Secondary transforms with flexible coefficient support region

Abstract

Secondary transform coefficients for a current block of image data are determined. An index identifying a support region of the current block is determined, where the support region comprises a proper subset of positions of the current block. An inverse of a secondary transform is applied to the secondary transform coefficients to obtain primary transform coefficients. The primary transform coefficients are arranged according to the support region of the current block. A primary transform for the current block is determined, and the primary transform coefficients are inverse transformed using the primary transform to obtain a residual block of pixels corresponding to the current block. The current block is reconstructed using the residual block.

Description

Atty. Doc. No. GOGL-2272-A-WO PATENTSECONDARY TRANSFORMS WITH FLEXIBLE COEFFICIENT SUPPORTREGIONBACKGROUND

[0001] Digital images and video can be used, for example, on the internet, for remote business meetings via video conferencing, high-definition video entertainment, video advertisements, or sharing of user-generated content. Due to the large amount of data involved in transferring and processing image and video data, high-performance compression may be advantageous for transmission and storage. Accordingly, it would be advantageous to provide high-resolution image and video transmitted over communications channels having limited bandwidth.SUMMARY

[0002] This application relates to encoding and decoding of image data, video stream data, or both for transmission, storage, or both. Disclosed herein are aspects of systems, methods, and apparatuses for encoding and decoding using secondary transforms with a flexible coefficient support region.

[0003] An aspect of the teachings herein is a method that includes determining secondary transform coefficients for a current block of image data and determining an index identifying a support region of the current block, wherein the support region comprises a proper subset of positions of the current block. The support region can also be called a transform support region because the positions are positions of the transform coefficients of the block. The method also includes applying an inverse of a secondary transform to the secondary transform coefficients to obtain primary transform coefficients, arranging the primary transform coefficients according to the support region of the current block, determining a primary transform for the current block, inverse transforming the primary transform coefficients using the primary transform to obtain a residual block of pixels corresponding to the current block, and reconstructing the current block using the residual block.

[0004] In some implementations of the method, the method includes determining the secondary transform for the current block by using the index in a lookup table to identify the secondary transform.

[0005] In some implementations of the method, determining the secondary transformcoefficients includes decoding, from an encoded bitstream, the secondary transform coefficients.

[0006] In some implementations of the method, the secondary transform is a non- separable transform, and the primary transform is separable transforms.

[0007] In some implementations of the method, the primary transform coefficients are quantized transform coefficients, and the method includes dequantizing the primary transform coefficients before the inverse transforming.

[0008] In some implementations of the method, the support region is one of multiple available support regions. At least some of the multiple available support regions have a same cardinality (e.g., number) of positions and form a set of support regions having a respective different proper subset of positions of the current block. In a variation of this method, the multiple available support regions include support regions having different cardinalities of positions.

[0009] In some implementations of the method, a kernel dimension of the secondary transform is KxM, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, and the support region is one of a set of support regions, each support region having K positions arranged different from each other.

[0010] In some implementations of the method, the index is a 4-ary signal and a 7-ary signal, and one available value of the 4-ary signal indicates that no secondary transform is used for coding a block. In such a case (where no secondary transform is used), the 7-ary signal may be omitted from the bitstream.

[0011] An aspect of the teachings herein is another method that can include determining secondary transform coefficients for a current block of image data; determining a secondary transform index from an encoded bitstream, where the secondary transform index identifies a secondary transform kernel and a corresponding support region; applying an inverse of the secondary transform kernel identified by the secondary transform index to the secondary transform coefficients to obtain primary transform coefficients; arranging the primary transform coefficients according to the support region of the current block; determining a primary transform for the current block; inverse transforming the primary transform coefficients using the primary transform to obtain a residual block of pixels corresponding to the current block; and reconstructing the current block using the residual block.

[0012] In some implementations, the method includes deriving the support region for the current block from the secondary transform index.

[0013] In some implementations of the method, deriving the support region may includeusing a lookup table that associates the secondary transform index with a specific support region pattern.

[0014] In some implementations of the method, the lookup table is organized based on a type of primary transform applied to the current block, and different lookup table mappings are used for different primary transform types.

[0015] An aspect of the teachings herein is another method that can include transforming a residual block of pixels corresponding to a current block using a primary transform to obtain primary transform coefficients, applying a secondary transform to primary transform coefficients within a support region of the current block to obtain secondary transform coefficients, wherein the support region comprises a proper subset of positions of the current block, transmitting, within an encoded bitstream, the secondary transform coefficients, and transmitting, within the encoded bitstream, an index identifying the support region of the current block.

[0016] In some implementations of this method, the secondary transform is a non- separable transform, the primary transform is separable transforms, the index identifies the secondary transform, and the method includes quantizing the primary transform coefficients before applying the secondary transform.

[0017] In some implementations of this method, a kernel dimension of the secondary transform is KxM, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, the support region is one of a set of support regions, and each support region has K positions arranged different from each other.

[0018] In some implementations of this method, the method includes setting a defined number of the secondary transform coefficients to zero before transmitting remaining secondary transform coefficients.

[0019] An aspect of the teachings herein is an apparatus includes a processor that is configured to perform any of the methods described herein.

[0020] An aspect of the teachings herein is a non-transitory computer-readable medium storing instructions that, when executed, cause a processor to perform any of the methods described herein.

[0021] An aspect of the teachings herein is a non-transitory computer-readable medium storing an encoded bitstream. The encoded bitstream includes encoded secondary transform coefficients for a current block of image date, an index identifying a support region of the current block, wherein the support region comprises a proper subset of positions of the current block, and an identifier of a primary transform for the current block to be applied toprimary transform coefficients arranged according to the support region of the current block and resulting from applying an inverse of a secondary transform to the secondary transform coefficients, wherein the primary transform is used for inverse transforming the primary transform coefficients to obtain a residual block of pixels corresponding to the current block

[0022] In some implementations of this non-transitory computer-readable medium, the index identifies the secondary transform.

[0023] Variations in these and other aspects will be described in additional detail hereafter.BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The description herein refers to the accompanying drawings wherein like reference numerals refer to like parts throughout the several views unless otherwise noted or otherwise clear from context.

[0025] FIG. 1 is a diagram of a computing device in accordance with implementations of this disclosure.

[0026] FIG. 2 is a diagram of a computing and communications system in accordance with implementations of this disclosure.

[0027] FIG. 3 is a diagram of a video stream for use in encoding and decoding in accordance with implementations of this disclosure.

[0028] FIG. 4 is a block diagram of an encoder in accordance with implementations of this disclosure.

[0029] FIG. 5 is a block diagram of a decoder in accordance with implementations of this disclosure.

[0030] FIG. 6 is a block diagram of a representation of a portion of a frame in accordance with implementations of this disclosure.

[0031] FIG. 7 is a diagram of three support regions for performing a secondary transform in accordance with implementations of this disclosure.

[0032] FIG. 8 is a diagram of three further support regions for performing a secondary transform in accordance with implementations of this disclosure.

[0033] FIG. 9 is a flowchart of a technique of reconstructing a current block using secondary transforms with flexible coefficient support region.

[0034] FIG. 10 is a flowchart of a technique of reconstructing a block using secondary transforms with flexible coefficient support region in accordance with implementations of this disclosure.

[0035] FIG. 11 is a flowchart of a technique of encoding a current block using secondarytransforms with flexible coefficient support region.DETAILED DESCRIPTION

[0036] Image and video compression schemes may include breaking an image, or frame, into smaller portions, such as blocks, and generating an output bitstream using techniques to minimize the bandwidth utilization of the information included for each block in the output. In some implementations, the information included for each block in the output may be limited by reducing spatial redundancy, reducing temporal redundancy, or a combination thereof. For example, temporal or spatial redundancies may be reduced by predicting a frame, or a portion thereof, based on information available to both the encoder and decoder, and including information representing a difference, or residual, between the predicted frame and the original frame in the encoded bitstream. The residual information may be further compressed by transforming the residual information into transform coefficients (e.g., energy compaction), quantizing the transform coefficients, and entropy coding the quantized transform coefficients. Other coding information, such as motion information, may be included in the encoded bitstream, which may include transmitting differential information based on predictions of the encoding information, which may be entropy coded to further reduce the corresponding bandwidth utilization. An encoded bitstream can be decoded to reconstruct the blocks and the source images from the limited information. In some implementations, the accuracy, efficiency, or both, of coding a block using either interprediction or intra-prediction may be limited.

[0037] In transforming the residual information (also called residuals or residual blocks or residues), separable transforms may be applied (e.g., horizontally and vertically transforms) to a (e.g., rectangular) block to decorrelate the residues. These transforms are referred to as primary transforms herein. It is possible to then apply non-separable secondary transforms to achieve better energy compactness.

[0038] To capture a high diversity of residue patterns, multiple secondary transforms may be defined in a codec. The codec can, for example, search and signal the best secondary transform type. In an example described in further detail below, for transform blocks greater than or equal to 8x8 samples, one among 42 secondary transforms may be used. Depending on the size of the transform blocks and given a large number of transform kernels, memory requirements and the computational complexity of the secondary transform operations (e.g., transformation and inverse transformation) can be very high. For example, where each of the 42 transforms has a dimension of 64x32, which takes in 64 primary transform coefficients, 32output coefficients are generated. The memory requirement for the kernels is equal to 42*64*32.

