Image encoding / decoding method, method for transmitting bitstream, and recording medium storing bitstream thereon

Abstract

An image encoding / decoding method, a bit stream transmission method, and a computer-readable recording medium on which a bit stream is stored are provided. According to the present disclosure, an image decoding method is a method by which an image decoding apparatus decodes an image, and comprises the steps of: acquiring first information indicating whether at least one output picture corresponding to at least one input picture is generated from a post neural network filter characteristic (NNPFC) supplemental enhancement information (SEI) message; and acquiring, from a post-neural network filter activation (NNPFA) SEI message, second information indicating whether the generated output picture is output, in which a value of the second information is determined based on whether a post-neural network filter (NNPF) target is picture rate up-sampling or whether the number of input pictures is multiple.

Description

TECHNICAL FIELD

[0001] The disclosure relates to an image encoding / decoding method, a method of transmitting a bitstream, and a recording medium storing a bitstream, and more particularly, to a method of determining whether a picture is output. BACKGROUND

[0002] Recently, the demand for high-resolution and high-quality images such as high-definition (HD) images and ultra-high-definition (UHD) images is increasing in various fields. As the resolution and quality of image data are improved, the amount of information or bits transmitted is relatively increased compared to existing image data. The increase in the amount of information or bits transmitted results in an increase in transmission and storage costs.

[0003] Accordingly, there is a demand for an efficient image compression technique to effectively transmit, store, and reproduce information on high-resolution and high-quality images. SUMMARY

[0004] TECHNICAL PROBLEM

[0005] An object of the disclosure is to provide an image encoding / decoding method and apparatus having improved encoding / decoding efficiency.

[0006] Another object of the disclosure is to provide a method of determining whether a picture is output.

[0007] Another object of the disclosure is to solve a problem in which an activated NNPFA does not output any picture.

[0008] Another object of the disclosure is to provide a non-transitory recording medium storing a bitstream generated by an image encoding method according to the disclosure.

[0009] Another object of the disclosure is to provide a non-transitory recording medium storing a bitstream generated by an image encoding method according to the disclosure.

[0010] Another object of the disclosure is to provide a method of transmitting a bitstream generated by an image encoding method according to the disclosure.

[0011] The technical problems solved by the disclosure are not limited to the above-mentioned technical problems, and other technical problems not described herein will become apparent to those skilled in the art from the following description.

[0012] TECHNICAL SOLUTION

[0013] An image decoding method according to one aspect of the disclosure is performed by an image decoding apparatus. The image decoding method can include obtaining first information from an NNPFC (Neural Network Post Filter Characteristic) SEI (Supplemental Enhancement Information) message, the first information specifying whether at least one output picture corresponding to at least one input picture is generated, and obtaining second information from an NNPFA (Neural Network Post Filter Activation) SEI message, the second information specifying whether the generated output picture is output. A value of the second information can be determined based on at least one of whether a purpose of an NNPF (Neural Network Post Filter) is picture rate up-sampling or whether a number of input pictures is multiple.

[0014] An image encoding method according to another aspect of the disclosure is performed by an image encoding apparatus. The image encoding method can include encoding first information from an NNPFC (Neural Network Post Filter Characteristic) SEI (Supplemental Enhancement Information) message, the first information specifying whether at least one output picture corresponding to at least one input picture is generated, and encoding second information from an NNPFA (Neural Network Post Filter Activation) SEI message, the second information specifying whether the generated output picture is output. A value of the second information can be determined based on at least one of whether a purpose of an NNPF (Neural Network Post Filter) is picture rate up-sampling or whether a number of input pictures is multiple.

[0015] A computer-readable recording medium according to another aspect of the disclosure can store a bitstream generated by the image encoding method or apparatus of the disclosure.

[0016] A transmission method according to another aspect of the disclosure can transmit a bitstream generated by the image encoding method or apparatus of the disclosure.

[0017] The features described above with respect to the brief summary of the disclosure are merely exemplary aspects of the following detailed description of the disclosure, and do not limit the scope of the disclosure.

[0018] Advantageous Effects

[0019] According to the disclosure, it is possible to provide an image encoding / decoding method and apparatus having improved encoding / decoding efficiency.

[0020] Further, according to the disclosure, it is possible to determine whether a picture is output under a predetermined condition.

[0021] Further, according to the disclosure, it is possible to solve a problem in which an activated NNPFA does not output any picture.

[0022] Further, according to the disclosure, it is possible to provide a non-transitory computer-readable recording medium storing a bitstream generated by the image encoding method according to the disclosure.

[0023] Further, according to the present disclosure, it is possible to provide a non-transitory computer-readable recording medium storing a bitstream received, decoded, and used for reconstructing an image by an image decoding apparatus according to the present disclosure.

[0024] Further, according to the present disclosure, it is possible to provide a method of transmitting a bitstream generated by an image encoding method according to the present disclosure.

[0025] Those skilled in the art will appreciate that the effects achievable by the present disclosure are not limited to what has been particularly described hereinabove and other advantages of the present disclosure will be more clearly understood from the detailed description. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a view schematically showing a video coding system to which embodiments of the present disclosure are applicable.

[0027] Figure 2 is a view schematically showing an image encoding apparatus to which embodiments of the present disclosure are applicable.

[0028] Figure 3 is a view schematically showing an image decoding apparatus to which embodiments of the present disclosure are applicable.

[0029] Figure 4 is a view showing a deinterleaving method for a luminance channel.

[0030] Figure 5 is a flowchart illustrating an image encoding method according to an embodiment of the present disclosure.

[0031] Figure 6 is a flowchart illustrating an image decoding method according to an embodiment of the present disclosure.

[0032] Figure 7 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the present disclosure.

[0033] Figure 8 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the present disclosure.

[0034] Figure 9 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the present disclosure.

[0035] Figure 10 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the present disclosure.

[0036] Figure 11 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the present disclosure.

[0037] Figure 12 is a flowchart illustrating an image encoding / decoding method according to another embodiment of the disclosure.

[0038] Figure 13 is a view showing a content streaming system to which an embodiment of the disclosure is applicable. DETAILED DESCRIPTION

[0039] Hereinafter, embodiments of the disclosure will be described in detail with reference to the accompanying drawings so as to be easily practiced by those skilled in the art. However, the disclosure can be implemented in various different forms and is not limited to the embodiments described herein.

[0040] In describing the disclosure, if it is determined that a detailed description of related known functions or configurations makes the scope of the disclosure unnecessarily obscure, a detailed description thereof will be omitted. In the drawings, parts irrelevant to the description of the disclosure are omitted, and like reference numerals are attached to like parts.

[0041] In the disclosure, when a component is "connected", "coupled", or "linked" to another component, it can include not only a direct connection relationship but also an indirect connection relationship in which a middle component exists. In addition, when a component "includes" or "has" another component, it means that a further component can be further included, rather than excluding the other component, unless otherwise specified.

[0042] In the disclosure, the terms first, second, and the like are used only for the purpose of distinguishing one component from other components, and do not limit the order or importance of the components, unless otherwise specified. Accordingly, within the scope of the disclosure, a first component in one embodiment can be referred to as a second component in another embodiment, and similarly, a second component in one embodiment can be referred to as a first component in another embodiment.

[0043] In the disclosure, components distinguished from each other are intended to clearly describe each feature, and do not mean that the components must be separated. That is, a plurality of components can be integrated and implemented in one hardware or software unit, or one component can be distributed and implemented in a plurality of hardware or software units. Therefore, even if not specifically described, embodiments in which components are integrated or components are distributed are included in the scope of the disclosure.

[0044] In the disclosure, components described in various embodiments do not necessarily mean essential components, and some components can be optional components. Therefore, embodiments consisting of a subset of components described in the embodiments are also included in the scope of the disclosure. In addition, embodiments including other components in addition to the components described in various embodiments are included in the scope of the disclosure.

[0045] The disclosure relates to encoding and decoding of an image, and unless redefined in the disclosure, the terms used in the disclosure can have a general meaning commonly used in the technical field to which the disclosure belongs.

[0046] In the disclosure, a "picture" generally means a basis representing one image in a specific time period, and a slice / tile is a coding basis constituting a part of a picture. One picture can be composed of one or more slices / tiles. In addition, a slice / tile can include one or more coding tree units (CTU).

[0047] In the disclosure, a "pixel" or "pel" can mean a minimum unit constituting one picture (or image). In addition, a "sample" can be used as a term corresponding to a pixel. A sample can generally represent a pixel or a value of a pixel, and can represent only a pixel / pixel value of a luminance component or only a pixel / pixel value of a chrominance component.

[0048] In the disclosure, a "unit" can mean a basic unit of image processing. The unit can include at least one of a specific area of a picture and information related to the area. One unit can include one luminance block and two chrominance (e.g., Cb, Cr) blocks. In some cases, the unit can be used interchangeably with terms such as "sample array", "block", or "area". In general, an MxN block can include a set (or array) of M columns and N rows of samples (or sample array) or transform coefficients.

