid changes in the syntax elements signaling. Coding may be performed with the restriction that bi-prediction (e.g. all the syntax associated with the bi-prediction) cannot be signaled in a bitstream if weighted mode is used. For example, a bi-directional motion vector, bi-directional inter direction and two reference indices (for list L0 and list L0) cannot be present in a bitstream if weighted mode is enabled (e.g., when weighted_mode_flag is equal to 1). Rather, only urn-directional motion information (a urn-directional motion vector, a uni-directional inter direction, and only one reference index) can be present in the bitstream. This restriction implies that associated syntax such as merge_index, MVPidx, MVD and the like also conform to the constraint and do not lead to bi-prediction.
The block size and motion prediction limitations and suppression techniques described above with respect to FIGS. 6-7 may be combined. Weighted mode may utilize bi-direction prediction only for particular block sizes and partition modes. As one specific, non-limiting example, bi-direction prediction with weighted mode can be used, for example, for CU size starting from 16×16 and partition mode 2N×2N. For 16×16 CU and partition modes other than 2N×2N, only uni-directional prediction may be used for the weighted mode, as described above. In addition, merge mode motion vector candidates can be converted to urn-directional candidates, and inter direction signaling for AMVP mode can be restricted to urn-direction for those blocks, as also described above. In some embodiments, for certain CUs (e.g., having a size of 16×16 with a non-2N×2N partition mode) bi-directional prediction may be blocked or suppressed for only certain prediction units in a given CU (e.g., only for the first prediction unit), and bi-prediction can be used for the others (e.g., for the second PU). Although the forgoing examples are described in the context of 8×8 and 16×16 CU block sizes, the techniques may be readily applied to blocks having any other size.
In some embodiments, weighted mode may be suppressed or restricted without syntax modifications by changing the behavior of the weighted mode prediction according to the block size or inter direction. For example, a bi-directional motion vector can be signaled to decoder and, if bi-prediction is restricted for weighted mode to reduce complexity, then normal inter prediction is performed instead of weighted mode even if weighted mode is signaled (e.g., “weighted_mode_flag” is equal to 1). If a uni-directional motion vector is present and weighted mode is signaled, then weighted mode prediction is performed.
To keep “worst case” complexity within the current “worst case” for HEVC single layer coding, conditional signaling of weighted mode may be used. For example, weighted mode may be suppressed or restricted for 8×4 and 4×8 PUs. However, weighted mode may still be signaled for all CU sizes. When weighted mode is signaled (e.g., “weighted_mode_flag” is equal to 1), a check of the PU size may be performed at the prediction stage. If the PU size is 4×8 or 8×4, then normal inter prediction is performed (e.g., uni-directional or bi-directional prediction) in lieu of weighted mode prediction, even though weighted mode was signaled. Only if a uni-directional motion vector is signaled and the PU block size is not 8×4 or 4×8 will weighted prediction be performed. Motion vector signaling can be done either through merge or AMVP modes. In some embodiments, weighted mode may be similarly restricted for blocks sizes other than 8×4 or 4×8. Additionally, restriction of weighted mode may be based on the partition mode, or on whether merge or AMVP mode signaling is used for motion information with the weighted mode.
Any of the foregoing techniques may be employed alone or in combination to reduce bandwidth and/or coding complexity.
Turning now to FIGS. 8-15 , generalized residual prediction (GRP) mode will be described in greater detail. As described above, weighted mode may use any type of prediction, including GRP. The GRP techniques described below may address issues relating to inter-layer residual prediction and difference domain motion compensation in SVC. In GRP, weights are used for various components of a generalized residual prediction, as described below. In an inter predicted block, weighted mode can be used for all GRP weights. In some embodiments, weighted mode may be applied to only certain GRP weights. For example, weighted mode may only be used only when the GRP weights equal 2. In this case, the weighted mode flag or other weighted mode information should be signaled after the GRP weights to exploit the mode dependency. In this example, weighted mode flag signaling can be skipped or suppressed for certain GRP weights, for example when GRP weights≠2, thereby reducing bandwidth and/or storage requirements. In another embodiment, weighted mode is used with certain GRP weights, for example when GRP weights=1 or 2, and therefore it would be unnecessary to signal weighted mode flags or information for GRP weights which are not used in combined mode, for example when GRP weights=zero. Other combinations of GRP weights and weighted mode suppress may be used.
The SVC extension to H.264 requires single-loop decoding for motion compensation in order to maintain low complexity for the decoder. In general, motion compensation is performed by adding the temporal prediction and the residue for the current block as follows:

Î=r+P (4)

where Î denotes the current frame, r denotes the residue, and P denotes the temporal prediction. In single-loop decoding, each supported layer in SVC can be decoded with a single motion compensation loop. In order to achieve this, all layers that are used to inter-layer intra predict higher layers are coded using constrained intra-prediction. In constrained intra prediction, intra mode blocks are intra-coded without referring to any samples from neighboring inter-coded blocks. On the other hand, HEVC allows multi-loop decoding for SVC, in which an SVC layer may be decoded using multiple motion compensation loops. For example, the base layer is fully decoded first, and then the enhancement layer is decoded.