[0039] As used herein, a “secondary transform” refers to a transform that is applied to a set of primary transform coefficients to achieve improved energy compaction. Whereas primary transforms are applied to pixel-domain samples to obtain frequency-domain transform coefficients, secondary transforms operate on a collection of frequency-domain transform coefficients to obtain therefrom another collection of frequency -domain transform coefficients. Secondary transforms are typically non- separable transforms. Unlike primary transforms which are typically separable and applied independently in horizontal and vertical directions, secondary transforms are typically non-separable and operate on a collection of coefficients simultaneously using a transform kernel matrix.

[0040] It is desirable to reduce kernel dimensions while keeping a wide diversity of secondary transform kernels. Using a flexible coefficient support region with the secondary transform kernels may allow such an improvement to transformation. Further details of the teachings herein are described below after an initial environment in which the teachings may be implemented.

[0041] FIG. 1 is a diagram of a computing device 100 in accordance with implementations of this disclosure. The computing device 100 shown includes a memory 110, a processor 120, a user interface (UI) 130, an electronic communication unit 140, a sensor 150, a power source 160, and a bus 170. As used herein, the term “computing device” includes any unit, or a combination of units, capable of performing any method, or any portion or portions thereof, disclosed herein.

[0042] The computing device 100 may be a stationary computing device, such as a personal computer (PC), a server, a workstation, a minicomputer, or a mainframe computer; or a mobile computing device, such as a mobile telephone, a personal digital assistant (PDA), a laptop, or a tablet PC. Although shown as a single unit, any one element or elements of the computing device 100 can be integrated into any number of separate physical units. For example, the user interface 130 and processor 120 can be integrated in a first physical unit and the memory 110 can be integrated in a second physical unit.

[0043] The memory 110 can include any non-transitory computer-usable or computer- readable medium, such as any tangible device that can, for example, contain, store, communicate, or transport data 112, instructions 114, an operating system 116, or any information associated therewith, for use by or in connection with other components of the computing device 100. The non-transitory computer-usable or computer-readable mediumcan be, for example, a solid-state drive, a memory card, removable media, a read-only memory (ROM), a random-access memory (RAM), any type of disk including a hard disk, a floppy disk, an optical disk, a magnetic or optical card, an application- specific integrated circuits (ASICs), or any type of non-transitory media suitable for storing electronic information, or any combination thereof.

[0044] Although shown a single unit, the memory 110 may include multiple physical units, such as one or more primary memory units, such as random-access memory units, one or more secondary data storage units, such as disks, or a combination thereof. For example, the data 112, or a portion thereof, the instructions 114, or a portion thereof, or both, may be stored in a secondary storage unit and may be loaded or otherwise transferred to a primary storage unit in conjunction with processing the respective data 112, executing the respective instructions 114, or both. In some implementations, the memory 110, or a portion thereof, may be removable memory.

[0045] The data 112 can include information, such as input audio data, encoded audio data, decoded audio data, or the like. The instructions 114 can include directions, such as code, for performing any method, or any portion or portions thereof, disclosed herein. The instructions 114 can be realized in hardware, software, or any combination thereof. For example, the instructions 114 may be implemented as information stored in the memory 110, such as a computer program, which may be executed by the processor 120 to perform any of the respective methods, algorithms, aspects, or combinations thereof, as described herein.

[0046] Although shown as included in the memory 110, in some implementations, the instructions 114, or a portion thereof, may be implemented as a special purpose processor, or circuitry, that can include specialized hardware for carrying out any of the methods, algorithms, aspects, or combinations thereof, as described herein. Portions of the instructions 114 can be distributed across multiple processors on the same machine or different machines or across a network such as a local area network, a wide area network, the Internet, or a combination thereof.

[0047] The processor 120 can include any device or system capable of manipulating or processing a digital signal or other electronic information now-existing or hereafter developed, including optical processors, quantum processors, molecular processors, or a combination thereof. For example, the processor 120 can include a special purpose processor, a central processing unit (CPU), a digital signal processor (DSP), a plurality of microprocessors, one or more microprocessor in association with a DSP core, a controller, a microcontroller, an Application Specific Integrated Circuit (ASIC), a Field ProgrammableGate Array (FPGA), a programmable logic array, programmable logic controller, microcode, firmware, any type of integrated circuit (IC), a state machine, or any combination thereof. As used herein, the term “processor” includes a single processor or multiple processors.

[0048] The user interface 130 can include any unit capable of interfacing with a user, such as a virtual or physical keypad, a touchpad, a display, a touch display, a speaker, a microphone, a video camera, a sensor, or any combination thereof. For example, the user interface 130 may be an audio-visual display device, and the computing device 100 may present audio, such as decoded audio, using the user interface 130 audio-visual display device, such as in conjunction with displaying video, such as decoded video. Although shown as a single unit, the user interface 130 may include one or more physical units. For example, the user interface 130 may include an audio interface for performing audio communication with a user, and a touch display for performing visual and touch-based communication with the user.

[0049] The electronic communication unit 140 can transmit, receive, or transmit and receive signals via a wired or wireless electronic communication medium 180, such as a radio frequency (RF) communication medium, an ultraviolet (UV) communication medium, a visible light communication medium, a fiber optic communication medium, a wireline communication medium, or a combination thereof. For example, as shown, the electronic communication unit 140 is operatively connected to an electronic communication interface 142, such as an antenna, configured to communicate via wireless signals.

[0050] Although the electronic communication interface 142 is shown as a wireless antenna in FIG. 1, the electronic communication interface 142 can be a wireless antenna, as shown, a wired communication port, such as an Ethernet port, an infrared port, a serial port, or any other wired or wireless unit capable of interfacing with a wired or wireless electronic communication medium 180. Although FIG. 1 shows a single electronic communication unit 140 and a single electronic communication interface 142, any number of electronic communication units and any number of electronic communication interfaces can be used.

[0051] The sensor 150 may include, for example, an audio-sensing device, a visible lightsensing device, a motion sensing device, or a combination thereof. For example, lOOthe sensor 150 may include a sound-sensing device, such as a microphone, or any other soundsensing device now existing or hereafter developed that can sense sounds in the proximity of the computing device 100, such as speech or other utterances, made by a user operating the computing device 100. In another example, the sensor 150 may include a camera, or any other image-sensing device now existing or hereafter developed that can sense an image suchas the image of a user operating the computing device. Although a single sensor 150 is shown, the computing device 100 may include a number of sensors 150. For example, the computing device 100 may include a first camera oriented with a field of view directed toward a user of the computing device 100 and a second camera oriented with a field of view directed away from the user of the computing device 100.

[0052] The power source 160 can be any suitable device for powering the computing device 100. For example, the power source 160 can include a wired external power source interface; one or more dry cell batteries, such as nickel-cadmium (NiCd), nickel-zinc (NiZn), nickel metal hydride (NiMH), lithium-ion (Li-ion); solar cells; fuel cells; or any other device capable of powering the computing device 100. Although a single power source 160 is shown in FIG. 1, the computing device 100 may include multiple power sources 160, such as a battery and a wired external power source interface.

[0053] Although shown as separate units, the electronic communication unit 140, the electronic communication interface 142, the user interface 130, the power source 160, or portions thereof, may be configured as a combined unit. For example, the electronic communication unit 140, the electronic communication interface 142, the user interface 130, and the power source 160 may be implemented as a communications port capable of interfacing with an external display device, providing communications, power, or both.

[0054] One or more of the memory 110, the processor 120, the user interface 130, the electronic communication unit 140, the sensor 150, or the power source 160, may be operatively coupled via a bus 170. Although a single bus 170 is shown in FIG. 1, a computing device 100 may include multiple buses. For example, the memory 110, the processor 120, the user interface 130, the electronic communication unit 140, the sensor 150, and the bus 170 may receive power from the power source 160 via the bus 170. In another example, the memory 110, the processor 120, the user interface 130, the electronic communication unit 140, the sensor 150, the power source 160, or a combination thereof, may communicate data, such as by sending and receiving electronic signals, via the bus 170.

[0055] Although not shown separately in FIG. 1, one or more of the processor 120, the user interface 130, the electronic communication unit 140, the sensor 150, or the power source 160 may include internal memory, such as an internal buffer or register. For example, the processor 120 may include internal memory (not shown) and may read data 112 from the memory 110 into the internal memory (not shown) for processing.

[0056] Although shown as separate elements, the memory 110, the processor 120, the user interface 130, the electronic communication unit 140, the sensor 150, the power source 160,and the bus 170, or any combination thereof can be integrated in one or more electronic units, circuits, or chips.

[0057] FIG. 2 is a diagram of a computing and communications system 200 in accordance with implementations of this disclosure. The computing and communications system 200 shown includes computing and communication devices 100A, 100B, 100C, access points 210A, 210B, and a network 220. For example, the computing and communication system 200 can be a multiple access system that provides communication, such as voice, audio, data, video, messaging, broadcast, or a combination thereof, to one or more wired or wireless communicating devices, such as the computing and communication devices 100A, 100B, 100C. Although, for simplicity, FIG. 2 shows three computing and communication devices 100A, 100B, 100C, two access points 210A, 210B, and one network 220, any number of computing and communication devices, access points, and networks can be used.