[0049] In the disclosure, a "current block" can mean one of a "current coding block", a "current coding unit", a "coding target block", a "decoding target block", or a "processing target block". When performing prediction, the "current block" can mean a "current prediction block" or a "prediction target block". When performing transform (inverse transform) / quantization (dequantization), the "current block" can mean a "current transform block" or a "transform target block". When performing filtering, the "current block" can mean a "filtering target block".

[0050] In addition, in the disclosure, unless explicitly stated as a chrominance block, the "current block" can mean a block including a luminance component block and a chrominance component block or a "luminance block of the current block". The luminance component block of the current block can be expressed by including an explicit description of the luminance component block such as "luminance block" or "current luminance block". In addition, the "chrominance component block of the current block" can be expressed by including an explicit description of the chrominance component block such as "chrominance block" or "current chrominance block".

[0051] In the disclosure, the term “ / ” or “,” can be interpreted to indicate “and / or”. For example, the expressions “A / B” and “A, B” can mean “A and / or B”. Also, “A / B / C” and “A, B, C” can mean “at least one of A, B, and / or C”.

[0052] In the disclosure, the term “or” should be interpreted to indicate “and / or”. For example, the expression “A or B” can include 1) only “A”, 2) only “B”, or 3) both “A and B”. In other words, in the disclosure, the term “or” should be interpreted to indicate “additionally or alternatively”.

[0053] Overview of a video encoding system

[0054] Figure 1 is a view schematically showing a video encoding system to which embodiments of the disclosure are applicable.

[0055] A video encoding system according to an embodiment can include an encoding apparatus 10 and a decoding apparatus 20. The encoding apparatus 10 can deliver encoded video and / or image information or data in the form of a file or streaming to the decoding apparatus 20 via a digital storage medium or a network.

[0056] The encoding apparatus 10 according to an embodiment can include a video source generator 11, an encoding unit (encoder) 12, and a transmitter 13. The decoding apparatus 20 according to an embodiment can include a receiver 21, a decoding unit (decoder) 22, and a renderer 23. The encoding unit 12 can be referred to as a video / image encoding apparatus, and the decoding unit 22 can be referred to as a video / image decoding apparatus. The transmitter 13 can be included in the encoding unit 12. The receiver 21 can be included in the decoding unit 22. The renderer 23 can include a display and the display can be configured as a separate device or an external component.

[0057] The video source generator 11 can acquire a video / image through a process of capturing, synthesizing, or generating a video / image. The video source generator 11 can include a video / image capturing device and / or a video / image generating device. The video / image capturing device can include, for example, one or more cameras, a video / image archive including previously captured videos / images, etc. The video / image generating device can include, for example, a computer, a tablet, and a smart phone, and can (electronically) generate a video / image. For example, a virtual video / image can be generated by a computer, etc. In this case, the video / image capturing process can be replaced by a process of generating related data.

[0058] The encoding unit 12 can encode an input video / image. For compression and coding efficiency, the encoding unit 12 can perform a series of processes such as prediction, transform, and quantization. The encoding unit 12 is capable of outputting encoded data (encoded video / image information) in the form of a bitstream.

[0059] The transmitter 13 can obtain the encoded video / image information or data output in the form of a bitstream and forward it to the receiver 21 of the decoding apparatus 20 or another external object in the form of a file or streaming through a digital storage medium or a network. The digital storage medium can include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD, etc. The transmitter 13 can include an element for generating a media file through a predetermined file format and can include an element for transmission through a broadcasting / communication network. The transmitter 13 can be provided as a transmission device separate from the encoding apparatus 12, and in this case, the transmission device can include at least one processor that acquires the encoded video / image information or data output in the form of a bitstream, and a transmission unit for transmitting the same in the form of a file or streaming. The receiver 21 can extract / receive a bitstream from a storage medium or a network and transmit the bitstream to the decoding unit 22.

[0060] The decoding unit 22 can decode a video / image by performing a series of processes such as dequantization, inverse transform, and prediction corresponding to the operations of the encoding unit 12.

[0061] The renderer 23 can render the decoded video / image. The rendered video / image can be displayed through a display.

[0062] Overview of an image encoding apparatus

[0063] Figure 2 is a view schematically showing an image encoding apparatus to which embodiments of the disclosure are applicable.

[0064] As shown in Figure 2 The image encoding apparatus 100 can include an image partitioner 110, a subtractor 115, a transformer 120, a quantizer 130, a dequantizer 140, an inverse transformer 150, an adder 155, a filter 160, a memory 170, an inter prediction unit 180, an intra prediction unit 185, and an entropy encoder 190. The inter prediction unit 180 and the intra prediction unit 185 can be collectively referred to as a "prediction unit". The transformer 120, the quantizer 130, the dequantizer 140, and the inverse transformer 150 can be included in a residual processor. The residual processor can further include the subtractor 115.

[0065] In some embodiments, all or at least some of the plurality of components configuring the image encoding apparatus 100 can be configured by one hardware component (e.g., an encoder or a processor). In addition, the memory 170 can include a decoded picture buffer (DPB) and can be configured by a digital storage medium.

[0066] The image partitioner 110 can partition an input image (or a picture or a frame) input to the image encoding apparatus 100 into one or more processing units. For example, the processing units can be referred to as coding units (CUs). The coding units can be obtained by recursively partitioning a coding tree unit (CTU) or a largest coding unit (LCU) according to a quadtree-binarytree-triple tree (QT / BT / TT) structure. For example, one coding unit can be partitioned into a plurality of coding units of a deeper depth based on a quadtree structure, a binary tree structure, and / or a triple tree structure. For the partitioning of the coding units, the quadtree structure can be applied first, and later the binary tree structure and / or the triple tree structure can be applied. The encoding process according to the present disclosure can be performed based on the final coding units that are not partitioned any more. The largest coding unit can be used as the final coding unit, or a coding unit of a deeper depth obtained by partitioning the largest coding unit can be used as the final coding unit. Here, the encoding process can include the processes of prediction, transform, and reconstruction that will be described later. As another example, the processing units of the encoding process can be prediction units (PUs) or transform units (TUs). The prediction units and the transform units can be partitioned or divided from the final coding units. The prediction units can be units of sample prediction, and the transform units can be units for deriving transform coefficients and / or units for deriving a residual signal from the transform coefficients.

[0067] The prediction unit (inter prediction unit 180 or intra prediction unit 185) can perform prediction on a block (current block) to be processed and generate a prediction block including prediction samples for the current block. The prediction unit can determine whether to apply intra prediction or inter prediction on a basis of the current block or CU. The prediction unit can generate various information related to prediction of the current block and transmit the generated information to the entropy encoder 190. The information about prediction can be encoded in the entropy encoder 190 and output in the form of a bitstream.

[0068] The intra prediction unit (intra predictor) 185 can predict the current block by referring to samples in the current picture. The reference samples can be located in the neighbors of the current block or can be placed separately depending on the intra prediction mode and / or the intra prediction technique. The intra prediction modes can include a plurality of non-directional modes and a plurality of directional modes. The non-directional modes can include, for example, a DC mode and a planar mode. The directional modes can include, for example, 33 directional prediction modes or 65 directional prediction modes depending on the level of detail of the prediction direction. However, this is merely an example, and more or less directional prediction modes can be used depending on the setting. The intra prediction unit 185 can determine the prediction mode applied to the current block by using the prediction mode applied to the neighboring block.

[0069] The inter prediction unit (inter predictor) 180 can derive a prediction block for the current block based on a reference block (reference sample array) specified by a motion vector on a reference picture. In this case, in order to reduce the amount of motion information transmitted in the inter prediction mode, the motion information can be predicted in units of a block, a sub-block, or a sample based on the correlation of the motion information between the neighboring block and the current block. The motion information can include a motion vector and a reference picture index. The motion information can further include inter prediction direction (L0 prediction, L1 prediction, Bi prediction, etc.) information. In the case of inter prediction, the neighboring block can include a spatial neighboring block existing in the current picture and a temporal neighboring block existing in the reference picture. The reference picture including the reference block and the reference picture including the temporal neighboring block can be the same or different. The temporal neighboring block can be referred to as a collocated reference block, a colCU, etc. The reference picture including the temporal neighboring block can be referred to as a collocated picture (colPic). For example, the inter prediction unit 180 can construct a motion information candidate list based on the neighboring block and generate information indicating which candidate to use to derive the motion vector and / or the reference picture index of the current block. The inter prediction can be performed based on various prediction modes. For example, in the case of a skip mode and a merge mode, the inter prediction unit 180 can use the motion information of the neighboring block as the motion information of the current block. In the case of the skip mode, unlike the merge mode, a residual signal can not be transmitted. In the case of a motion vector prediction (MVP) mode, the motion vector of the neighboring block can be used as a motion vector predictor, and the motion vector of the current block can be signaled by encoding a motion vector difference and an indicator for the motion vector predictor. The motion vector difference can mean the difference between the motion vector of the current block and the motion vector predictor.