Residual prediction formulated in equation (1), described above, may be an efficient technique in H.264 SVC extension. However, its performance can be further improved in the HEVC SVC extension, especially when multi-loop decoding is used in the HEVC SVC extension.
In the case of multi-loop decoding, difference domain motion compensation may be used in place of residual prediction. In SVC, an enhancement layer may be coded using pixel domain coding or difference domain coding. In pixel domain coding, the input pixels for an enhancement layer pixels may be coded, as for a non-SVC HEVC layer. On the other hand, in difference domain coding, difference values for an enhancement layer may be coded. The difference values may be the difference between the input pixels for the enhancement layer and the corresponding scaled base layer reconstructed pixels. Such difference values may be used in motion compensation for difference domain motion compensation.
For inter coding using difference domain, the current predicted block is determined based on the difference values between the corresponding predicted block samples in the enhancement layer reference picture and the corresponding predicted block samples in the scaled base layer reference picture. The difference values may be referred to as the difference predicted block. The co-located base layer reconstructed samples are added to the difference predicted block in order to obtain enhancement layer reconstructed samples.
However, using difference domain motion compensation in inter-layer prediction introduces two sets of motion estimation and motion compensation, since motion estimation and motion compensation are often used for both pixel domain and difference domain. Introducing two sets of motion estimation and motion compensation can lead to higher buffer and computational cost, which may not be practical for an encoder or a decoder. In addition, coding two sets of motion vectors may reduce coding efficiency since motion field may become irregular when the two sets of motion vectors have different properties and are interleaved at coding unit (CU) level. Moreover, motion estimation in difference domain requires that the base layer and enhancement layer share the same motion. Further, difference domain motion compensation does not work with single-loop decoding since the derivation of differential pictures between two layers is based on fully reconstructed pictures of each layer. Accordingly, it would be advantageous to avoid redundancy in having two sets of motion estimation and motion compensation when using difference domain motion compensation. Also, it would be advantageous to extend difference domain motion compensation in single-loop decoding.
As explained above, inter-layer residual prediction uses the residue of the reference layer in predicting the current video unit, for example, a block or a frame. In generalized residual prediction, the inter-layer residual prediction of the current video unit may be based on the residue of the current layer, the temporal prediction of the current layer, and the residue of the reference layer. The residue of the reference layer may be adjusted by a weighting factor. The weighting factor may be based on and include various types of information. Examples of such information may include number of weighting candidates, weighting step, weighting index, and weighting table.
The GRP framework according to aspects of this disclosure may accommodate various types of residual prediction by incorporating a weighting factor. Adjusting the weighting factor appropriately may lead to significant coding gains for residual prediction. In addition, in the GRP framework, residual prediction may be performed using a reference layer that is not necessarily the base layer in traditional residual prediction. For example, the reference layer may be derived from the current enhancement layer. The GRP may also accommodate traditional residual prediction when the weighting factor is set to 1. The GRP framework may be used with both single-loop decoding and multi-loop decoding. In addition, in the GRP framework, motion estimation in difference domain may not be necessary, and therefore, the current layer and the enhancement layer do not have to share the same motion for motion estimation. The GRP framework can apply to many different types of residual prediction, and the traditional residual prediction as defined in Equation (1) and difference domain motion compensation are two specific scenarios of using the GRP framework. The techniques may improve the performance of motion compensation in scalable extension of HEVC and may also apply to 3D video coding extension of HEVC.
According to certain aspects, the differential value between the predictions of the current (e.g., enhancement) layer and the reference (e.g., base) layer may be employed as a difference predictor. In some embodiments, the differential value may be referred to as differential pixels. Because the enhancement layer and the base layer may have different quality target, the motion of the temporal predictions of the current layer and the base layer may be different. In some situations, reconstruction using the differential value may be more efficient and/or yield better results. For example, when there is a scene change or cut, such that sequential frames can be very different from one another, reconstruction using the difference predictor may be preferred. A weighting factor can be applied to the differential value. Such technique may be referred to as weighted difference prediction (WDP). In some embodiments, WDP can be implemented as an extension to the GRP framework.
FIGS. 8A and 8B are flowcharts illustrating example methods for generalized residual prediction using residual pixels (e.g., Î r −P r ) and differential pixels (e.g., P c −P r ), respectively. The techniques described in this disclosure may provide a generalized residual prediction (GRP) framework. As explained above, inter-layer residual prediction uses the residue of the reference layer in predicting the current video unit, for example, a frame. In generalized residual prediction, the inter-layer residual prediction of the current video unit may be based on the residue of the current layer, the temporal prediction of the current layer, and the residue of the reference layer. The residue of the reference layer may be adjusted by a weighting factor. The GRP scheme may be defined as follows:

Î c =r c +P c +w·r r (5)

where Î c denotes the reconstruction of the current frame, r c denotes the residual prediction from the current layer, P c denotes the temporal prediction from the same layer, r r denotes the residual prediction from the reference layer, and w denotes a weighting factor.

The weighting factor may be based on and include various types of information. Examples of such information may include number of weighting candidates, weighting step, weighting index, and weighting table. The number of weighting candidates may indicate the number of different weighting factors that are available for applying to the residue of the reference layer. The weighting step may indicate the size of the increment or unit between the available weighting factors. The weighting index may indicate a particular weighting factor among the available weighting factors. The weighting table can include information about the weighting factor and may be accessed by the weighting index, similar to a lookup table. In a specific example, three weighting factor candidates may be available: 0.0, 0.5, and 1.0. In this example, the number of weighting candidates is 3 because three weighting factor candidates are available. The weighting step between the 3 weighting candidates is 0.5. Each weighting candidate can be identified by a weighting index. Weighting factor 0 is identified by index 0, weighting factor 0.5 by index 1, and weighting factor 1.0 by index 2. The weighting step and the index may be used to derive the weighting factor since signaling fractions may be costly.
The GRP framework according to aspects of this disclosure may accommodate various types of residual prediction by incorporating a weighting factor. Adjusting the weighting factor appropriately may lead to significant coding gains for residual prediction. The GRP may improve coding performance while reducing the amount of memory and computational cost by incorporating weighting information for the reference layer in residual prediction. For example, the GRP can improve coding performance since the weighted residue prediction is more accurate. Also, the amount of memory and computational cost can be reduced, for example, because two sets of motion compensation loops are not typically used as in difference domain motion compensation. In addition, in the GRP framework, residual prediction may be performed using a reference layer that is not necessarily the base layer in traditional residual prediction. For example, the reference layer may be derived from the enhancement layer of the current layer. The GRP may also accommodate traditional residual prediction when the weighting factor is set to 1. The GRP framework may be used with both single-loop decoding and multi-loop decoding.
With respect to difference domain motion compensation, the GRP framework may be applied in single-loop decoding. As explained above, in H.264, difference domain motion compensation cannot be employed in a single-loop decoding scenario since differential pictures between layers have to be calculated based on the fully reconstructed picture of each layer. In order to obtain the difference picture in difference domain motion compensation, full reconstruction of each layer is often used, and for each layer, one motion compensation loop may be used for full reconstruction. For example, two motion compensation loops are often used to have full reconstruction of two layers. Accordingly, difference domain motion compensation cannot be employed in single-loop decoding. In contrast, the GRP may support both single-loop decoding and multi-loop decoding. In addition, in the GRP framework, motion estimation in difference domain may not be necessary. Therefore, the current layer and the enhancement layer do not have to share the same motion for motion estimation. The GRP framework is applicable to many different types of residual prediction, and the traditional residual prediction as defined in Equation (1) and difference domain motion compensation are two specific scenarios of using the GRP framework.
The example method for generalized residual prediction according to aspects of this disclosure will now be explained with reference to FIG. 8A . The process 800 A may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 800 A are described with respect to the decoder 30 in FIG. 3 , but the process 800 A may be performed by other components, such as an encoder, as mentioned above.
At block 801 A, the decoder 30 applies a weighting factor to the residual prediction from the reference layer. As explained above, the generalized residual prediction (GRP) may apply a weighting factor to the residue from the reference layer. The weighting factor can be determined to be optimal for a particular scenario, such as single-loop decoding. The weighting factor may include information such as the number of weighting candidates, the weighting step, the weighting index, and the weighting table.
At block 802 A, the decoder 30 obtains the residual prediction from the enhancement layer. At block 803 A, the decoder 30 obtains the temporal prediction from the enhancement layer.
At block 804 A, the decoder 30 determines the current video unit based on the residual prediction from the reference layer adjusted by the weighting factor, the residual prediction from the enhancement layer, and the temporal prediction from the enhancement layer. As explained above, in GRP, the current video unit may be predicted according to Equation (5).
The example method for generalized residual prediction according to aspects of this disclosure will now be explained with reference to FIG. 8B . The process 800 B may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 800 B are described with respect to the decoder 30 in FIG. 3 , but the process 800 B may be performed by other components, such as an encoder, as mentioned above.
At block 801 B, the decoder 30 applies a weighting factor to the difference prediction. The generalized residual prediction (GRP) may apply a weighting factor to the difference between the predictions of the current or enhancement layer (P e ) and the reference or base layer (P b ). The weighting factor can be determined to be optimal for a particular scenario, such as multi-loop decoding. The weighting factor may include information such as the number of weighting candidates, the weighting step, the weighting index, and the weighting table.
At block 802 B, the decoder 30 obtains the residual prediction from the enhancement layer. At block 803 B, the decoder 30 obtains the reconstruction of the current picture in the reference layer.
At block 804 B, the decoder 30 determines the current video unit based on the difference prediction adjusted by the weighting factor, the residual prediction from the enhancement layer, and the reference layer reconstruction. The current video unit may be predicted according to Equation (7B), discussed below.
The example method for generalized residual prediction according to aspects of this disclosure described with respect to FIG. 8A and FIG. 8B may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIGS. 8A and 8B may be implemented separately, or in combination with one another.
FIG. 9 is a flowchart illustrating an example method for generalized residual prediction using single-loop decoding according to aspects of this disclosure. As explained above, in single-loop decoding, one loop is used for motion compensation of the enhancement layer. In the scenario of single-loop decoding, no full reconstruction of the base layer is available. Accordingly, normalized residual of base layer may be directly employed as the base residual predictor. For the enhancement layer, the reconstruction Î e may be determined as follows:

Î e =r e +P e +w·r b =r e +P e +w·r b ′·( Q e /Q b ) (6)

where r e and P e denote de-quantized residue and temporal prediction of the enhancement layer, r b denotes the normalized base layer residual predictor (up-sampled in spatial scalable case), r b ′ denotes the base layer residue, and Q e and Q b denote the quantization step of the enhancement layer and the base layer, respectively.

The example method for generalized residual prediction using single-loop decoding according to aspects of this disclosure will now be explained with reference to FIG. 9 . The process 900 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 900 are described with respect to the decoder 30 in FIG. 3 , but the process 900 may be performed by other components, such as an encoder, as mentioned above. At block 901 , the decoder 30 determines a weighting factor for the residual prediction from the reference layer in single-loop decoding for the GRP framework. At block 902 , the decoder 30 determines the current video unit based on the residual prediction from the RL adjusted by the weighting factor, the residual prediction from the EL, and the temporal prediction from the EL. For example, as explained above with respect to Equation (6), the normalized base layer residue may be used for the RL residual prediction. The example method for generalized residual prediction using single-loop decoding according to aspects of this disclosure described with respect to FIG. 9 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 9 may be implemented separately, or in combination with one another.
FIG. 10A and FIG. 10B are flowcharts illustrating example methods for generalized residual prediction using multi-loop decoding according to aspects of this disclosure. As explained above, in multi-loop decoding, multiple loops are used for motion compensation of the enhancement layer. In the scenario of multi-loop decoding, full reconstruction of base layer is available when encoding/decoding the enhancement layer. Accordingly, the differential value between the reconstruction of the previously coded enhancement layer and base layer (up-sampled if necessary) may be employed as the residual predictor. For the enhancement layer, the reconstruction Î e may be determined as follows:

Î e =r e +P e +w ·( Î b −P b ) (7A)

where r e indicates the de-quantized residue of the current video unit in the enhancement layer, P e and P b indicate the temporal prediction for the current video unit in the enhancement layer and the base layer, respectively, and Î b indicates the full reconstruction of the current video unit in the base layer. Because the enhancement layer and the base layer may have different quality target, the motion of the temporal predictions P e and P b may be different.

If the base layer and the enhancement layer have the same motion, the motion of the temporal predictions P e and P b are the same, and Equation (7A) can be employed directly. When decoding an inter video unit of the enhancement layer, its enhancement layer and base layer temporal predictions P e and P b are both available. The base layer reconstruction Î b is also available. Accordingly, the reconstruction Î e can be obtained from the de-quantized residue r e and w, which may be signaled or derived as explained in more detail with respect to FIGS. 11 and 12 .
If the base layer and the enhancement layer have different motion, the motion of the enhancement layer and base layer temporal predictions P e and P b is different, and the residual of the base layer and the residual of the enhancement layer may not be correlated. In such case, residual prediction may not lead to good results. In order to improve performance of residual prediction, it may be assumed that the enhancement layer and base layer temporal predictions share the same motion. In addition to or instead of assuming that the EL and the BL temporal predictions share the same motion, either the motion of the base layer or the motion of the enhancement layer may be forced to apply to another layer to generate the residual predictor. For example, the motion of the enhancement layer temporal prediction P e may be applied to the base layer to get P b . In such case, two motion compensations are often used to decode the enhancement layer since both P e and P b may be generated with the motion of P e .
In another embodiment, the differential value between the predictions of the current (e.g., enhancement) layer and the reference (e.g., base) layer may be employed as a difference predictor. For the enhancement layer, the reconstruction Î e may be determined as follows:

Î e =r e +Î b +w ·( P e −P b ) (7B)

where r e indicates the de-quantized residue of the current video unit in the enhancement layer, P e and P b indicate the temporal prediction for the current video unit in the enhancement layer and the base layer, respectively, and Î b indicates the full reconstruction of the current video unit in the base layer. Because the enhancement layer and the base layer may have different quality target, the motion of the temporal predictions P e and P b may be different. In many situations, the reconstruction according to equation (7A) will be more efficient than according to equation (7B). However, in some situation, reconstruction according to equation (7B) will be more efficient and/or yield better results. For example, when there is a scene change or cut, such that sequential frames are very different from one another, reconstruction according to equation (7B) is preferred.

In one embodiment, different weighting indexes are assigned to normal GRP weighting factors and WDP weighting factors. For example, in one embodiment, four weighting indexes are allowed at the CU level. Weighting indexes 0, 1, and 2 indicate that equation (7A) is used for the prediction calculation, where w=0, 0.5, and 1, respectively. Weighting index 3 indicates that equation (7B) is used for the prediction calculation, and w=0.5. In another embodiment, GRP weighting factors (e.g., equation (7A)) are all disabled and only WDP weighting factors (e.g., equation (7B)) are allowed. All the methods describe herein with respect to GRP, including but not limited to parameter signaling/derivation methods, weighting factor determination methods, related picture/motion upsampling, downsampling methods, etc., may be applied to WDP, as well.
Example methods for generalized residual prediction using multi-loop decoding according to aspects of this disclosure will now be explained with reference to FIG. 10A and FIG. 10B . The process 1000 A and process 1000 B may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1000 A and process 1000 B are described with respect to the decoder 30 in FIG. 3 , but the process 1000 A and process 1000 B may be performed by other components, such as an encoder, as mentioned above.
Referring to FIG. 10A , at block 1001 A, the decoder 30 determines a weighting factor for the residual prediction from the reference layer in multi-loop decoding for the GRP framework. At block 1002 A, the decoder 30 determines the current video unit based on the residual prediction from the RL adjusted by the weighting factor, the residual prediction from the EL, and the temporal prediction from the EL. For example, as explained above with respect to Equation (7), Î b −P b may be used for the RL residual prediction.
Referring to FIG. 10B , at block 1001 B, the decoder 30 determines a weighting factor for the difference prediction in multi-loop decoding for the GRP framework. At block 1002 B, the decoder 30 determines the current video unit based on the difference prediction adjusted by the weighting factor, the residual prediction from the EL, and the RL, reconstruction (for example, the full reconstruction of the current picture in the reference layer). For example, as explained above with respect to Equation (7B), P e −P b (or P c −P r ) may be used for the difference prediction.
The example method for generalized residual prediction using multi-loop decoding according to aspects of this disclosure described with respect to FIG. 10A and FIG. 10B may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 10A and FIG. 10B may be implemented separately, or in combination with one another.
In some situations, the residual pixels (e.g., Î b −P b ) and differential pixels (e.g., P e −P b ) may extend beyond an allocated or desired bit-depth. For example, in some situations, these pixels may not be able to be represented in 8- or 16-bits. This can create complications for hardware implementation. Therefore, in some embodiments, clipping is performed to truncate the residual or differential pixels to make sure each falls within a desired range, such as, but not limited to, 8-bit or 16-bit representation.
FIG. 11 is a flowchart illustrating an example method for signaling generalized residual prediction parameters according to aspects of this disclosure. As explained above, weighting information may include the number of weighting candidates, the weighting step (or the weighting table), and the weighting index. The weighting factor w may be determined based on such weighting information. The number of weighting candidates may be denoted by N w . The weighting step may be denoted by S w , and the weighting table by W T . The weighting index may be denoted by i w . In one embodiment, the weighting factor w is derived based on the weighting step S w and the weighting index i w as follows:

w=S w ·i w (8)

In another embodiment, w may be obtained from a lookup table W T according to the index i w .