[0058] A computing and communication device 100A, 100B, 100C can be, for example, a computing device, such as the computing device 100 shown in FIG. 1. For example, the computing and communication devices 100A, 100B may be user devices, such as a mobile computing device, a laptop, a thin client, or a smartphone, and the computing and communication device 100C may be a server, such as a mainframe or a cluster. Although the computing and communication device 100A and the computing and communication device 100B are described as user devices, and the computing and communication device 100C is described as a server, any computing and communication device may perform some or all of the functions of a server, some, or all, of the functions of a user device, or some or all of the functions of a server and a user device. For example, the server computing and communication device 100C may receive, encode, process, store, transmit, or a combination thereof audio data and one or both of the computing and communication device 100A and the computing and communication device 100B may receive, decode, process, store, present, or a combination thereof the audio data.

[0059] Each computing and communication device 100A, 100B, 100C, which may include a user equipment (UE), a mobile station, a fixed or mobile subscriber unit, a cellular telephone, a personal computer, a tablet computer, a server, consumer electronics, or any similar device, can be configured to perform wired or wireless communication, such as via the network 220. For example, the computing and communication devices 100A, 100B, 100C can be configured to transmit or receive wired or wireless communication signals. Although each computing and communication device 100A, 100B, 100C is shown as a single unit, a computing and communication device can include any number of interconnected elements.

[0060] Each access point 210A, 210B can be any type of device configured to communicate with a computing and communication device 100A, 100B, 100C, a network 220, or both via wired or wireless communication links 180A, 180B, 180C. For example, an access point 210A, 210B can include a base station, a base transceiver station (BTS), a Node- B, an enhanced Node-B (eNode-B), a Home Node-B (HNode-B), a wireless router, a wired router, a hub, a relay, a switch, or any similar wired or wireless device. Although each access point 210A, 210B is shown as a single unit, an access point can include any number of interconnected elements.

[0061] The network 220 can be any type of network configured to provide services, such as voice, data, applications, voice over internet protocol (VoIP), or any other communications protocol or combination of communications protocols, over a wired or wireless communication link. For example, the network 220 can be a local area network (FAN), wide area network (WAN), virtual private network (VPN), a mobile or cellular telephone network, the Internet, or any other means of electronic communication. The network can use a communication protocol, such as the transmission control protocol (TCP), the user datagram protocol (UDP), the internet protocol (IP), the real-time transport protocol (RTP), the HyperText Transport Protocol (HTTP), or a combination thereof.

[0062] The computing and communication devices 100A, 100B, 100C can communicate with each other via the network 220 using one or more a wired or wireless communication links, or via a combination of wired and wireless communication links. For example, as shown the computing and communication devices 100A, 100B can communicate via wireless communication links 180A, 180B, and computing and communication device 100C can communicate via a wired communication link 180C. Any of the computing and communication devices 100A, 100B, 100C may communicate using any wired or wireless communication link, or links. For example, a first computing and communication device 100A can communicate via a first access point 210A using a first type of communication link, a second computing and communication device 100B can communicate via a second access point 210B using a second type of communication link, and a third computing and communication device 100C can communicate via a third access point (not shown) using a third type of communication link. Similarly, the access points 210A, 210B can communicate with the network 220 via one or more types of wired or wireless communication links 230A, 230B. Although FIG. 2 shows the computing and communication devices 100A, 100B, 100C in communication via the network 220, the computing and communication devices 100A, 100B, 100C can communicate with each other via any number of communication links, suchas a direct wired or wireless communication link.

[0063] In some implementations, communications between one or more of the computing and communication device 100A, 100B, 100C may omit communicating via the network 220 and may include transferring data via another medium (not shown), such as a data storage device. For example, the server computing and communication device 100C may store audio data, such as encoded audio data, in a data storage device, such as a portable data storage unit, and one or both of the computing and communication device 100A or the computing and communication device 100B may access, read, or retrieve the stored audio data from the data storage unit, such as by physically disconnecting the data storage device from the server computing and communication device 100C and physically connecting the data storage device to the computing and communication device 100A or the computing and communication device 100B.

[0064] Other implementations of the computing and communications system 200 are possible. For example, in an implementation, the network 220 can be an ad-hoc network and can omit one or more of the access points 210A, 210B. The computing and communications system 200 may include devices, units, or elements not shown in FIG. 2. For example, the computing and communications system 200 may include many more communicating devices, networks, and access points.

[0065] FIG. 3 is a diagram of a video stream 300 for use in encoding and decoding in accordance with implementations of this disclosure. A video stream 300, such as a video stream captured by a video camera or a video stream generated by a computing device, may include a video sequence 310. The video sequence 310 may include a sequence of adjacent frames 320. Although three adjacent frames 320 are shown, the video sequence 310 can include any number of adjacent frames 320.

[0066] Each frame 330 from the adjacent frames 320 may represent a single image from the video stream. Although not shown in FIG. 3, a frame 330 may include one or more segments, tiles, or planes, which may be coded, or otherwise processed, independently, such as in parallel. A frame 330 may include one or more tiles 340. Each of the tiles 340 may be a rectangular region of the frame that can be coded independently. Each of the tiles 340 may include respective blocks 350. Although not shown in FIG. 3, a block can include pixels. For example, a block can include a 16x16 group of pixels, an 8x8 group of pixels, an 8x16 group of pixels, or any other group of pixels. Unless otherwise indicated herein, the term ‘block’ can include a superblock, a macroblock, a segment, a slice, or any other portion of a frame. A frame, a block, a pixel, or a combination thereof can include display information, such asluminance information, chrominance information, or any other information that can be used to store, modify, communicate, or display the video stream or a portion thereof.

[0067] FIG. 4 is a block diagram of an encoder 400 in accordance with implementations of this disclosure. Encoder 400 can be implemented in a device, such as the computing device 100 shown in FIG. 1 or the computing and communication devices 100A, 100B, 100C shown in FIG. 2, as, for example, a computer software program stored in a data storage unit, such as the memory 110 shown in FIG. 1. The computer software program can include machine instructions that may be executed by a processor, such as the processor 120 shown in FIG. 1, and may cause the device to encode video data as described herein. The encoder 400 can be implemented as specialized hardware included, for example, in computing device 100.

[0068] The encoder 400 can encode an input video stream 402, such as the video stream 300 shown in FIG. 3, to generate an encoded (compressed) bitstream 404. In some implementations, the encoder 400 may include a forward path for generating the compressed bitstream 404. The forward path may include an intra / inter prediction unit 410, a transform unit 420, a quantization unit 430, an entropy encoding unit 440, or any combination thereof. In some implementations, the encoder 400 may include a reconstruction path (indicated by the broken connection lines) to reconstruct a frame for encoding of further blocks. The reconstruction path may include a dequantization unit 450, an inverse transform unit 460, a reconstruction unit 470, a filtering unit 480, or any combination thereof. Other structural variations of the encoder 400 can be used to encode the video stream 402.

[0069] For encoding the video stream 402, each frame within the video stream 402 can be processed in units of blocks. Thus, a current block may be identified from the blocks in a frame, and the current block may be encoded.

[0070] At the intra / inter prediction unit 410, the current block can be encoded using either intra-frame prediction, which may be within a single frame, or inter-frame prediction, which may be from frame to frame. Intra-prediction may include generating a prediction block from samples in the current frame that have been previously encoded and reconstructed. Interprediction may include generating a prediction block from samples in one or more previously constructed reference frames. Generating a prediction block for a current block in a current frame may include performing motion estimation to generate a motion vector indicating an appropriate reference portion of the reference frame.

[0071] The intra / inter prediction unit 410 may subtract the prediction block from the current block (raw block) to produce a residual block. The transform unit 420 may perform a block-based transform, which may include transforming the residual block into transformcoefficients in, for example, the frequency domain. Examples of block-based transforms include the Karhunen-Loeve Transform (KLT), the Discrete Cosine Transform (DCT), the Singular Value Decomposition Transform (SVD), and the Asymmetric Discrete Sine Transform (ADST). In an example, the DCT may include transforming a block into the frequency domain. The DCT may include using transform coefficient values based on spatial frequency, with the lowest frequency (i.e., DC) coefficient at the top-left of the matrix and the highest frequency coefficient at the bottom-right of the matrix.

[0072] The quantization unit 430 may convert the transform coefficients into discrete quantum values, which may be referred to as quantized transform coefficients or quantization levels. The quantized transform coefficients can be entropy encoded by the entropy encoding unit 440 to produce entropy-encoded coefficients. Entropy encoding can include using a probability distribution metric. The entropy-encoded coefficients and information used to decode the block, which may include the type of prediction used, motion vectors, and quantizer values, can be output to the compressed bitstream 404. The compressed bitstream 404 can be formatted using various techniques, such as run-length encoding (RLE) and zerorun coding.

[0073] The reconstruction path can be used to maintain reference frame synchronization between the encoder 400 and a corresponding decoder, such as the decoder 500 shown in FIG. 5. The reconstruction path may be similar to the decoding process discussed below and may include decoding the encoded frame, or a portion thereof, which may include decoding an encoded block, which may include dequantizing the quantized transform coefficients at the dequantization unit 450 and inverse transforming the dequantized transform coefficients at the inverse transform unit 460 to produce a derivative residual block. The reconstruction unit 470 may add the prediction block generated by the intra / inter prediction unit 410 to the derivative residual block to create a decoded block. The filtering unit 480 can be applied to the decoded block to generate a reconstructed block, which may reduce distortion, such as blocking artifacts. Although one filtering unit 480 is shown in FIG. 4, filtering the decoded block may include loop filtering, deblocking filtering, or other types of filtering or combinations of types of filtering. The reconstructed block may be stored or otherwise made accessible as a reconstructed block, which may be a portion of a reference frame, for encoding another portion of the current frame, another frame, or both, as indicated by the broken line at 482. Coding information, such as deblocking threshold index values, for the frame may be encoded, included in the compressed bitstream 404, or both, as indicated by the broken line at 484.