[0070] The prediction unit can generate a prediction signal based on various prediction methods and prediction techniques described below. For example, the prediction unit can not only apply intra prediction or inter prediction, but also simultaneously apply both intra prediction and inter prediction in order to predict the current block. The prediction method that simultaneously applies both intra prediction and inter prediction to predict the current block can be referred to as combined inter and intra prediction (CIIP). In addition, the prediction unit can perform intra block copy (IBC) for the prediction of the current block. Intra block copy can be used for content image / video coding of games and the like, for example, screen content coding (SCC). IBC is a method of predicting a current picture using a previously reconstructed reference block in the current picture at a position separated by a predetermined distance. When IBC is applied, the position of the reference block in the current picture can be coded as a vector (block vector) corresponding to the predetermined distance. IBC basically performs prediction in the current picture, but can be performed similarly to inter prediction because the reference block is derived within the current picture. That is, IBC can use at least one of the inter prediction techniques described in the present disclosure.

[0071] The prediction signal generated by the prediction unit can be used to generate a reconstructed signal or to generate a residual signal. The subtracter 115 can generate a residual signal (a residual block or a residual sample array) by subtracting the prediction signal (a prediction block or a prediction sample array) output from the prediction unit from the input image signal (an original block or an original sample array). The generated residual signal can be transmitted to the transformer 120.

[0072] The transformer 120 can generate transform coefficients by applying a transform technique to the residual signal. For example, the transform technique can include at least one of a discrete cosine transform (DCT), a discrete sine transform (DST), a karhunen-loève transform (KLT), a graph-based transform (GBT), or a conditional non-linear transform (CNT). Here, GBT means a transform obtained from a graph when relationship information between pixels is represented by a graph. CNT refers to a transform obtained based on a prediction signal generated using all previously reconstructed pixels. In addition, the transform process can be applied to a square pixel block having the same size or can be applied to a block having a variable size other than a square.

[0073] The quantizer 130 can quantize the transform coefficients and transmit them to the entropy encoder 190. The entropy encoder 190 can encode the quantized signal (information about the quantized transform coefficients) and output a bitstream. The information about the quantized transform coefficients can be referred to as residual information. The quantizer 130 can rearrange the quantized transform coefficients in a block type into a one-dimensional vector form based on a coefficient scan order, and generate information about the quantized transform coefficients based on the quantized transform coefficients in the one-dimensional vector form.

[0074] The entropy encoder 190 can perform various encoding methods such as, for example, exponential Golomb, context-adaptive variable length coding (CAVLC), context-adaptive binary arithmetic coding (CABAC), and / or the like. The entropy encoder 190 can encode information necessary for video / image reconstruction, other than quantized transform coefficients (e.g., values of syntax elements, and / or the like), together or individually. The encoded information (e.g., encoded video / image information) can be transmitted or stored in the form of a bitstream in units of network abstraction layer (NAL). The video / image information can further include information on various parameter sets such as an adaptation parameter set (APS), a picture parameter set (PPS), a sequence parameter set (SPS), or a video parameter set (VPS). In addition, the video / image information can further include general constraint information. The signaled information, transmitted information, and / or syntax elements described in the disclosure can be encoded through the above-described encoding processes and included in the bitstream.

[0075] The bitstream can be transmitted through a network or can be stored in a digital storage medium. The network can include a broadcasting network and / or a communication network, and the digital storage medium can include various storage media such as USB, SD, CD, DVD, Blu-ray, HDD, SSD, and / or the like. A transmitter (not shown) that transmits a signal output from the entropy encoder 190 and / or a storage unit (not shown) that stores the signal can be included as an internal / external element of the image encoding apparatus 100. Alternatively, the transmitter can be provided as a component of the entropy encoder 190.

[0076] The quantized transform coefficients output from the quantizer 130 can be used to generate a residual signal. For example, the residual signal (residual block or residual sample) can be reconstructed by applying dequantization and inverse transform to the quantized transform coefficients via the dequantizer 140 and the inverse transformer 150.

[0077] The adder 155 adds the reconstructed residual signal to the prediction signal output from the inter prediction unit 180 or the intra prediction unit 185 to generate a reconstructed signal (reconstructed picture, reconstructed block, reconstructed sample array). If there is no residual of a block to be processed, such as the case where a skip mode is applied, the prediction block can be used as the reconstructed block. The adder 155 can be referred to as a reconstructor or a reconstructed block generator. The generated reconstructed signal can be used for intra prediction of a next block to be processed in the current picture, and can be used for inter prediction of a next picture by filtering as described below.

[0078] Filter 160 can improve subjective / objective image quality by applying filtering to the reconstructed signal. For example, filter 160 can generate a modified reconstructed image by applying various filtering methods to the reconstructed image and store the modified reconstructed image in memory 170, specifically in the DPB of memory 170. Various filtering methods may include, for example, deblocking filtering, sample adaptive shifting, adaptive loop filtering, bilateral filtering, etc. Filter 160 can generate various filtering-related information and send the generated information to entropy encoder 190, as described later in the description of each filtering method. The filtering-related information can be encoded by entropy encoder 190 and output as a bitstream.

[0079] The modified reconstructed image sent to memory 170 can be used as a reference image in inter-frame prediction unit 180. When inter-frame prediction is applied by image coding device 100, prediction mismatch between image coding device 100 and image decoding device can be avoided and coding efficiency can be improved.

[0080] The DPB of memory 170 can store modified reconstructed images for use as reference images in inter-frame prediction unit 180. Memory 170 can store motion information of blocks from which motion information in the current image is derived (or encoded) and / or motion information of already reconstructed blocks in the image. The stored motion information can be sent to inter-frame prediction unit 180 and used as motion information for spatially or temporally neighboring blocks. Memory 170 can store reconstructed samples of reconstructed blocks in the current image and can transmit the reconstructed samples to intra-frame prediction unit 185.

[0081] Overview of an image decoding apparatus

[0082] Figure 3 This is a schematic view illustrating an image decoding apparatus to which embodiments of the present disclosure may be applied.

[0083] like Figure 3 As shown, the image decoding device 200 may include an entropy decoder 210, a dequantizer 220, an inverse transformer 230, an adder 235, a filter 240, a memory 250, an inter-frame predictor 260, and an intra-frame predictor 265. The inter-frame predictor (inter-frame predictor unit) 260 and the intra-frame predictor (intra-frame predictor unit) 265 may be collectively referred to as a "prediction unit (predictor)". The dequantizer 220 and the inverse transformer 230 may be included in a residual processor.

[0084] According to an embodiment, all or at least some of the plurality of components of the image decoding apparatus 200 can be configured by hardware components (e.g., decoders or processors). In addition, the memory 250 can include a decoded picture buffer (DPB) or can be configured by a digital storage medium.

[0085] The image decoding apparatus 200 that has received a bitstream including video / image information can reconstruct an image by performing a process corresponding to a process performed by the image encoding apparatus 100. For example, the image decoding apparatus 200 can perform decoding using a processing unit applied in the image encoding apparatus. Accordingly, the processing unit for decoding can be, for example, a coding unit. The coding unit can be obtained by partitioning a coding tree unit or a largest coding unit. A reconstructed image signal decoded and output by the image decoding apparatus 200 can be reproduced through a reproduction apparatus (not shown). Figure 2

[0086] The image decoding apparatus 200 can receive a bitstream in the form of a stream from the image encoding apparatus 100, and can decode and output the bitstream. Figure 2 ​The received signal can be decoded by the entropy decoder 210. For example, the entropy decoder 210 can parse a bitstream to derive information (e.g., video / image information) necessary for image reconstruction (or picture reconstruction). The video / image information can further include information on various parameter sets such as an adaptation parameter set (APS), a picture parameter set (PPS), a sequence parameter set (SPS), or a video parameter set (VPS). In addition, the video / image information can further include general constraint information. The image decoding apparatus can further decode a picture based on the information on the parameter sets and / or the general constraint information. The information and / or the syntax elements described in the disclosure and received / sent by signal can be decoded by a decoding process and obtained from the bitstream. For example, the entropy decoder 210 decodes information in the bitstream based on an encoding method such as exponential Golomb coding, CAVLC, or CABAC, and outputs values of syntax elements necessary for image reconstruction and quantized values of transform coefficients for a residual. More specifically, the CABAC entropy decoding method can receive bins corresponding to each syntax element in the bitstream, determine a context model using decoded information of a decoding target syntax element, a neighboring block, and a decoding target block, or information of decoded symbols / bins in a previous stage, and perform arithmetic decoding on bins according to a probability of occurrence of the bins determined by the context model, and generate symbols corresponding to values of each syntax element. In this case, the CABAC entropy decoding method can update the context model by using information of the decoded symbols / bins for the context model of the next symbol / bin after determining the context model. Among the information decoded by the entropy decoder 210, prediction-related information can be provided to the prediction unit (inter-predictor 260 and intra-predictor 265), and residual values, i.e., quantized transform coefficients and related parameter information, on which entropy decoding is performed in the entropy decoder 210 can be input to the dequantizer 220. In addition, among the information decoded by the entropy decoder 210, information on filtering can be provided to the filter 240. Meanwhile, a receiver (not shown) for receiving a signal output from the image encoding apparatus can be further configured as an internal / external element of the image decoding apparatus 200, or the receiver can be a component of the entropy decoder 210.