The weighting factor information, which may include, but is not limited to, N w , S w , W T , and i w , may signaled in various ways. In some embodiments, the weighting step S w or the weighting table W T may be hard coded or signaled. S w or W T may be signaled at the sequence level or picture level. The weighting index i w may be signaled at a lower level, such as CU and PU.
In one embodiment, the weighting step S w is represented with 3-bit quantization (S w may be ⅛, 2/8, . . . , 8/8) and unsigned integer Exp-Golomb coded in Sequence Parameter Set (SPS). Considering that N w ≧1, (N w −1) is also unsigned integer Exp-Golomb coded in SPS. The weighting index i w , is first binarized with truncated unary code (with N w as the maximum number) and then CABAC coded. In CABAC coding, the first bin is coded with one context, and rest of the bins are coded with another context. To code the weighting index i w , the context may depend on previously coded parameters. For example, i w of spatially neighboring CUs, such as left and top CUs, may be used as the context for the weighting index i w of the current CU. Also, the type of the current CU, such as whether the current CU is skip or merge coded, or the size of the current CU may be used as the context.
In other embodiments, different CU modes may have different weighting factor signaling methods. For example, for skip and merge modes, three weighting factors (such as w=0, w=0.5, and w=1) may be selected and signaled. For inter modes other than skip and merge modes, only two weighting factors (such as w=0 and w=1) may be selected and signaled. Alternatively, for inter modes other skip and merge modes, only one fixed weighting factor may be applied. In such case, no additional signaling for weighting factor may be used.
The example method for signaling generalized residual prediction parameters according to aspects of this disclosure will now be explained with reference to FIG. 11 . The process 1100 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1100 are described with respect to the encoder 20 in FIG. 2 , but the process 1100 may be performed by other components, such as a decoder, as mentioned above. At block 1101 , the encoder 20 signals the weighting step or the weighting table. At block 1102 , the encoder 20 signals the number of weighting candidates. At block 1103 , the encoder 20 signals the weighting index. The steps in the process 1100 may be performed in a different order. For example, the number of weighting candidates may be signaled before the weighting step (or the weighting table). The example method for signaling generalized residual prediction parameters according to aspects of this disclosure described with respect to FIG. 11 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 11 may be implemented separately, or in combination with one another.
FIG. 12 is a flowchart illustrating an example method for deriving generalized residual prediction parameters according to aspects of this disclosure. The GRP parameters may be signaled as explained with respect to FIG. 11 . The GRP parameters may also be derived from information included in the bitstream. The GRP parameters may be derived fully or partially from the information in the bitstream. In one embodiment, the weighting step S w is derived at CU level according to the related CU size. An example mapping between the weighting step S w and the CU size is listed in the table below.