[0074] Other variations of the encoder 400 can be used to encode the compressed bitstream 404. For example, a non-transform-based encoder 400 can quantize the residual block directly without the transform unit 420. In some implementations, the quantization unit 430 and the dequantization unit 450 may be combined into a single unit.

[0075] FIG. 5 is a block diagram of a decoder 500 in accordance with implementations of this disclosure. The decoder 500 can be implemented in a device, such as the computing device 100 shown in FIG. 1 or the computing and communication devices 100A, 100B, 100C shown in FIG. 2, as, for example, a computer software program stored in a data storage unit, such as the memory 110 shown in FIG. 1. The computer software program can include machine instructions that may be executed by a processor, such as the processor 120 shown in FIG. 1, and may cause the device to decode video data as described herein. The decoder 500 can be implemented as specialized hardware included, for example, in computing device 100.

[0076] The decoder 500 may receive a compressed bitstream 502, such as the compressed bitstream 404 shown in FIG. 4, and may decode the compressed bitstream 502 to generate an output video stream 504. The decoder 500 may include an entropy decoding unit 510, a dequantization unit 520, an inverse transform unit 530, an intra / inter prediction unit 540, a reconstruction unit 550, a filtering unit 560, or any combination thereof. Other structural variations of the decoder 500 can be used to decode the compressed bitstream 502.

[0077] The entropy decoding unit 510 may decode data elements within the compressed bitstream 502 using, for example, Context Adaptive Binary Arithmetic Decoding, to produce a set of quantized transform coefficients. The dequantization unit 520 can dequantize the quantized transform coefficients, and the inverse transform unit 530 can inverse transform the dequantized transform coefficients to produce a derivative residual block, which may correspond to the derivative residual block generated by the inverse transform unit 460 shown in FIG. 4. Using header information decoded from the compressed bitstream 502, the intra / inter prediction unit 540 may generate a prediction block corresponding to the prediction block created in the encoder 400. At the reconstruction unit 550, the prediction block can be added to the derivative residual block to create a decoded block. The filtering unit 560 can be applied to the decoded block to reduce artifacts, such as blocking artifacts, which may include loop filtering, deblocking filtering, or other types of filtering or combinations of types of filtering, and which may include generating a reconstructed block, which may be output as the output video stream 504.

[0078] Other variations of the decoder 500 can be used to decode the compressed bitstream 502. For example, the decoder 500 can produce the output video stream 504without the filtering unit 560.

[0079] FIG. 6 is a block diagram of a representation of a portion 600 of a frame, such as the frame 330 shown in FIG. 3, in accordance with implementations of this disclosure. As shown, the portion 600 of the frame includes four 64x64 blocks 610, in two rows and two columns in a matrix or Cartesian plane. In some implementations, a 64x64 block may be a maximum coding unit, N=64. Each 64x64 block may include four 32x32 blocks 620. Each 32x32 block may include four 16x16 blocks 630. Each 16x16 block may include four 8x8 blocks 640. Each 8x8 block 640 may include four 4x4 blocks 650. Each 4x4 block 650 may include 16 pixels, which may be represented in four rows and four columns in each respective block in the Cartesian plane or matrix. The pixels may include information representing an image captured in the frame, such as luminance information, color information, and location information. In some implementations, a block, such as a 16x16 pixel block as shown, may include a luminance block 660, which may include luminance pixels 662; and two chrominance blocks 670, 680, such as a U or Cb chrominance block 670, and a V or Cr chrominance block 680. The chrominance blocks 670, 680 may include chrominance pixels 690. For example, the luminance block 660 may include 16x16 luminance pixels 662 and each chrominance block 670, 680 may include 8x8 chrominance pixels 690 as shown. Although one arrangement of blocks is shown, any arrangement may be used. Although FIG.6 shows NxN blocks, in some implementations, NxM blocks may be used. For example, 32x64 blocks, 64x32 blocks, 16x32 blocks, 32x16 blocks, or any other size blocks may be used. In some implementations, Nx2N blocks, 2NxN blocks, or a combination thereof may be used.

[0080] In some implementations, video coding may include ordered block-level coding. Ordered block-level coding may include coding blocks of a frame in an order, such as rasterscan order, wherein blocks may be identified and processed starting with a block in the upper left comer of the frame, or portion of the frame, and proceeding along rows from left to right and from the top row to the bottom row, identifying each block in turn for processing. For example, the 64x64 block in the top row and left column of a frame may be the first block coded and the 64x64 block immediately to the right of the first block may be the second block coded. The second row from the top may be the second row coded, such that the 64x64 block in the left column of the second row may be coded after the 64x64 block in the rightmost column of the first row.

[0081] In some implementations, coding a block may include using quad-tree coding, which may include coding smaller block units within a block in raster- scan order. Forexample, the 64x64 block shown in the bottom left comer of the portion of the frame shown in FIG. 6, may be coded using quad-tree coding wherein the top left 32x32 block may be coded, then the top right 32x32 block may be coded, then the bottom left 32x32 block may be coded, and then the bottom right 32x32 block may be coded. Each 32x32 block may be coded using quad- tree coding wherein the top left 16x16 block may be coded, then the top right 16x16 block may be coded, then the bottom left 16x16 block may be coded, and then the bottom right 16x16 block may be coded. Each 16x16 block may be coded using quad- tree coding wherein the top left 8x8 block may be coded, then the top right 8x8 block may be coded, then the bottom left 8x8 block may be coded, and then the bottom right 8x8 block may be coded. Each 8x8 block may be coded using quad-tree coding wherein the top left 4x4 block may be coded, then the top right 4x4 block may be coded, then the bottom left 4x4 block may be coded, and then the bottom right 4x4 block may be coded. In some implementations, 8x8 blocks may be omitted for a 16x16 block, and the 16x16 block may be coded using quad-tree coding wherein the top left 4x4 block may be coded, then the other 4x4 blocks in the 16x16 block may be coded in raster- scan order.

[0082] In some implementations, video coding may include compressing the information included in an original, or input, frame by, for example, omitting some of the information in the original frame from a corresponding encoded frame. For example, coding may include reducing spectral redundancy, reducing spatial redundancy, reducing temporal redundancy, or a combination thereof.

[0083] In some implementations, reducing spectral redundancy may include using a color model based on a luminance component (Y) and two chrominance components (U and V or Cb and Cr), which may be referred to as the YUV or YCbCr color model, or color space. Using the YUV color model may include using a relatively large amount of information to represent the luminance component of a portion of a frame and using a relatively small amount of information to represent each corresponding chrominance component for the portion of the frame. For example, a portion of a frame may be represented by a high- resolution luminance component, which may include a 16x16 block of pixels, and by two lower resolution chrominance components, each of which represents the portion of the frame as an 8x8 block of pixels. A pixel may indicate a value, for example, a value in the range from 0 to 255, and may be stored or transmitted using, for example, eight bits. Although this disclosure is described in reference to the YUV color model, any color model may be used.

[0084] In some implementations, reducing spatial redundancy may include transforming a block into the frequency domain using, for example, a discrete cosine transform (DCT). Forexample, a unit of an encoder, such as the transform unit 420 shown in FIG. 4, may perform a DCT using transform coefficient values based on spatial frequency.

[0085] In some implementations, reducing temporal redundancy may include using similarities between frames to encode a frame using a relatively small amount of data based on one or more reference frames, which may be previously encoded, decoded, and reconstructed frames of the video stream. For example, a block or pixel of a current frame may be similar to a spatially corresponding block or pixel of a reference frame. In some implementations, a block or pixel of a current frame may be similar to block or pixel of a reference frame at a different spatial location and reducing temporal redundancy may include generating motion information indicating the spatial difference, or translation, between the location of the block or pixel in the current frame and corresponding location of the block or pixel in the reference frame.

[0086] In some implementations, reducing temporal redundancy may include identifying a portion of a reference frame that corresponds to a current block or pixel of a current frame. For example, a reference frame, or a portion of a reference frame, which may be stored in memory, may be searched to identify a portion for generating a prediction to use for encoding a current block or pixel of the current frame with maximal efficiency. For example, the search may identify a portion of the reference frame for which the difference in pixel values between the current block and a prediction block generated based on the portion of the reference frame is minimized and may be referred to as motion searching. In some implementations, the portion of the reference frame searched may be limited. For example, the portion of the reference frame searched, which may be referred to as the search area, may include a limited number of rows of the reference frame. In an example, identifying the portion of the reference frame for generating a prediction may include calculating a cost function, such as a sum of absolute differences (SAD), between the pixels of portions of the search area and the pixels of the current block.

[0087] In some implementations, the spatial difference between the location of the portion of the reference frame for generating a prediction in the reference frame and the current block in the current frame may be represented as a motion vector. The difference in pixel values between the prediction block and the current block may be referred to as differential data, residual data, a prediction error, or as a residual block. In some implementations, generating motion vectors may be referred to as motion estimation, and a pixel of a current block may be indicated based on location using Cartesian coordinates asfx y. Similarly, a pixel of the searcharea of the reference frame may be indicated based on location using Cartesian coordinates such as rx,y. A motion vector (MV) for the current block may be determined based on, for example, a SAD between the pixels of the current frame and the corresponding pixels of the reference frame.