[0087] Meanwhile, the image decoding apparatus according to the disclosure can be referred to as a video / image / picture decoding apparatus. The image decoding apparatus can be classified into an information decoder (video / image / picture information decoder) and a sample decoder (video / image / picture sample decoder). The information decoder can include the entropy decoder 210. The sample decoder can include at least one of the dequantizer 220, the inverse transformer 230, the adder 235, the filter 240, the memory 250, the inter-prediction unit 160, or the intra-prediction unit 265.

[0088] The dequantizer 220 can dequantize the quantized transform coefficients and output the transform coefficients. The dequantizer 220 can rearrange the quantized transform coefficients in the form of a two-dimensional block. In this case, the rearrangement can be performed based on a coefficient scan order performed in the image encoding apparatus. The dequantizer 220 can perform dequantization on the quantized transform coefficients by using a quantization parameter (e.g., quantization step length information) and obtain the transform coefficients.

[0089] The inverse transformer 230 can inverse-transform the transform coefficients to obtain a residual signal (a residual block, a residual sample array).

[0090] The prediction unit can perform prediction on the current block and generate a prediction block including prediction samples for the current block. The prediction unit can determine whether to apply intra prediction or inter prediction to the current block based on information about prediction output from the entropy decoder 210, and can determine a specific intra / inter prediction mode (prediction technique).

[0091] The same as described in the prediction unit of the image encoding apparatus 100 is that the prediction unit can generate a prediction signal based on various prediction methods (techniques) described later.

[0092] The intra predictor 265 can predict the current block by referring to samples in the current picture. The description of the intra prediction unit 185 is equally applied to the intra predictor 265.

[0093] The inter predictor 260 can derive a prediction block for the current block based on a reference block (a reference sample array) specified by a motion vector on a reference picture. In this case, in order to reduce the amount of motion information transmitted in the inter prediction mode, the motion information can be predicted in units of a block, a sub-block, or a sample based on the correlation of the motion information between neighboring blocks and the current block. The motion information can include a motion vector and a reference picture index. The motion information can further include inter prediction direction (L0 prediction, L1 prediction, Bi prediction, etc.) information. In the case of inter prediction, the neighboring blocks can include spatial neighboring blocks present in the current picture and temporal neighboring blocks present in the reference picture. For example, the inter predictor 260 can configure a motion information candidate list based on the neighboring blocks, and derive a motion vector and / or a reference picture index of the current block based on received candidate selection information. The inter prediction can be performed based on various prediction modes, and information about prediction can include information indicating an inter prediction mode for the current block.

[0094] The adder 235 can generate a reconstructed block by adding the obtained residual signal to a prediction signal (prediction block, prediction sample array) output from the prediction unit (including the inter-predictor 260 and / or the intra-predictor 265). If there is no residual for the block to be processed, such as when a skip mode is applied, the prediction block can be used as the reconstructed block. The description of the adder 155 is equally applicable to the adder 235. The adder 235 can be referred to as a reconstructor or a reconstructed block generator. The generated reconstructed signal can be used for intra-prediction of the next block to be processed in the current picture, and can be used for inter-prediction of the next picture by filtering as described below.

[0095] The filter 240 can improve subjective / objective image quality by applying filtering to the reconstructed signal. For example, the filter 240 can generate a modified reconstructed picture by applying various filtering methods to the reconstructed picture, and store the modified reconstructed picture in the memory 250, specifically, in the DPB of the memory 250. The various filtering methods can include, for example, deblocking filtering, sample adaptive offset, adaptive loop filter, bilateral filter, etc.

[0096] The (modified) reconstructed picture stored in the DPB of the memory 250 can be used as a reference picture in the inter-predictor 260. The memory 250 can store motion information of a block from which motion information of the current picture is derived (or decoded), and / or motion information of a block that has been reconstructed in the picture. The stored motion information can be transmitted to the inter-predictor 260 so as to be utilized as motion information of a spatial neighboring block or motion information of a temporal neighboring block. The memory 250 can store reconstructed samples of the reconstructed block in the current picture and transfer the reconstructed samples to the intra-predictor 265.

[0097] In the present disclosure, the embodiments described in the filter 160, the inter-predictor 180, and the intra-predictor 185 of the image encoding apparatus 100 can be equally or correspondingly applied to the filter 240, the inter-predictor 260, and the intra-predictor 265 of the image decoding apparatus 200.

[0098] Neural network post filter characteristic (NNPFC)

[0099] The combination of Table 1 to Table 3 represents the NNPFC syntax structure.

[0100] [Table 1]

[0101] [Table 2]

[0102] [Table 3]

[0103] The NNPFC syntax structure of Table 1 to Table 3 can be signaled in the form of a supplemental enhancement information (SEI) message. The SEI message that signals the NNPFC syntax structure of Table 1 to Table 3 can be referred to as a NNPFC SEI message.

[0104] The neural network post filter characteristic (NNPFC) SEI message specifies a neural network that is used as a post-processing filter. The use of the specified neural network post-processing filter (NNPF) for a particular picture is indicated with the neural network post filter activation (NNPFA) SEI message. Here, "post-processing filter" and "post filter" can have the same meaning.

[0105] The following variables need to be defined for the use of this SEI message: - The input picture width and height in luma samples, denoted here by CroppedWidth and CroppedHeight, respectively.

[0106] - The luma sample array CroppedYPic [ idx ] and the chroma sample array CroppedCbPic [ idx ] (when present) of the input picture with index idx in the range of 0 to numInputPics - 1 (inclusive) are used as input to the NNPF.

[0107] - BitDepth Y may represent the bit depth used for the luma sample array of the input picture.

[0108] - BitDepth C may represent the bit depth used for the chroma sample array (if any) of the input picture.

[0109] - ChromaFormatIdc may represent the chroma format identifier.

[0110] - When nnpfc_auxiliary_inp_idc is equal to 1, the filter strength control value array StrengthControlVal [ idx ] shall contain a real number in the range of 0 to 1 (inclusive) for the input picture with index idx in the range of 0 to numInputPics - 1 (inclusive).

[0111] An input picture with index 0 can correspond to a picture for which the NNPF defined by this NNPFC SEI message is activated by the NNPFA SEI message. An input picture with index i in the range of 1 to numInputPics - 1, inclusive, can precede, in output order, an input picture with index i - 1.

[0112] nnpfc_purpose indicates the purpose of the NNPF as specified in Table 4. In Table 4, (nnpfc_purpose & bitMask) not equal to 0 can indicate that the NNPF of Table 4 has a purpose associated with the bitMask value. When nnpfc_purpose is greater than 0 and (nnpfc_purpose & bitMask) is equal to 0, the purpose associated with the bitMask value does not apply to this NNPF. When nnpfc_purpose is equal to 0, the NNPF can be used as determined by the application. In the bitstream, the value of nnpfc_purpose shall be in the range of 0 to 63, inclusive. Values of 64 to 65535, inclusive, for nnpfc_purpose are reserved for future use. Decoders shall ignore NNPFC SEI messages with nnpfc_purpose in the range of 64 to 65535, inclusive.

[0113] [Table 4]

[0114] The variables chromaUpsamplingFlag specifying whether nnpfc_purpose indicates the purpose of a NNPF including chroma upsampling, resolutionResamplingFlag specifying whether nnpfc_purpose indicates the purpose of a NNPF including resolution resampling, pictureRateUpsamplingFlag specifying whether nnpfc_purpose indicates the purpose of a NNPF including picture rate upsampling, bitDepthUpsamplingFlag indicating the purpose of a NNPF including bit depth upsampling, and colourizationFlag specifying the purpose of a NNPF including colourization are derived as shown in Table 5 below.

[0115] [Table 5]

[0116] When ChromaFormatIdc equals 3, chromaUpsamplingFlag should equal 0. When ChromaFormatIdc or chromaUpsamplingFlag is not equal to 0, colourizationFlag should equal 0. When pictureRateUpsamplingFlag equals 1, and an input picture with index 0 is associated with a frame packing arrangement SEI message with fp_arrangement_type equal to 5, all input pictures can be associated with a frame packing arrangement SEI message with the same value as fp_arrangement_type and fp_current_frame_is_frame0_flag.