TABLE 1



Example Mapping Between Weighting Step and CU Size






CU size
S W





64 × 64
⅛


32 × 32
¼


16 × 16
½


8 × 8
½

























In another embodiment, the maximum number of weighting candidates is adjusted at CU level based on previously coded information, such as CU mode, CU size, and quantization. For example, for CUs smaller than 16×16, only two weighting candidates may be allowed, e.g., in order to save signaling cost.
The example method for deriving generalized residual prediction parameters according to aspects of this disclosure will now be explained with reference to FIG. 12 . The process 1200 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1200 are described with respect to the decoder 30 in FIG. 3 , but the process 1200 may be performed by other components, such as an encoder, as mentioned above.
At block 1201 , the decoder 30 obtains information from the bitstream or obtains previously coded information in order to determine the weighting information. For example, as explained above, the GRP parameters may be based on the CU size. Or the GRP parameters may be based on previously coded information, such as CU mode, CU size, and quantization. At block 1202 , the decoder 30 determines one or more parameters for generalized residual prediction based on the information obtained at block 1201 . For example, the decoder 30 may determine the weighting step S w based on the CU size. The decoder 30 may also determine the number of weighting candidates N w based on the CU size. The decoder 30 may also adjust the weighting information based on previously coded information, such as CU mode, CU size, and quantization. The example method for deriving generalized residual prediction parameters according to aspects of this disclosure described with respect to FIG. 8 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 12 may be implemented separately, or in combination with one another.
FIG. 13 is a flowchart illustrating an example method for up-sampling or down-sampling a layer in generalized residual prediction according to aspects of this disclosure. In inter-layer prediction process, an up-sampling or down-sampling filtering process is applied to the base layer picture to match the spatial aspect ratio of the enhancement layer. A filtering process, such as a smoothing filter, can be also applied even when the picture size of the base layer and the enhancement layer is identical. In general, one fixed upsampling, downsampling, and smoothing filter set is used and hard-coded. The filter may be selected from the set according to the fractional pixel shift (sometimes referred to as phase), which is calculated based on the spatial aspect ratio between the base layer and enhancement layer pictures.
In the GRP framework, variant filtering sets may be applied to improve inter-layer prediction performance. The filtering sets may be hard coded or signaled at sequence or picture level. The filter set index may be signaled or derived at a lower level, such as CU and PU. The filter set index may be derived based on the value of the weighting factor w, or may be derived based on the weighting index i w . The derivation mapping table between the filtering set index and the weighting factor w, or between the filtering set and the weighting index i w may be hard coded or signaled at sequence or picture level.
The example method for up-sampling or down-sampling a layer in generalized residual prediction according to aspects of this disclosure will now be explained with reference to FIG. 13 . The process 1300 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1300 are described with respect to the decoder 30 in FIG. 3 , but the process 1300 may be performed by other components, such as an encoder, as mentioned above.
At block 1301 , the decoder 30 determines whether to up-sample the reference layer or down-sample the enhancement layer. In spatial scalability, such up-sampling and down-sampling is performed so that inter-layer prediction can be performed at the same resolution. If it is determined that the reference layer will be up-sampled at block 1302 , the decoder 30 up-samples the reference layer to the resolution of the enhancement layer at block 1303 . On the other hand, if it is determined that the enhancement layer will be down-sampled at block 1302 , the decoder 30 down-samples the enhancement layer to the resolution of the reference layer at block 1304 . At block 1305 , the decoder 30 applies a smoothing filter to the up-sampled or down-sampled pictures. The smoothing filter may be applied even if the enhancement layer and the reference layer pictures are the same. The smoothing filter may be selected appropriately. At block 1306 , the decoder 30 determines the current video unit using GRP based on the up-sampled or down-sampled pictures.
The example method for up-sampling or down-sampling a layer in generalized residual prediction according to aspects of this disclosure described with respect to FIG. 13 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 13 may be implemented separately, or in combination with one another.
FIG. 14 is a flowchart illustrating an example method for remapping, up-sampling, or down-sampling motion information in generalized residual prediction according to aspects of this disclosure. In some cases, when applying motion information of one layer to another layer to generate the residual predictor, the reference available in one layer may not be available in another layer. In such case, motion remapping is necessary. In one embodiment, if a reference is only available in one layer, it is marked as unavailable so that this reference will not be used to generate residual predictor in the proposed GRP framework. In another embodiment, the unavailable reference is replaced by the reference at the beginning of the related reference list, and the motion is set to zero motion.
In 3D video coding, SVC video data also includes video data for different views. Because the views may relate to different angles, disparity may exist between the different views. If motion is remapped in the context of 3D video coding, the disparity vector may be considered in remapping the motion.
In spatial scalable cases, motion vector may be up-sampled or down-sampled due to different resolutions between the enhancement layer and the base layer. In one embodiment, motion vector scaling is directly based on the resolution ratio. In another embodiment, additional phase shift (+1 or −1) may be applied subsequent to direct scaling. The additional phase shift may be signaled in the bitstream or derived based on previously coded information, such as PU size, motion vector, CU depth, etc.
The example method for remapping, up-sampling, or down-sampling motion information according to aspects of this disclosure will now be explained with reference to FIG. 14 . The process 1400 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1400 are described with respect to the decoder 30 in FIG. 3 , but the process 1400 may be performed by other components, such as an encoder, as mentioned above. At block 1401 , if a reference for motion information is not available in one of the layers, the decoder 30 remaps motion information at block 1402 . For example, the decoder 30 can mark a reference as unavailable if the corresponding reference in another layer is unavailable. Or the decoder 30 may remap the reference to a reference in a related reference list. If a reference for motion information available in the layers used for inter-prediction at block 1401 , the decoder 30 may not perform further processing as shown in block 1403 . At block 1404 , if spatial SVC is used, the decoder 30 determines whether to up-sample the reference layer motion information or to down-sample the enhancement layer motion information at block 1405 . If spatial scalability is not used, the decoder 30 may not perform any further processing as shown at block 1406 . At block 1407 , if it is determined that the reference layer motion information will be up-sampled, the decoder 30 up-samples the reference layer motion information to the resolution of the enhancement layer at block 1008 . On the other hand, if it is determined that the enhancement layer motion information will be down-sampled at block 1407 , the decoder 30 down-samples the enhancement layer motion information to the resolution of the reference layer at block 1409 . At block 1410 , the decoder 30 determines the current video unit using GRP using the up-samples or down-sampled pictures.
The example method for remapping, up-sampling, or down-sampling motion information according to aspects of this disclosure described with respect to FIG. 10 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 14 may be implemented separately, or in combination with one another.
FIG. 15 is a flowchart illustrating an example method for determining a weighting factor for encoding in generalized residual prediction according to aspects of this disclosure. The example method may apply to encoder-side optimization. In one embodiment, the best weighting factor w for each CU is determined by checking the CU rate-distortion cost with each weighting factor candidate. The weighting factor with minimal cost is selected as the weighting factor w for the CU. In another embodiment, the residual predictor is derived by applying the motion of the enhancement layer temporal prediction P e to the base layer temporal prediction P b . The weighting factor w may be determined as follows:







w
=



∑

x
,
y


⁢

{


(

I
-

P
e


)