[0088] Although described herein with reference to matrix or Cartesian representation of a frame for clarity, a frame may be stored, transmitted, processed, or any combination thereof, in any data structure such that pixel values may be efficiently represented for a frame or image. For example, a frame may be stored, transmitted, processed, or any combination thereof, in a two-dimensional data structure such as a matrix as shown, or in a onedimensional data structure, such as a vector array. In an implementation, a representation of the frame, such as a two-dimensional representation as shown, may correspond to a physical location in a rendering of the frame as an image. For example, a location in the top left corner of a block in the top left comer of the frame may correspond with a physical location in the top left corner of a rendering of the frame as an image.

[0089] In some implementations, block-based coding efficiency may be improved by partitioning input blocks into one or more prediction partitions, which may be rectangular, including square, partitions for prediction coding. In some implementations, video coding using prediction partitioning may include selecting a prediction partitioning scheme from among multiple candidate prediction partitioning schemes. For example, in some implementations, candidate prediction partitioning schemes for a 64x64 coding unit may include rectangular size prediction partitions ranging in sizes from 4x4 to 64x64, such as 4x4, 4x8, 8x4, 8x8, 8x16, 16x8, 16x16, 16x32, 32x16, 32x32, 32x64, 64x32, or 64x64. In some implementations, video coding using prediction partitioning may include a full prediction partition search, which may include selecting a prediction partitioning scheme by encoding the coding unit using each available candidate prediction partitioning scheme and selecting the best scheme, such as the scheme that produces the least rate-distortion error.

[0090] In some implementations, encoding a video frame may include identifying a prediction partitioning scheme for encoding a current block, such as block 610. In some implementations, identifying a prediction partitioning scheme may include determining whether to encode the block as a single prediction partition of maximum coding unit size, which may be 64x64 as shown, or to partition the block into multiple prediction partitions, which may correspond with the sub-blocks, such as the 32x32 blocks 620 the 16x16 blocks 630, or the 8x8 blocks 640, as shown, and may include determining whether to partition intoone or more smaller prediction partitions. For example, a 64x64 block may be partitioned into four 32x32 prediction partitions. Three of the four 32x32 prediction partitions may be encoded as 32x32 prediction partitions and the fourth 32x32 prediction partition may be further partitioned into four 16x16 prediction partitions. Three of the four 16x16 prediction partitions may be encoded as 16x16 prediction partitions and the fourth 16x16 prediction partition may be further partitioned into four 8x8 prediction partitions, each of which may be encoded as an 8x8 prediction partition. In some implementations, identifying the prediction partitioning scheme may include using a prediction partitioning decision tree.

[0091] In some implementations, video coding for a current block may include identifying an optimal prediction coding mode from multiple candidate prediction coding modes, which may provide flexibility in handling video signals with various statistical properties and may improve the compression efficiency. For example, a video coder may evaluate each candidate prediction coding mode to identify the optimal prediction coding mode, which may be, for example, the prediction coding mode that minimizes an error metric, such as a rate-distortion cost, for the current block. In some implementations, the complexity of searching the candidate prediction coding modes may be reduced by limiting the set of available candidate prediction coding modes based on similarities between the current block and a corresponding prediction block. In some implementations, the complexity of searching each candidate prediction coding mode may be reduced by performing a directed refinement mode search. For example, metrics may be generated for a limited set of candidate block sizes, such as 16x16, 8x8, and 4x4, the error metric associated with each block size may be in descending order, and additional candidate block sizes, such as 4x8 and 8x4 block sizes, may be evaluated.

[0092] In some implementations, block-based coding efficiency may be improved by partitioning a current residual block into one or more transform partitions, which may be rectangular, including square, partitions for transform coding. In some implementations, video coding, such as video coding using transform partitioning, may include selecting a uniform transform partitioning scheme. For example, a current residual block, such as block 610, may be a 64x64 block and may be transformed without partitioning using a 64x64 transform.

[0093] Although not expressly shown in FIG. 6, a residual block may be transform partitioned using a uniform transform partitioning scheme. For example, a 64x64 residual block may be transform partitioned using a uniform transform partitioning scheme including four 32x32 transform blocks, using a uniform transform partitioning scheme includingsixteen 16x16 transform blocks, using a uniform transform partitioning scheme including sixty-four 8x8 transform blocks, or using a uniform transform partitioning scheme including 256 4x4 transform blocks.

[0094] In some implementations, video coding, such as video coding using transform partitioning, may include identifying multiple transform block sizes for a residual block using multiform transform partition coding. In some implementations, multiform transform partition coding may include recursively determining whether to transform a current block using a current block size transform or by partitioning the current block and multiform transform partition coding each partition. For example, the bottom left block 610 shown in FIG. 6 may be a 64x64 residual block, and multiform transform partition coding may include determining whether to code the current 64x64 residual block using a 64x64 transform or to code the 64x64 residual block by partitioning the 64x64 residual block into partitions, such as four 32x32 blocks 620, and multiform transform partition coding each partition. In some implementations, determining whether to transform partition the current block may be based on comparing a cost for encoding the current block using a current block size transform to a sum of costs for encoding each partition using partition size transforms.

[0095] As referenced briefly above, non-separable secondary transforms may be used to improve energy compactness of a set of primary transform coefficients of respective transform blocks. Generally, a non-separable secondary transform may be applied to some or all primary transform coefficients of a transform block (e.g., applied to the top-left comer of the transform coefficients matrix). One non-separable secondary transform may be selected and signaled per transform block by an encoder from a set of different predefined secondary transform kernels. The one non-separable secondary transform may be selected using ratedistortion optimization (RDO) or some other technique that balances quality with bitrate.

[0096] A wide diversity of secondary transform kernels allows improved energy compactness for a variety of input signals. The trade-offs for this energy compactness are relatively high memory requirements and relatively complex calculations.

[0097] The framework described herein defines multiple secondary transforms that can be performed on top of different subsets of the primary transform coefficients. This technique allows for the reduction in the dimensions of secondary transforms while keeping a high flexibility to compress a variety of input signals. First described is an example, which is followed by a general process for reducing kernel dimensions according to the teachings herein.

[0098] Secondary transforms can be applied to different portions of primary transformcoefficients depending on the size and characteristics of the transform block. Primary transform coefficients, as used herein, refers to the transform coefficients resulting from applying the primary transforms to a residual block. The residual block may be of various sizes, such as 4x4, 4x8, 8x4, 8x8, or larger blocks of pixel values, or may be a larger residual block partitioned into smaller transform blocks.

[0099] In some implementations, the secondary transform processing depends on the transform block size. For transform blocks larger than a threshold size (e.g., 8x8 blocks or larger), secondary transforms process all coefficients within the block (e.g., all 64 coefficients in an 8x8 region). For smaller transform blocks such as 4xN or Nx4 blocks (where N > 4), secondary transforms are configured to process only a subset of coefficients. For example, given a 4x8 transform block (32 coefficients total), a secondary transform processes only the top-left 4x4 region (16 coefficients) of the primary transform coefficients. The 4x4 non- separable secondary transform uses the 16 primary transform coefficients as input and outputs a reduced number of secondary transform coefficients, achieving improved energy compactness.

[0100] If the kernel dimension of the secondary transforms is instead 64x32, for example, all primary transform coefficients of an 8x8 transform block may be input to a 64x32 non- separable secondary transform to output 32 secondary transform coefficients. Desirably, the secondary transform coefficients that are output have a higher compactness of energy as compared to the primary transform coefficients. Moreover, the number of coefficients to be coded into a bitstream are reduced. The cardinality of secondary transforms available may be 42 in some implementations. The cardinality of secondary transforms may be split into groups depending on one or more block characteristics such as the block size, prediction mode, primary transform kernel, etc. In an example, 42 secondary transform kernels are available in two groups of 21 secondary transform kernels, and which group is used for a particular set of primary transform coefficients depends on the primary transform kernel. The memory requirement for the kernels is 42*64*32 for the 64x32 kernels. Accordingly, both memory requirements and computational complexity are increased over the previous example. However, compression performance is improved in most cases.

[0101] It is desirable to reduce the kernel dimensions to reduce memory requirements and the complexity of quantization calculations but to maximize compression performance while doing so.

[0102] In the example where a kernel dimension of the secondary transforms is 64x32, it may be desirable to reduce the kernel dimensions from 64x32 to Kx32, where K<64. Thiswill reduce memory requirements and the complexity of quantization calculations.

[0103] Initially, a framework for reduced kernel dimensions is determined by defining a set of C support regions. Each support region is indexed by i = 1, 2, ..., C, and each is a proper subset of a region (here, an 8x8 region) with size K (i.e., having K pixels). Thereafter, a number N_i of secondary transforms of the dimension Kx32 that is applied to the i-th support region is defined. In particular, such a secondary transform takes in the K coefficients within the support region and produces 32 output coefficients. The result is a framework of N_1 + N_2 + ... + N_C secondary transform kernels of size Kx32.