[0117] nnpfc_id can contain an identifier that can be used to identify NNPY. The value of nnpfc_id should be between 0 and 2. 32 - 2 (including 0 and 2) 32 - 2) range. From 256 to 511 (inclusive) and from 2 31 to 2 32 - 2 (including 2) 31 and 2 32 - 2) The value of nnpfc_id can be reserved for future use. When encountering an nnpfc_id in the range of 256 to 511 (inclusive) or in the range of 2... 31 Up to 2 32 – 2 (including 2) 31 and 2 32 – 2) The decoder of NNPFC SEI messages for nnpfc_id within the range should ignore SEI messages.

[0118] When the NNPFC SEI message is the first NNPFC SEI message in the current CLVS with a specific nnpfc_id value in decoding order, the following applies: - This SEI message specifies the underlying NNPF.

[0119] This SEI message contains information about the currently decoded image in output order, as well as all subsequent decoded images of the current layer, until the current CLVS ends.

[0120] An NNPFC SEI message can be a repetition of a previous NNPFC SEI message in the current CLVS in the order of decoding, and subsequent semantics can be applied as if this SEI message were the only NNPFC SEI message with the same content in the current CLVS.

[0121] nnpfc_mode_idc equal to 0 can specify that the SEI message can contain a bitstream representing a base NNPF or can represent an update relative to a base NNPF with the same nnpfc_id value.

[0122] When the NNPFC SEI message is the first NNPFC SEI message in decoding order within the current CLVS with a particular nnpfc_id value, nnpfc_mode_idc equal to 1 can specify that the base NNPF associated with the nnpfc_id value is a neural network and that the neural network can be the neural network identified by the format identified by the tag URI nnpfc_tag_uri using the URI represented by nnpfc_uri.

[0123] If the NNPFC SEI message is neither the first NNPFC SEI message in decoding order within the current CLVS with a particular nnpfc_id value nor a repetition of the first NNPFC SEI message, nnpfc_mode_idc equal to 1 can specify that the update relative to the base NNPF with the same nnpfc_id value is defined by the format identified by the tag URI nnpfc_tag_uri using the URI represented by nnpfc_uri.

[0124] nnpfc_base_flag equal to 1 can specify that the SEI message specifies a base NNPF. nnpf_base_flag equal to 0 can specify that the SEI message specifies an update relative to a base NNPF.

[0125] The following constraint applies to the value of nnpfc_base_flag: - When the NNPFC SEI message is the first NNPFC SEI message in decoding order within the current CLVS with a particular nnpfc_id value, the value of nnpfc_base_flag shall be equal to 1.

[0126] When the NNPFC SEI message nnpfcB is not the first NNPFC SEI message in decoding order within the current CLVS with a particular nnpfc_id value and the value of nnpfc_base_flag is equal to 1, the NNPFC SEI message shall be a repetition of the first NNPFC SEI message in decoding order with the same nnpfc_id value nnpfcA, i.e., the payload content of nnpfcB shall be the same as nnpfcA.

[0127] When nnpfc_base_flag is equal to 0, the following applies: - This SEI message defines an update with respect to the previous base NNPF in decoding order with the same nnpfc id value. The update is not cumulative, but each update is applied to the base NNPF, which is the NNPF specified by the first NNPFC SEI message in decoding order with the particular nnpfc id value within the current CLVS. The NNPF defined by this SEI message can be obtained by applying the update defined by this SEI message to the base NNPF with the same nnpfc id value.

[0128] - This SEI message is with respect to the current decoded picture of the current layer in output order and all subsequent decoded pictures up to the end of the current CLVS. In this case, the decoded pictures following the current decoded picture in output order within the current CLVS can be excluded. This SEI message is associated with the subsequent NNPFC SEI message in decoding order with nnpfc base flag equal to 0 and the particular nnpfc id value within the current CLVS (whichever is earlier).

[0129] nnpfc mode idc equal to 0 can indicate that this SEI message contains a bitstream that specifies a base NNPF (when nnpfc base flag is equal to 1) or an update with respect to the base NNPF with the same nnpfc id value (when nnpfc base flag is equal to 0). When nnpfc base flag is equal to 1, nnpfc mode idc equal to 1 can specify that the base NNPF associated with the nnpfc id value is a neural network identified by a URI. In this case, the URI can be indicated by nnpfc uri with the format identified by the label URI nnpfc tag uri. When nnpfc base flag is equal to 0, nnpfc mode idc equal to 1 can specify that the update with respect to the base NNPF with the same nnpfc id value is defined by a URI. In this case, the URI can be indicated by nnpfc uri with the format identified by the label URI nnpfc tag uri.

[0130] The value of nnpfc mode idc shall be in the range of 0 to 1, inclusive, in the bitstream. The values of 2 to 255, inclusive, for nnpfc mode idc can be reserved for future use and shall not appear in the bitstream. Decoders shall ignore NNPFC SEI messages with nnpfc mode idc in the range of 2 to 255, inclusive. Values of nnpfc mode idc greater than 255 shall not appear in the bitstream and can not be reserved for future use.

[0131] In the bitstream, nnpfc_reserved_zero_bit_a shall be equal to 0. Decoders shall ignore NNPFC SEI messages in which nnpfc_reserved_zero_bit_a is not equal to 0.

[0132] nnpfc_tag_uri can contain a tag URI with syntax and semantics as specified in IETF RFC 4151 that identifies a neural network that is used as the underlying NNPF or an update relative to the underlying NNPF with the same nnpfc_id value as specified by nnpfc_uri. nnpfc_tag_uri enables uniquely identifying the format of the neural network data specified by nnrpf_uri without the need for a central registration authority. nnpfc_tag_uri equal to "tag:iso.org,2023:15938-17" can indicate that the neural network data identified by nnpfc_uri conforms to ISO / IEC 15938-17.

[0133] nnpfc_uri can contain a URI with syntax and semantics as specified in IETF Internet Standard 66 that identifies a neural network that is used as the underlying NNPF or an update relative to the underlying NNPF with the same nnpfc_id value.

[0134] nnpfc_property_present_flag equal to 1 can specify that syntax elements related to filter usage, input formatting, output formatting, and complexity are present. nnpfc_property_present_flag equal to 0 can specify that syntax elements related to filter usage, input formatting, output formatting, and complexity are not present. When nnpfc_base_flag is equal to 1, nnpfc_property_present_flag shall be equal to 1. When nnpfc_property_present_flag is equal to 0, the values of all syntax elements that can be present only when nnpfc_property_present_flag is equal to 1 can be inferred to be equal to their corresponding syntax elements in the NNPFC SEI message of the underlying NNPF for which this SEI message provides an update, respectively.

[0135] When the NNPFC SEI message nnpfcCurr is not the first NNPFC SEI message with a particular nnpfc id value in decoding order within the current CLVS, is not a repetition of the first NNPFC SEI message with that particular nnpfc id (i.e., the value of nnpfc base flag is equal to 0), and the value of nnpfc property present flag is equal to 1, the following restrictions apply: - The value of nnpfc purpose in the NNPFC SEI message shall be the same as the value of nnpfc purpose in the first NNPFC SEI message with that particular nnpfc id value in decoding order within the current CLVS.

[0136] - The values of the syntax elements in the NNPFC SEI message that are in decoding order subsequent to nnpfc property present flag and prior to nnpfc complexity info present flag shall be the same as the values of the corresponding syntax elements in the first NNPFC SEI message with that particular nnpfc id value in decoding order within the current CLVS.

[0137] - In the first NNPFC SEI message with a particular nnpfc id value in decoding order within the current CLVS, nnpfc complexity info present flag shall be equal to 0 or nnpfc complexity info present flag shall be equal to 0 or 1 and all of the following apply: (1) nnpfc parameter type idc in nnpfcCurr shall be equal to nnpfc parameter type idc in nnpfcBase.

[0138] (2) nnpfc log2 parameter bit length minus3 in nnpfcCurr (when present) shall be less than or equal to nnpfc log2 parameter bit length minus3 in nnpfcBase.

[0139] (3) If nnpfc num parameters idc in nnpfcBase is equal to 0, nnpfc num parameters idc in nnpfcCurr shall be equal to 0.

[0140] (4) Otherwise (nnpfc num parameters idc in nnpfcBase is greater than 0), nnpfc num parameters idc in nnpfcCurr shall be greater than 0 and less than or equal to nnpfc num parameters idc in nnpfcBase.

[0141] (5) If nnpfc num kmac operations idc in nnpfcBase is equal to 0, nnpfc num kmac operations idc in nnpfcCurr shall be equal to 0.

[0142] (6) Otherwise (nnpfc num kmac operations idc in nnpfcBase is greater than 0), nnpfc num kmac operations idc in nnpfcCurr shall be greater than 0 and less than or equal to nnpfc num kmac operations idc in nnpfcBase.

[0143] (7) If nnpfc total kilobyte size in nnpfcBase is equal to 0, nnpfc total kilobyte size in nnpfcCurr shall be equal to 0.