·

(



I
^

b

-

P
b


)


}




∑

x
,
y


⁢

{


(



I
^

b

-

P
b


)

2

}







(
9
)







where I indicates the source picture for the enhancement layer, Σ x,y {(I−P e )·(Î b −P b )} indicates the sum of dot product of the differential block (I−P e ) and (Î b −P b ).
The example method for determining a weighting factor for encoding in generalized residual prediction according to aspects of this disclosure will now be explained with reference to FIG. 15 . The process 1500 may be performed by an encoder (e.g., the encoder as shown in FIG. 2 ), a decoder (e.g., the decoder as shown in FIG. 3 ), or any other component. The steps of the process 1500 are described with respect to the encoder 20 in FIG. 2 , but the process 1500 may be performed by other components, such as a decoder, as mentioned above. At block 1501 , the encoder 20 derives the residual prediction of the EL by applying motion of the EL temporal prediction to the BL temporal prediction. At block 1502 , the decoder 30 derives the weighting factor based on the derived residual prediction. The example method for determining a weighting factor for encoding in generalized residual prediction according to aspects of this disclosure described with respect to FIG. 15 may be implemented at various coding levels, such as sequence, picture, a group of frames, frame, a group of slices, slice, a group of coding units (CUs), coding unit (CU), a group of prediction units (PUs), prediction unit (PU), blocks, or a region of pixels. In addition, all embodiments described with respect to FIG. 15 may be implemented separately, or in combination with one another.
It is to be recognized that depending on the example, certain acts or events of any of the techniques described herein can be performed in a different sequence, may be added, merged, or left out altogether (e.g., not all described acts or events are necessary for the practice of the techniques). Moreover, in certain examples, acts or events may be performed concurrently, e.g., through multi-threaded processing, interrupt processing, or multiple processors, rather than sequentially.
In one or more examples, the functions described may be implemented in hardware, software, firmware, or any combination thereof. If implemented in software, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium and executed by a hardware-based processing unit. Computer-readable media may include computer-readable storage media, which corresponds to a tangible medium such as data storage media, or communication media including any medium that facilitates transfer of a computer program from one place to another, e.g., according to a communication protocol. In this manner, computer-readable media generally may correspond to (1) tangible computer-readable storage media which is non-transitory or (2) a communication medium such as a signal or carrier wave. Data storage media may be any available media that can be accessed by one or more computers or one or more processors to retrieve instructions, code and/or data structures for implementation of the techniques described in this disclosure. A computer program product may include a computer-readable medium.
By way of example, and not limitation, such computer-readable storage media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage, or other magnetic storage devices, flash memory, or any other medium that can be used to store desired program code in the form of instructions or data structures and that can be accessed by a computer. Also, any connection is properly termed a computer-readable medium. For example, if instructions are transmitted from a website, server, or other remote source using a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless technologies such as infrared, radio, and microwave, then the coaxial cable, fiber optic cable, twisted pair, DSL, or wireless technologies such as infrared, radio, and microwave are included in the definition of medium. It should be understood, however, that computer-readable storage media and data storage media do not include connections, carrier waves, signals, or other transitory media, but are instead directed to non-transitory, tangible storage media. Disk and disc, as used herein, includes compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk and Blu-ray disc, where disks usually reproduce data magnetically, while discs reproduce data optically with lasers. Combinations of the above should also be included within the scope of computer-readable media.
Instructions may be executed by one or more processors, such as one or more digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), or other equivalent integrated or discrete logic circuitry. Accordingly, the term “processor,” as used herein may refer to any of the foregoing structure or any other structure suitable for implementation of the techniques described herein. In addition, in some aspects, the functionality described herein may be provided within dedicated hardware and/or software modules configured for encoding and decoding, or incorporated in a combined codec. Also, the techniques could be fully implemented in one or more circuits or logic elements.
The techniques of this disclosure may be implemented in a wide variety of devices or apparatuses, including a wireless handset, an integrated circuit (IC) or a set of ICs (e.g., a chip set). Various components, modules, or units are described in this disclosure to emphasize functional aspects of devices configured to perform the disclosed techniques, but do not necessarily require realization by different hardware units. Rather, as described above, various units may be combined in a codec hardware unit or provided by a collection of interoperative hardware units, including one or more processors as described above, in conjunction with suitable software and/or firmware.
Various examples have been described. These and other examples are within the scope of the following claims.