[0104] This may be seen with reference to FIG. 7, which shows an example of three support regions (i.e., C = 3), each having a size Ki of 44 pixels. In FIG. 7, the support regions are shaded within each 8x8 block and form respective proper subsets of the 8x8 block. The first support region 702 may be suitable for coefficient blocks where the energy is mostly distributed in the left side of the block. The second support region 704 may be suitable where most energy is distributed on the top. The third support region 706 may be suitable where energy is present in both the top right and the bottom left corners.

[0105] To illustrate, it is possible to design 14 secondary transforms for each support region (7 per primary transform type), which leads to the same cardinality of secondary transform kernels as originally provided (i.e., 21 transform kernels per primary transform type) of a reduced size 44x32. In this example, the memory requirement for the same number of kernels is reduced by 31.25% (42*44*32 according to the teachings herein compared to 42*64*32 for the 64x32 kernels). A 21 -ary symbol may still be used, although it is not required to signal (e.g., an index for) the secondary transform for a block. Thus, the excess bandwidth associated with transmitting the index for a smaller set of secondary transforms can be used to signal, in an encoded bitstream, the index for the support region, the size K for the block (e.g., where more than one value for the size K is available), or some combination thereof. In the example shown and described with respect to FIG. 7, the encoded bitstream may receive a first ternary symbol identifying the support region and a 7-ary symbol identifying the actual transform type for that support region. In general, a part of the signaling space is used to determine a reduced set of transform coefficients for the second transform, and remaining signaling space can be used to determine the actual secondary transform type (or kernel).

[0106] In some implementations, the size K may be defined a priori (e.g., based on transform block size) and only the index is signaled. In some implementations, multiple values for the size K may be available, and an index is signaled instead of the size K (e.g.,along with the index for the support region and an index for the secondary transform). In some implementations, one or both of the index for the support region or the size K for a block may not be signaled and instead may be determined by the decoder from other information within the bitstream, such as the block size, the transform type, the color plane, etc.

[0107] These values above are described by example only. Other combinations may be used. For example, size K may be 40 such that the transform kernels for an 8x8 block are 40x32. Other block sizes may be used, e.g., to match the original transform block sizes.

[0108] In some implementations, the support regions can be designed using data-driven approaches. For example, one may obtain 8x8 regions of primary transform coefficients as training samples, classify them based on their energy distributions, and update the support regions based on the energy distributions of the samples within each class.

[0109] For example, an embodiment may define a different support region for each secondary transform, in which case C=M and N_i=l for all i. When M=42, this means that there are 42 secondary transform kernels, each of which uses a different support region. One can select a proper number of K that maximally reduces the kernel dimension while keeping the flexibility of the transforms.

[0110] FIG. 8 is a diagram of three further support regions for performing a secondary transform in accordance with implementations of this disclosure. There are three support regions (i.e., C = 3), each having a size K_i of 40 pixels. These support regions were developed through training using video samples from training libraries. Multiple support regions were tested for the region size, and the three regions resulting in the best compression were maintained. The support regions are unshaded within each 8x8 block and form respective proper subsets of the 8x8 block. The first support region 802 is suitable for coefficient blocks where the energy is present in both the top right and the bottom left comers. The second support region 804 is suitable where most energy is distributed in the left side of the block. The third support region 806 can capture a case where the most energy is distributed at the top side of the block.

[0111] FIG. 9 is a flowchart of a technique 900 of reconstructing a block using secondary transforms with flexible coefficient support region in accordance with implementations of this disclosure. The technique 900 may be implemented in a reconstruction path of an encoder, such as the encoder 400 shown in FIG. 4, or in a decoder, such as the decoder 500 shown in FIG. 5.

[0112] The technique 900 includes determining secondary transform coefficients for acurrent block of image data at operation 902. For example, at a decoder, determining the secondary transform coefficients can include decoding, from an encoded bitstream, the secondary transform coefficients.

[0113] At operation 904, the technique 900 determines an index identifying a support region of the current block, wherein the support region comprises a proper subset of positions of the current block. Note that operation 904 may be performed before operation 902 where, for example, the index is encoded into a header associated with the current block, such as a block, slice, or frame header.

[0114] At operation 906, a secondary transform for the current block is determined. The secondary transform may be determined by decoding an identifier of the secondary transform from the encoded bitstream, such as from the header described above in some implementations, and then using that identifier (index, etc.) to access a list of possible secondary transforms known to each of the encoder and decoder. The secondary transform may be a non- separable transform, and the primary transform may be or include one or more separable transforms.

[0115] In some implementations, determining the secondary transform for the current block includes using the index identifying the support region in a lookup table to identify the secondary transform. The lookup table may associate each support region index with one or more available secondary transforms for that support region. For example, the lookup table may be implemented as a two-dimensional array where the first dimension corresponds to the support region index, and the second dimension corresponds to the secondary transform identifier within that region. Alternatively, the lookup table may be implemented as a hierarchical structure where the support region index is used to select a subset of available secondary transforms, and the secondary transform identifier is then used to select the specific transform from that subset.

[0116] At operation 908, the secondary transform is applied to the secondary transform coefficients to obtain primary transform coefficients, more specifically, the inverse of the secondary transform is applied. This may also be referred to as an inverse transformation, although the domain is not changing from the frequency domain to the pixel domain. That is, the inverse transformation occurs in the frequency domain.

[0117] At operation 910, the primary transform coefficients are arranged according to the support region of the current block. That is, knowing the scan order and a template representing the support region identified by the index of operation 904, positions of the current block within the support region are filled with the primary transform coefficients.Remaining positions may be filled with the value 0.

[0118] At operation 912, a primary transform for the current block is determined. The primary transform may be determined by decoding an identifier of the primary transform from the encoded bitstream, such as from the header described above in some implementations, and then using that identifier (index, etc.) to access a list of possible primary transforms known to each of the encoder and decoder. In some implementations, the primary transform may be determined from other data within the bitstream related to the current block, such as the prediction mode, the size, the color plane, etc., or some combination thereof.

[0119] At operation 914, the primary transform coefficients are inversed transformed using the primary transform to obtain a residual block of pixels (e.g., in the pixel domain) corresponding to the current block. Optionally, where the primary transform coefficients are quantized transform coefficients (e.g., the primary transform coefficients were quantized before the secondary transform was applied during encoding), the primary transform coefficients may be dequantized before inverse transformation. In some implementations, it is possible that the second transform coefficients were quantized in addition to or instead of the primary transform coefficients. In such implementations, the second transform coefficients may be dequantized before applying the secondary transform.

[0120] At operation 916, the current block is reconstructed using the residual block as described above with regards to the decoder 500 and the reconstruction path of the encoder 400.

[0121] The support region may be one of multiple available support regions. At least some of the multiple available support regions have a same cardinality of positions and form a set of support regions having a respective different proper subset of positions of the current block. In some implementations, the multiple available support regions include support regions having different cardinalities of positions.

[0122] A kernel dimension of the secondary transform may be generalized as KxM, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, and the support region is one of a set of support regions, each having K positions arranged different from each other. M and K are integers, as is C described previously.

[0123] An encoder can substantially perform the inverse of the technique 900 except that the encoder must select the primary transform, support region, and secondary transform, along with conventional prediction modes, etc. Any technique can be used for the selection,including but not limited to RDO.

[0124] In some implementations, the application of secondary transforms may be controlled at different signalling levels within the encoded bitstream. For example, a syntax element which may be coded at a group of blocks level may indicate that secondary transforms are to be applied for each block within the group of blocks. In such cases, where C represents the number of available support regions, the index identifying the support region can include a value (e.g., signal) that ranges from 0 to C-l, where each value corresponds to a different support region. As such, in the case that C=3, the index can be or include a 3-ary signal.

[0125] On the other hand, if the codec performs such a determination at the block level, then the index can range from 0 to C, where the value 0 indicates that no secondary transform is applied to the block and values 1 to C correspond to the C available support regions. As such, in the case that C=3, the index can be or include a 4-ary signal. The group of blocks level syntax may indicate that applying the secondary transform is a block-level decision (e.g., encoded in the block header). The group of blocks may be a superblock, a tile, a segment, or a frame.

[0126] In some implementations, and as already described, the index includes a 4-ary signal and a 7-ary signal. When secondary transform is enabled, three values of the 4-ary signal indicate different support regions, and the 7-ary signal indicates the specific secondary transform within the selected support region. When the 4-ary signal indicates that no secondary transform is used for coding the block, the 7-ary signal may be omitted from the bitstream.

[0127] In some implementations, where there are N secondary transforms available for a given support region, the identifier of the secondary transform may have values ranging from 0 to N-l. For example, where N=7 secondary transforms are available for a particular support region, the identifier may have values 0 through 6. In some implementations, different support regions may have different numbers of available secondary transforms. For example, a first support region may have Ni available secondary transforms while a second support region may have N2 available secondary transforms, where Ni N2. Accordingly, the number of bits required to encode the identifier of the secondary transform may depend on the index of the support region, with regions having more available transforms requiring more bits for signalling.

[0128] In some implementations, the available support regions and secondary transforms may depend on characteristics of the current block, such as block size, prediction mode, or- l-primary transform type. For example, different sets of support regions may be available for intra-predicted versus inter-predicted blocks.

[0129] In some implementations, the support regions may have different sizes (i.e., different numbers of coefficient positions). For example, a first support region may comprise Ki positions while a second support region may comprise K2 positions, where Ki K2. Additionally, the support regions need not comprise contiguous sets of coefficient positions within the transform block. For example, a support region may include non-adjacent coefficient positions, creating gaps or holes within the support region pattern, thereby allowing for more flexible adaptation to different energy distribution patterns in the transform coefficients.