[0144] (8) Otherwise (nnpfc total kilobyte size in nnpfcBase is greater than 0), nnpfc total kilobyte size in nnpfcCurr shall be greater than 0 and less than or equal to nnpfc total kilobyte size in nnpfcBase.

[0145] nnpfc num input pics minusl + 1 can specify the number of pictures used as input for the NNPF. The value of nnpfc num input pics minusl shall be in the range of 0 to 63, inclusive. When pictureRateUpsamplingFlag is equal to 1, the value of nnpfc num input pics minusl shall be greater than 0.

[0146] The variable numInputPics, which specifies the number of pictures used as input for the NNPF, can be derived as follows: [Equation 1]

[0147] nnpfc_input_pic_output_flag[ i ] equal to 1 can indicate that for the i-th input picture, the NNPF generates a corresponding output picture. nnpfc_input_pic_output_flag[ i ] equal to 0 can indicate that for the i-th input picture, the NNPF does not generate a corresponding output picture. When nnpfc_num_input_pics_minus1 is equal to 0, nnpfc_input_pic_output_flag[ 0 ] is inferred to be equal to 1. When pictureRateUpsamplingFlag is equal to 0 and nnpfc_num_input_pics_minus1 is greater than 0, for at least one value of i in the range of 0 to nnpfc_num_input_pics_minus1, inclusive, nnpfc_input_pic_output_flag[ i ] shall be equal to 1. nnpfc_input_pic_output_flag[ i ] can be referred to as nnpfc_input_pic_filtering_flag[ i ].

[0148] nnpfc_absent_input_pic_zero_flag equal to 1 can indicate that an input picture that is not expected to be present in the bitstream is to be represented by a sample array with sample values equal to 0. nnpfc_absent_input_pic_flag equal to 0 can indicate that an input picture that is not expected to be present in the bitstream is to be represented by the closest input picture in output order within the bitstream.

[0149] When chromaUpsamplingFlag is equal to 1, nnpfc_out_sub_c_flag can specify the values of the variables outSubWidthC and outSubHeightC. nnpfc_out_sub_c_flag equal to 1 can specify that outSubWidthC is equal to 1 and outSubHeightC is equal to 1. nnpfc_out_sub_c_flag equal to 0 can specify that outSubWidthC is equal to 2 and outSubHeightC is equal to 1. When ChromaFormatldc is equal to 2 and nnpfc_out_sub_c_flag is present, the value of nnpfc_out_sub_c_flag shall be equal to 1.

[0150] When colourizationFlag is equal to 1, nnpfc_out_colour_format_idc can specify the colour format of the NNPF output and accordingly the values of the variables outSubWidthC and outSubHeightC. nnpfc_out_colour_format_idc equal to 1 can specify that the colour format of the NNPF output is 4:2:0 format and both outSubWidthC and outSubHeightC are equal to 2. nnpfc_out_colour_format_idc equal to 2 can specify that the colour format of the NNPF output is 4:2:2 format and outSubWidthC is equal to 2 and outSubHeightC is equal to 1. nnpfc_out_colour_format_idc equal to 3 can specify that the colour format of the NNPF output is 4:4:4 format and both outSubWidthC and outSubHeightC are equal to 1. The value of nnpfc_out_colour_format_idc shall not be equal to 0. When both chromaUpsamplingFlag and colourizationFlag are equal to 0, outSubWidthC and outSubHeightC can be inferred to be equal to SubWidthC and SubHeightC, respectively.

[0151] nnpfc_pic_width_num_minus1 + 1 and nnpfc_pic_width_denom_minus1 + 1 can specify the numerator and the denominator, respectively, of the resampling ratio for the NNPF output picture width relative to CroppedWidth. The value of (nnpfc_pic_width_num_minus1 + 1) ÷ (nnpfc_pic_width_denom_minus1 + 1) shall be in the range of 1 ÷ 16 to 16, inclusive. When nnpfc_pic_width_num_minus1 and nnpfc_pic_width_denom_minus1 are not present, the values of nnpfc_pic_width_num_minus1 and nnpfc_pic_width_denom_minus1 can both be inferred to be equal to 0.

[0152] The variable nnpfcOutputPicWidth, representing the width of the luma sample array of the picture resulting from the application of the NNPF identified by nnpfc_id to the input picture, can be derived as follows: [Equation 2]

[0153] The remainder of nnpfcOutputPicWidth divided by outSubWidthC shall be 0.

[0154] nnpfc_pic_height_num_minus1 + 1 and nnpfc_pic_height_denom_minus1 + 1 can specify the numerator and the denominator, respectively, of the repeated sampling ratio for the NNPF output picture height relative to CroppedHeight. The value of (nnpfc_pic_height_num_minus1 + 1) ÷ (nnpfc_pic_height_denom_minus1 + 1) shall be in the range of 1 ÷ 16 to 16, inclusive. When nnpfc_pic_height_num_minus1 and nnpfc_pic_height_denom_minus1 are not present, the values of nnpfc_pic_height_num_minus1 and nnpfc_pic_height_denom_minus1 can both be inferred to be equal to 0.

[0155] The variable nnpfcOutputPicHeight, representing the height of the array of luma samples of the picture resulting from the application of the NNPF identified by nnpfc_id to the input picture, can be derived as follows: [Equation 3]

[0156] The remainder of nnpfcOutputPicHeight divided by outSubHeightC shall be equal to 0.

[0157] When nnpfc_pic_width_num_minus1, nnpfc_pic_width_denom_minus1, nnpfc_pic_height_num_minus1, and nnpfc_pic_height_denom_minus1 are present, at least one of the following shall be true: The value of nnpfcOutputPicWidth is not equal to CroppedWidth.

[0158] The value of nnpfcOutputPicHeight is not equal to CroppedHeight.

[0159] nnpfc_interpolated_pics[ i ] can specify the number of interpolated pictures generated by the NNPF between the i-th and the (i+1)-th picture that are used as input for the NNPF. The value of nnpfc_interpolated_pics[ i ] shall be in the range of 0 to 63, inclusive. The value of nnpfc_interpolated_pics[ i ] shall be greater than 0 for at least one value of i in the range of 0 to nnpfc_num_input_pics_minusl - 1, inclusive.

[0160] The variable NumInpPicsInOutputTensor that specifies the number of pictures that have corresponding input pictures and that are present in the output tensor of the NNPF, the variable InpIdx[ idx ] that specifies the input picture index of the idx-th picture that is present in the output tensor of the NNPF and that has a corresponding input picture, and the variable numOutputPics that specifies the total number of pictures that are present in the output tensor of the NNPF can be derived as shown in Table 6.

[0161] [Table 6]

[0162] nnpfc_component_last_flag equal to 1 can indicate that the last dimension in the input tensor inputTensor to the NNPF and in the output tensor outputTensor produced by the NNPF is used for the current channel. nnpfc_component_last_flag equal to 0 can indicate that the third dimension in the input tensor inputTensor to the NNPF and in the output tensor outputTensor produced by the NNPF is used for the current channel. The first dimension in the input tensor and in the output tensor can be used as a batch index, which is a common practice in some neural network frameworks. While the formula in the semantics of this SEI message uses a batch size corresponding to a batch index equal to 0, it depends on post-processing implementations that determine the batch size that is used as input to the neural network inference. For example, when nnpfc_inp_order_idc is equal to 3 and nnpfc_auxiliary_inp_idc is equal to 1, there are 7 channels in the input tensor, including four luma matrices, two chroma matrices, and one auxiliary input matrix. In this case, DeriveInputTensors( ) derives each of these 7 channels of the input tensor one by one, and when a particular channel of these channels is processed, the channel can be referred to as the current channel during the process.

[0163] nnpfc_inp_format_idc can indicate a method of converting sample values of an input picture to input values to the NNPF. When nnpfc_inp_format_idc is equal to 0, the input values to the NNPF are real numbers, and the functions InpY() and InpC() can be specified as shown in Equation 4.

[0164] [Equation 4]

[0165] When nnpfc_inp_format_idc is equal to 1, the input values to the NNPF are unsigned integers, and the functions InpY() and InpC() are specified as shown in Table 7.

[0166] [Table 7]

[0167] The variable inpTensorBitDepthY can be derived from the syntax element nnpfc_inp_tensor_luma_bitdepth_minus8 specified as follows. The variable inpTensorBitDepthC can be derived from the syntax element nnpfc_inp_tensor_chroma_bitdepth_minus8 specified as follows. Values of nnpfc_inp_format_idc greater than 1 can be reserved for future use and shall not occur in the bitstream. Decoders shall ignore NNPFC SEI messages containing reserved values of nnpfc_inp_format_idc.

[0168] nnpfc auxiliary inp idc greater than 0 can indicate that there is auxiliary input data in the input tensor of the NNPF. nnpfc auxiliary inp idc equal to 0 can indicate that there is no auxiliary input data in the input tensor. nnpfc auxiliary inp idc equal to 1 can specify that the auxiliary input data is derived by the method as shown in Table 10 to Table 12. In the bitstream, the value of nnpfc auxiliary inp idc shall be in the range of 0 to 1, inclusive. Values of 2 to 255, inclusive, for nnpfc auxiliary inp idc can be reserved for future use and shall not occur in the bitstream. Decoders shall ignore the NNPFC SEI message with nnpfc auxiliary inp idc in the range of 2 to 255, inclusive. Values of nnpfc auxiliary inp idc greater than 255 shall not occur in the bitstream and are not reserved for future use.

[0169] nnpfc inp order idc can indicate the method of ordering the sample arrays of the input pictures to form the input tensor of the NNPF. In the bitstream, the value of nnpfc inp order idc shall be in the range of 0 to 3, inclusive. Values of 4 to 255, inclusive, for nnpfc inp order idc can be reserved for future use and shall not occur in the bitstream. Decoders shall ignore the NNPFC SEI message with nnpfc inp order idc in the range of 4 to 255, inclusive. Values of nnpfc inp order idc greater than 255 shall not occur in the bitstream and are not reserved for future use. When ChromaFormatldc is not equal to 1, nnpfc inp order idc shall not be equal to 3. When ChromaFormatldc is equal to 0, nnpfc inp order idc shall be equal to 0. When chromaUpsamplingFlag is equal to 1, nnpfc inp order idc shall be equal to 0.

[0170] Table 8 shows the information description of nnpfc inp order idc values.

[0171] [Table 8]

[0172] nnpfc inp tensor luma bitdepth minus8 + 8 can specify the bit depth of luma sample values in the input integer tensor. The value of inpTensorBitDepthY is derived as shown in Equation 5.

[0173] [Equation 5]

[0174] A requirement for bitstream conformance is that the value of nnpfc inp tensor luma bitdepth minus8 shall be in the range of 0 to 24, inclusive.

[0175] nnpfc inp tensor chroma bitdepth minus8 + 8 can specify the bit depth of chroma sample values in the input integer tensor. The value of inpTensorBitDepthC can be derived as shown in Equation 6.

[0176] [Equation 6]

[0177] A requirement for bitstream conformance is that the value of nnpfc inp tensor chroma bitdepth minus8 shall be in the range of 0 to 24, inclusive.

[0178] When nnpfc auxiliary inp idc is equal to 1, the value of the variable strengthControlScaledVal can be derived as shown in Table 9.

[0179] [Table 9]

[0180] A tile is a rectangular array of samples of a component (e.g., luma or chroma component) from a picture.

[0181] A process DeriveInputTensor for deriving a given vertical sample coordinate cTop and horizontal sample coordinate cLeft for specifying the top-left sample position of a tile for an input tensor inputTensor can be specified as shown in the combination of Table 10 to Table 12.

[0182] [Table 10]

[0183] [Table 11]

[0184] [Table 12]

[0185] nnpfc_out_format_idc being equal to 0 can indicate that the sample values output through NNPF are real numbers, where the value range from 0 to 1 (including 0 and 1) is linearly mapped to the unsigned integer value range from 0 to (1 << bitDepth) - 1 (including 0 and (1 << bitDepth) - 1) for any desired bit depth bitDepth, for subsequent post - processing or display. nnpfc_out_format_idc being equal to 1 can indicate that the luminance sample values output by NNPF are unsigned integers within the range from 0 to (1 << outTensorBitDepth Y ) - 1 (including 0 and (1 << outTensorBitDepth Y ) - 1), and the chrominance sample values output by NNPF are unsigned integers within the range from 0 to (1 << outTensorBitDepth C ) - 1 (including 0 and (1 << outTensorBitDepth C ) - 1). Values of nnpfc_out_format_idc greater than are reserved for future use and shall not appear in the bitstream. The decoder shall ignore NNPFC SEI messages containing reserved values of nnpfc_out_format_idc.

[0186] nnpfc_out_order_idc can indicate the output order of the samples produced by NNPF. In the bitstream, the value of nnpfc_out_order_idc shall be within the range from 0 to 3 (including 0 and 3). Values from 4 to 255 (including 4 and 255) for nnpfc_out_order_idc are reserved for future use and shall not appear in the bitstream. The decoder shall ignore NNPFC SEI messages having nnpfc_out_order_idc within the range from 4 to 255 (including 4 and 255). Values of nnpfc_out_order_idc greater than 255 shall not appear in the bitstream and are not reserved for future use. When chromaUpsamplingFlag is equal to 1, nnpfc_out_order_idc shall not be equal to 0 or 3. When colourizationFlag is equal to 1, nnpfc_out_order_idc shall not be equal to 0.

[0187] Table 13 shows information specification of nnpfc_out_order_idc value.

[0188] [Table 13]

[0189] nnpfc_out_tensor_luma_bitdepth_minus8 + 8 can specify the bit depth of luma sample values in the output integer tensor. The value of nnpfc_out_tensor_luma_bitdepth_minus8 shall be in the range of 0 to 24, inclusive. outTensorBitDepth Y value can be derived as shown in Equation 7.

[0190] [Equation 7]

[0191] nnpfc_out_tensor_chroma_bitdepth_minus8 + 8 can specify the bit depth of chroma sample values in the output integer tensor. The value of nnpfc_out_tensor_chroma_bitdepth_minus8 shall be in the range of 0 to 24, inclusive. outTensorBitDepth C value can be derived as shown in Equation 8.

[0192] [Equation 8]

[0193] When bitDepthUpsamplingFlag is equal to 1, the value of nnpfc_out_format_idc shall be equal to 1 and at least one of the following conditions shall be true: - nnpfc_out_tensor_luma_bitdepth_minus8 + 8 is present and outTensorBitDepth Y is greater than BitDepth Y .

[0194] nnpfc_out_tensor_chroma_bitdepth_minus8 + 8 is present and outTensorBitDepth C is greater than BitDepth C .

[0195] When nnpfc inp tensor luma bitdepth minus8, nnpfc inp tensor chroma bitdepth minus8, nnpfc out tensor luma bitdepth minus8 and nnpfc out tensor chroma bitdepth minus8 are present and outTensorBitDepth Y is greater than inpTensorBitDepth Y , outTensorBitDepth C shall not be less than inpTensorBitDepth C . When nnpfc inp tensor luma bitdepth minus8, nnpfc inp tensor chroma bitdepth minus8, nnpfc out tensor luma bitdepth minus8 and nnpfc out tensor chroma bitdepth minus8 are present and outTensorBitDepth C is greater than inpTensorBitDepth C , outTensorBitDepth Y shall not be less than inpTensorBitDepth Y .

[0196] The process StoreOutputTensors( ) for deriving sample values in the filtered output sample arrays FilteredYPic, FilteredCbPic and FilteredCrPic from the output tensors specifying the top-left sample positions of the tiles for the samples included in the input tensors can be specified as shown in the combination of Table 14 and Table 15.

[0197] [Table 14]

[0198] [Table 15]

[0199] nnpfc_separate_colour_description_present_flag equal to 1 can indicate that a different combination of colour primaries, transfer characteristics, matrix coefficients and scaling and offset values associated with the matrix coefficients for pictures produced by the NNPF are specified in the SEI message syntax structure. nnpfc_separate_colour_description_present_flag equal to 0 indicates that the combination of colour primaries, transfer characteristics, matrix coefficients and scaling and offset values associated with the matrix coefficients for pictures produced by the NNPF are the same as specified in the VUI parameters for the CLVS.

[0200] nnpfc_colour_primaries can have the same semantics as defined for the vui_colour_primaries syntax element, with the following exceptions: - nnpfc_colour_primaries can specify the colour primaries of pictures produced by the NNPF as specified in the SEI message by the application, rather than the colour primaries used for the CLVS.

[0201] - When nnpfc_colour_primaries is not present in the NNPFC SEI message, the value of nnpfc_colour_primaries can be inferred to be equal to vui_colour_primaries.

[0202] nnpfc_transfer_characteristics can have the same semantics as defined for the vui_transfer_characteristics syntax element, with the following exceptions: - nnpfc_transfer_characteristics can specify the transfer characteristics of pictures produced by the NNPF as specified in the SEI message by the application, rather than the transfer characteristics used for the CLVS.

[0203] - When nnpfc_transfer_characteristics is not present in the NNPFC SEI message, the value of nnpfc_transfer_characteristics can be inferred to be equal to vui_transfer_characteristics.

[0204] nnpfc_matrix_coeffs can describe the equations used to derive the luma and chroma signals from the green, blue and red or Y, Z and X primary colors. The semantics of nnpfc_matrix_coeffs can be applied to pictures resulting from the NNPF specified by the application in the SEI message and BitDepthY and BitDepthC can be equal to outTensorBitDepthY and outTensorBitDepthC respectively as indicated for MatrixCoefficients. When nnpfc_matrix_coeffs is not present in the NNPFC SEI message, the value of nnpfc_matrix_coeffs can be inferred to be equal to vui_matrix_coeffs.