[0130] FIG. 10 is a flowchart of a technique 1000 of reconstructing a block using secondary transforms with flexible coefficient support region in accordance with implementations of this disclosure. The technique 1000 represents a variant on some aspects of the technique 900 of FIG. 9, particularly in how the support region and secondary transform are determined. Like technique 900, the technique 1000 may be implemented in a reconstruction path of an encoder, such as the encoder 400 shown in FIG. 4, or in a decoder, such as the decoder 500 shown in FIG. 5.

[0131] At operation 1002, secondary transform coefficients are determined for a current block of image data. For example, at a decoder, determining the secondary transform coefficients can include decoding, from an encoded bitstream, the secondary transform coefficients.

[0132] At operation 1004, a secondary transform index is determined from the encoded bitstream. The secondary transform index identifies which secondary transform kernel is to be applied to the current block and implicitly identifies the corresponding support region through a predetermined coupling. The index is decoded from the encoded bitstream, such as from a header associated with the current block, slice, or frame header.

[0133] At operation 1006, a support region for the current block is derived from the secondary transform index determined at operation 1004. Unlike the technique 900 where the support region may be independently signaled, in the technique 1000 the support region is derived using a predetermined mapping or lookup table that associates each secondary transform index with a corresponding support region pattern. This approach reduces signaling overhead by coupling the secondary transform selection with the support region determination. The lookup table may associate each secondary transform index with a specific support region pattern, such that once the secondary transform index is known, thecorresponding support region is deterministically derived.

[0134] The lookup table may be configured with a predetermined coupling between secondary transform kernels and their corresponding support regions. Each secondary transform index signaled in the bitstream simultaneously identifies both the specific secondary transform kernel to be used and the specific support region where that kernel should be applied. The lookup table may be organized based on the type of primary transform applied to the current block. To illustrate, different lookup table mappings may be used when the primary transform is a DCT_DCT transform versus when the primary transform is an ADST_ADST transform or some other primary transform type.

[0135] In some implementations, the lookup table may internally organize secondary transforms using hierarchical identifiers (e.g., stx_set and stx_type parameters). A single index value determines both the secondary transform kernel and its associated support region simultaneously. For example, when the primary transform is DCT_DCT, index value A may specify both a particular secondary transform kernel and support region R2, while index value B may specify a different secondary transform kernel and support region R3. This coupling ensures that secondary transform kernels are always applied to their intended support region patterns.

[0136] Operations 1008 through 1016 proceed similarly to operations 908 through 916 of technique 900. Specifically, at operation 1010, the primary transform coefficients are arranged according to the derived support region of the current block. At operation 1012, a primary transform for the current block is determined. At operation 1014, the primary transform coefficients are inverse transformed using the primary transform to obtain a residual block of pixels corresponding to the current block. At operation 1016, the current block is reconstructed using the residual block.

[0137] The coupled determination of secondary transform and support region in technique 1000 reflects implementations where these elements are inherently linked through predetermined associations, while the technique 900 provides a more general framework where these elements may be determined independently or through alternative signaling approaches.

[0138] Accordingly, the technique 1000 determines secondary transform coefficients for a current block of image data; determines a secondary transform index from an encoded bitstream, wherein the secondary transform index identifies a secondary transform kernel and a corresponding support region; applies an inverse of the secondary transform kernel identified by the secondary transform index to the secondary transform coefficients to obtainprimary transform coefficients; arranges the primary transform coefficients according to the support region of the current block; determines a primary transform for the current block; inverse transforms the primary transform coefficients using the primary transform to obtain a residual block of pixels corresponding to the current block; and reconstructs the current block using the residual block. The support region for the current block is derived from the secondary transform index. Deriving the support region includes using a lookup table that associates the secondary transform index with a specific support region pattern. The lookup table is organized based on a type of primary transform applied to the current block, and different lookup table mappings are used for different primary transform types.

[0139] FIG. 11 is a flowchart of a technique 1100 of encoding a current block using secondary transforms with flexible coefficient support region in accordance with implementations of this disclosure. The technique 1100 may be implemented in an encoder, such as the encoder 400 shown in FIG. 4.

[0140] At 1102, a residual block of pixels corresponding to a current block is transformed into the frequency domain using a primary transform to obtain primary transform coefficients. The residual block may be generated by subtracting a prediction block from the current block. At operation 1104, a support region is selected for the current block. The support region consists of a proper subset of positions of the current block.

[0141] At operation 1106, a secondary transform for the current block is selected. The secondary transform may be selected using various techniques including rate-distortion optimization, where multiple candidate secondary transforms are evaluated and the transform that provides the best rate-distortion performance is selected. In some implementations, the secondary transform may be selected based on a distribution of non-zero coefficients in the primary transform coefficients, with different transforms being more suitable for different sparsity patterns. The selection may also be based on characteristics of the current block such as the prediction mode (e.g., intra prediction mode, inter prediction mode), block size, or primary transform type. In some implementations, the technique 1100 may jointly select the support region and the secondary transform.

[0142] At operation 1108, the secondary transform is applied to primary transform coefficients within the support region of the current block to obtain secondary transform coefficients. A kernel dimension of the secondary transform may be MxK, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, and the support region is one of a set of support regions, each support region having K positions arranged different from each other. The secondary transform may be a non-separable transform, and the primary transform may be separable transforms.

[0143] At operation 1110, at least some of the secondary transform coefficients are encoded in a compressed bitstream, such as the compressed bitstream 404 of FIG. 4. Encoding the at least some of the secondary transform coefficients may include quantizing the secondary transform coefficients before encoding them into the encoded bitstream. For example, in some implementations, only a subset of the secondary transform coefficients are encoded in the encoded bitstream, while remaining coefficients are set to zero or otherwise omitted from transmission. For example, only the first P coefficients (where P < total number of secondary transform coefficients) may be encoded, with higher-frequency coefficients being zeroed out. Alternatively, coefficients below a certain magnitude threshold may be omitted from encoding. In another implementation, a predetermined number of coefficients in a specific scan order may be retained while others are discarded. This selective encoding reduces the bitrate required for transmission while maintaining compression efficiency.

[0144] At operation 1112, an index identifying the support region is encoded within the encoded bitstream. The index may be encoded using various techniques such as fixed-length coding, variable-length coding, or entropy coding depending on the number of available support regions.

[0145] As described above, the application of secondary transforms may be controlled at different signaling levels. For example, a syntax element which may be coded at a group of blocks level may indicate that secondary transforms are to be applied for each block within the group of blocks. In such cases, where C represents the number of available support regions, the index identifying the support region can range from 0 to C-l, where each value corresponds to a different support region. On the other hand, if the codec performs such a determination at the block level, then the index can range from 0 to C, where the value 0 indicates that no secondary transform is applied to the block and values 1 to C correspond to the C available support regions. For example, the group of blocks level syntax may indicate that applying the secondary transform is a block-level decision (e.g., encoded in the block header). The group of blocks may be a superblock, a tile, a segment, or a frame.

[0146] At operation 1114, an identifier of the secondary transform is encoded within the encoded bitstream. In some implementations, where there are N secondary transforms available for a given support region, the identifier of the secondary transform may have values ranging from 0 to N-l. In some implementations, different support regions may have different numbers of available secondary transforms. For example, a first support region may have Ni available secondary transforms while a second support region may have N2 availablesecondary transforms, where Ni N2. Accordingly, the number of bits required to encode the identifier of the secondary transform may depend on the index of the support region, with regions having more available transforms requiring more bits for signaling.

[0147] In an alternative implementation, encoding using secondary transforms may include transforming a residual block of pixels corresponding to a current block using a primary transform to obtain primary transform coefficients; applying a secondary transform to at least some of the primary transform coefficients within a support region of the current block to obtain secondary transform coefficients, wherein the support region includes a proper subset of positions of the current block; transmitting, within an encoded bitstream, the secondary transform coefficients; and transmitting, within the encoded bitstream, an index identifying the support region of the current block. In some implementations, a defined number of the secondary transform coefficients may be set to zero before transmitting the secondary transform coefficients.

[0148] The secondary transform may include a non- separable transform, the primary transform may include separable transforms, and the index may identifies the secondary transform. The primary transform coefficients may be quantized before applying the secondary transform. A kernel dimension of the secondary transform may be KxM, where K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, the support region is one of a set of support regions, and each support region can have K positions arranged different from each other.

[0149] While not specifically shown, in some implementations of the technique 1100, and as described with respect to the technique 1000 of FIG. 10, the index identifying the support region and the identifier of the secondary transform may be combined into a single composite index that is transmitted in the encoded bitstream. In such implementations, the encoder may encode both the support region information and the secondary transform selection as a unified signaling element, and the decoder may decode this composite index to simultaneously determine both the support region and the secondary transform to be applied. This approach can provide signaling efficiency by reducing the number of separate syntax elements that need to be transmitted, while still maintaining the flexibility to independently select support regions and secondary transforms during the encoding process.

[0150] For simplicity of explanation, the technique 900, the technique 1000, and the technique 1100 is each depicted and described as a respective series of steps or operations. However, the steps or operations in accordance with this disclosure can occur in various orders and / or concurrently. Additionally, other steps or operations not presented anddescribed herein may be used. Furthermore, not all illustrated steps or operations may be required to implement a technique in accordance with the disclosed subject matter.

[0151] As used herein, the terms “optimal”, “optimized”, “optimization”, or other forms thereof, are relative to a respective context and are not indicative of absolute theoretic optimization unless expressly specified herein.

[0152] As used herein, the term “set” indicates a distinguishable collection or grouping of zero or more distinct elements or members that may be represented as a one-dimensional array or vector, except as expressly described herein or otherwise clear from context.

[0153] The words “example”, “aspect”, or the like are used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein using these terms are not necessarily to be construed as preferred or advantageous over other aspects or designs. Rather, use of these words is intended to present concepts in a concrete fashion. As used in this application, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or”. That is, unless specified otherwise, or clear from context, “X includes A or B” is intended to mean any of the natural inclusive permutations. That is, if X includes A; X includes B; or X includes both A and B, then “X includes A or B” is satisfied under any of the foregoing instances. In addition, the articles “a” and “an” as used in this application and the appended claims should generally be construed to mean “one or more” unless specified otherwise or clear from context to be directed to a singular form. Moreover, use of the term “an embodiment” or “one embodiment” or “an implementation” or “one implementation” throughout is not intended to mean the same embodiment or implementation unless described as such. As used herein, the terms “determine” and “identify”, or any variations thereof, includes selecting, ascertaining, computing, looking up, receiving, determining, establishing, obtaining, or otherwise identifying or determining in any manner whatsoever using one or more of the devices shown in FIG. 1.

[0154] Further, for simplicity of explanation, although the figures and descriptions herein may include sequences or series of steps or stages, elements of the methods disclosed herein can occur in various orders and / or concurrently. Additionally, elements of the methods disclosed herein may occur with other elements not explicitly presented and described herein. Furthermore, one or more elements of the methods described herein may be omitted from implementations of methods in accordance with the disclosed subject matter.

[0155] The implementations of the transmitting computing and communication device 100A and / or the receiving computing and communication device 100B (and the algorithms, methods, instructions, etc. stored thereon and / or executed thereby) can be realized inhardware, software, or any combination thereof. The hardware can include, for example, computers, intellectual property (IP) cores, application-specific integrated circuits (ASICs), programmable logic arrays, optical processors, programmable logic controllers, microcode, microcontrollers, servers, microprocessors, digital signal processors or any other suitable circuit. In the claims, the term “processor” should be understood as encompassing any of the foregoing hardware, either singly or in combination. The terms “signal” and “data” are used interchangeably. Further, portions of the transmitting computing and communication device 100A and the receiving computing and communication device 100B do not necessarily have to be implemented in the same manner.

[0156] Further, in one implementation, for example, the transmitting computing and communication device 100A or the receiving computing and communication device 100B can be implemented using a computer program that, when executed, carries out any of the respective methods, algorithms and / or instructions described herein. In addition, or alternatively, for example, a special purpose computer / processor can be utilized which can contain specialized hardware for carrying out any of the methods, algorithms, or instructions described herein.

[0157] The transmitting computing and communication device 100A and receiving computing and communication device 100B can, for example, be implemented on computers in a real-time video system. Alternatively, the transmitting computing and communication device 100A can be implemented on a server and the receiving computing and communication device 100B can be implemented on a device separate from the server, such as a hand-held communications device. In this instance, the transmitting computing and communication device 100A can encode content using an encoder 400 into an encoded video signal and transmit the encoded video signal to the communications device. In turn, the communications device can then decode the encoded video signal using a decoder 500. Alternatively, the communications device can decode content stored locally on the communications device, for example, content that was not transmitted by the transmitting computing and communication device 100A. Other suitable transmitting computing and communication device 100A and receiving computing and communication device 100B implementation schemes are available. For example, the receiving computing and communication device 100B can be a generally stationary personal computer rather than a portable communications device and / or a device including an encoder 400 may also include a decoder 500.

[0158] Further, all or a portion of implementations can take the form of a computerprogram product accessible from, for example, a tangible computer-usable or computer- readable medium. A computer-usable or computer-readable medium can be any device that can, for example, tangibly contain, store, communicate, or transport the program for use by or in connection with any processor. The medium can be, for example, an electronic, magnetic, optical, electromagnetic, or a semiconductor device. Other suitable mediums are also available.

[0159] It will be appreciated that aspects can be implemented in any convenient form. For example, aspects may be implemented by appropriate computer programs which may be carried on appropriate carrier media which may be tangible carrier media (e.g., disks) or intangible carrier media (e.g. communications signals). Aspects may also be implemented using suitable apparatus which may take the form of programmable computers running computer programs arranged to implement the methods and / or techniques disclosed herein. Aspects can be combined such that features described in the context of one aspect may be implemented in another aspect.

[0160] The above-described implementations have been described to allow easy understanding of the application are not limiting. On the contrary, the application covers various modifications and equivalent arrangements included within the scope of the appended claims, which scope is to be accorded the broadest interpretation to encompass all such modifications and equivalent structure as is permitted under the law.

Claims

What is claimed is:

1. A method, comprising: determining secondary transform coefficients for a current block of image data; determining an index identifying a support region of the current block, wherein the support region comprises a proper subset of positions of the current block; applying an inverse of a secondary transform to the secondary transform coefficients to obtain primary transform coefficients; arranging the primary transform coefficients according to the support region of the current block; determining a primary transform for the current block; inverse transforming the primary transform coefficients using the primary transform to obtain a residual block of pixels corresponding to the current block; and reconstructing the current block using the residual block.

2. The method of claim 1, comprising: determining the secondary transform for the current block by using the index in a lookup table to identify the secondary transform.

3. The method of claim 1 or claim 2, wherein determining the secondary transform coefficients comprises decoding, from an encoded bitstream, the secondary transform coefficients.

4. The method of any one of claims 1 to 3, wherein the secondary transform comprises a non-separable transform, and the primary transform comprises separable transforms.

5. The method of any one of claims 1 to 4, comprising: dequantizing the primary transform coefficients before the inverse transforming, in a case that the primary transform coefficients are quantized coefficients.

6. The method of any one of claims 1 to 5, wherein the support region is one of multiple available support regions, at least some of the multiple available support regions having a same cardinality of positions and forming a set of support regions having a respective different proper subset of positions of the current block.

7. The method of claim 6, wherein the multiple available support regions include support regions having different cardinalities of positions.

8. The method of any one of claims 1 to 5, wherein a kernel dimension of the secondary transform is KxM, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, and the support region is one of a set of support regions, each support region having K positions arranged different from each other.

9. The method of any one of claims 1 to 8, wherein the index comprises a 4-ary signal and a 7-ary signal, and one available value of the 4-ary signal indicates that no secondary transform is used for coding a block.

10. A method, comprising: determining secondary transform coefficients for a current block of image data; determining a secondary transform index from an encoded bitstream, wherein the secondary transform index identifies a secondary transform kernel and a corresponding support region; applying an inverse of the secondary transform kernel identified by the secondary transform index to the secondary transform coefficients to obtain primary transform coefficients; arranging the primary transform coefficients according to the support region of the current block; determining a primary transform for the current block; inverse transforming the primary transform coefficients using the primary transform to obtain a residual block of pixels corresponding to the current block; and reconstructing the current block using the residual block.

11. The method of claim 10, further comprising: deriving the support region for the current block from the secondary transform index.

12. The method of claim 11, wherein deriving the support region comprises: using a lookup table that associates the secondary transform index with a specific support region pattern.

13. The method of claim 12, wherein the lookup table is organized based on a type of primary transform applied to the current block, and different lookup table mappings are used for different primary transform types.

14. A method, comprising: transforming a residual block of pixels corresponding to a current block using a primary transform to obtain primary transform coefficients; applying a secondary transform to at least some of the primary transform coefficients within a support region of the current block to obtain secondary transform coefficients, wherein the support region comprises a proper subset of positions of the current block; transmitting, within an encoded bitstream, the secondary transform coefficients; and transmitting, within the encoded bitstream, an index identifying the support region of the current block.

15. The method of claim 14, wherein the secondary transform comprises a non-separable transform, the primary transform comprises separable transforms, the index identifies the secondary transform, and the method comprises quantizing the primary transform coefficients before applying the secondary transform.

16. The method of claim 14 or claim 11, wherein a kernel dimension of the secondary transform is KxM, wherein K is equal to a cardinality of the proper subset of positions of the current block, M is less than or equal to K, the support region is one of a set of support regions, each support region having K positions arranged different from each other.

17. The method of any one of claim 14 to claim 12, comprising: setting a defined number of the secondary transform coefficients to zero before transmitting the secondary transform coefficients.

18. An apparatus, comprising: a processor configured to perform the method of any one of claim 1 to claim 17.

19. A non-transitory computer-readable medium storing instructions that, when executed, cause a processor to perform the method of any one of claim 1 to claim 17.

20. A non-transitory computer-readable medium storing an encoded bitstream that includes encoded secondary transform coefficients for a current block of image data, an index identifying a support region of the current block, wherein the support region comprises a proper subset of positions of the current block, and an identifier of a primary transform for the current block to be applied to primary transform coefficients arranged according to the support region of the current block and resulting from applying an inverse of a secondary transform to the secondary transform coefficients, wherein the primary transform is used for inverse transforming the primary transform coefficients to obtain a residual block of pixels corresponding to the current block21. The non-transitory computer-readable medium of claim 20, wherein the index identifies the secondary transform.

Images