[0205] nnpfc_matrix_coeffs shall not be equal to 0 unless both of the following conditions are true: - nnpfc_out_tensor_chroma_bitdepth_minus8 is equal to nnpfc_out_tensor_luma_bitdepth_minus8.

[0206] - nnpfc_out_order_idc is equal to 2, outSubHeightC is equal to 1 and outSubWidthC is equal to 1.

[0207] nnpfc_matrix_coeffs shall not be equal to 8 unless one of the following conditions is true: - nnpfc_out_tensor_chroma_bitdepth_minus8 is equal to nnpfc_out_tensor_luma_bitdepth_minus8.

[0208] - nnpfc_out_tensor_chroma_bitdepth_minus8 is equal to nnpfc_out_tensor_luma_bitdepth_minus8 + 1, nnpfc_out_order_idc is equal to 2, outSubHeightC is equal to 1 and outSubWidthC is equal to 1.

[0209] nnpfc_full_range_flag can indicate scaling and offset values applied in association with matrix coefficients as specified by nnpfc_matrix_coeffs. The semantics of nnpfc_full_range_flag are the same as specified for VideoFullRangeFlag. When nnpfc_full_range_flag is not present, the value of nnpfc_full_range_flag can be inferred to be equal to 0.

[0210] nnpfc_chroma_loc_info_present_flag equal to 1 can indicate the presence of the nnpfc_chroma_sample_loc_type_frame syntax element in the NNPFC SEI message. nnpfc_chroma_loc_info_present_flag equal to 0 can indicate the absence of the nnpfc_chroma_sample_loc_type_frame syntax element in the NNPFC SEI message. When colourizationFlag is equal to 0 or nnpfc_out_colour_format_idc is not equal to 1, the value of nnpfc_chroma_loc_info_present_flag shall be equal to 0.

[0211] When nnpfc_chroma_sample_loc_type_frame is not equal to 6 and nnpfc_out_colour_format_idc is equal to 1, nnpfc_chroma_sample_loc_type_frame can specify the location of chroma samples of the output picture. nnpfc_chroma_sample_loc_type_frame equal to 6 and nnpfc_out_colour_format_idc equal to 1 can indicate that the location of chroma samples is unknown, unspecified, or has been specified by other means. The value of nnpfc_chroma_sample_loc_type_frame shall be in the range of 0 to 6, inclusive.

[0212] nnpfc_overlap can indicate the horizontal and vertical sample counts of the overlap of adjacent input tensors of the NNPFC. The value of nnpfc_overlap shall be in the range of 0 to 16383, inclusive.

[0213] nnpfc_constant_patch_size_flag equal to 1 can indicate that the NNPF accepts exactly the patch size indicated by nnpfc_patch_width_minus1 and nnpfc_patch_height_minus1 as input. nnpfc_constant_patch_size_flag equal to 0 can indicate that the NNPF accepts an arbitrary patch size with width inpPatchWidth and height inpPatchHeight as input, such that the width of the extended patch (i.e., the patch plus the overlap region) is equal to inpPatchWidth + 2 nnpfc_overlap, is a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 nnpfc_overlap, is a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 nnpfc_overlap, is a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 nnpfc_overlap, is a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2

[0214] When nnpfc_constant_patch_size_flag is equal to 1, nnpfc_patch_width_minus1 + 1 can indicate the horizontal sample count of the required patch size to the input to the NNPF. The value of nnpfc_patch_width_minus1 shall be in the range of 0 to Min(32766, CroppedWidth - 1), inclusive.

[0215] When nnpfc_constant_patch_size_flag is equal to 1, nnpfc_patch_height_minus1 + 1 can indicate the vertical sample count of the required patch size to the input to the NNPF. The value of nnpfc_patch_height_minus1 shall be in the range of 0 to Min(32766, CroppedHeight - 1), inclusive.

[0216] When nnpfc_constant_patch_size_flag is equal to 0, nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 nnpfc_overlap can indicate a common factor of all allowed values of the width of the extended patch required for the input to the NNPF. The value of nnpfc_extended_patch_width_cd_delta_minus1 shall be in the range of 0 to Min(32766, CroppedWidth - 1), inclusive.

[0217] When nnpfc_constant_patch_size_flag is equal to 0, nnpfc_extended_patch_height_cd_delta_minus1 + 1 + 2 nnpfc_overlap can indicate a common factor of all allowed values of the height of the extended patch required for the input to the NNPF. The value of nnpfc_extended_patch_height_cd_delta_minus1 shall be in the range of 0 to Min(32766, CroppedHeight - 1), inclusive.

[0218] The variables inpPatchWidth and inpPatchHeight can be set to the patch size width and patch size height, respectively.

[0219] If nnpfc_constant_patch_size_flag is equal to 0, the following applies: - The values of inpPatchWidth and inpPatchHeight can be provided by an external device or set by the post-processor itself.

[0220] - inpPatchWidth + 2 The value of nnpfc_overlap shall be a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 nnpfc_overlap is a positive integer multiple of nnpfc_extended_patch_width_cd_delta_minus1 + 1 + 2 The value of nnpfc overlap shall be nnpfc extended patch height cd delta minusl + 1 + 2 A positive integer multiple of nnpfc overlap, and inpPatchHeight shall be less than or equal to CroppedHeight.

[0221] Otherwise (nnpfc_constant_patch_size_flag is equal to 1), the value of inpPatchWidth can be set equal to nnpfc patch width minusl + 1, and the value of inpPatchHeight can be set equal to nnpfc patch height minusl + 1.

[0222] The variables outPatchWidth, outPatchHeight, horCScaling, verCScaling, outPatchCWidth, and outPatchCHeight can be derived as specified in Table 16.

[0223] [Table 16]

[0224] The requirement for bitstream conformance is that outPatchWidth CroppedWidth shall be equal to nnpfcOutputPicWidth inpPatchWidth and outPatchHeight CroppedHeight shall be equal to nnpfcOutputPicHeight inpPatchHeight.

[0225] nnpfc padding type can indicate the process of padding when referring to sample positions outside the boundaries of the reference input picture, as specified in Table 17. The value of nnpfc padding type shall be in the range of 0 to 4, inclusive. Values of 5 to 15, inclusive, for nnpfc padding type can be reserved for future use and shall not appear in the bitstream. Decoders shall ignore NNPFC SEI messages with nnpfc padding type in the range of 5 to 15, inclusive. Values of nnpfc padding type greater than 15 shall not appear in the bitstream and are not reserved for future use.

[0226] [Table 17]

[0227] When nnpfc_padding_type is equal to 4, nnpfc_luma_padding_val can indicate the luma value to be used for padding. The value of nnpfc_luma_padding_val shall be in the range of 0 to (1 « BitDepthY) - 1, inclusive.

[0228] When nnpfc_padding_type is equal to 4, nnpfc_cb_padding_val can indicate the Cb value to be used for padding. The value of nnpfc_cb_padding_val shall be in the range of 0 to (1 « BitDepthC) - 1, inclusive.

[0229] When nnpfc_padding_type is equal to 4, nnpfc_cr_padding_val can indicate the Cr value to be used for padding. The value of nnpfc_cr_padding_val shall be in the range of 0 to (1 « BitDepthC) - 1, inclusive.

[0230] A function InpSampleVal(y, x, picHeight, picWidth, croppedPic, cldx) with inputs vertical sample position y, horizontal sample position x, picture height picHeight, picture width picWidth, sample array croppedPic, and component index cldx (equal to 0 for luma, equal to 1 for Cb, and equal to 2 for Cr) can return the value of sampleVal derived as shown in Table 18.

[0231] [Table 18]

[0232] NNPF PostProcessingFilter() is the target NNPF derived as in the semantics of the NNPFA SEI message.

[0233] The process of Table 19 can be used by the NNPF PostProcessingFilter( ) to generate filtered and / or interpolated pictures in a tile-by-tile manner, which can contain Y, Cb and Cr sample arrays FilteredYPic, FilteredCbPic and FilteredCrPic, respectively, as indicated by nnpfc_out_order_idc: [Table 19]

[0234] An NNPF-generated picture with index i can contain sample arrays FilteredYPic[i], FilteredCbPic[i] and FilteredCrPic[i] (when present). An NNPF-generated picture can not include overlapping regions.

[0235] The NNPF process consists of the process defined in Table 19 immediately followed by outputting the NNPF-generated pictures in their increasing index order, where all NNPF-generated pictures interpolated by the NNPF are output, and those NNPF-generated pictures corresponding to any input pictures to the NNPF can be output, as specified in the semantics of the NNPFA SEI message.

[0236] nnpfc_complexity_info_present_flag equal to 1 can specify that one or more syntax elements indicating the complexity of the NNPF associated with nnpfc_id are present. nnpfc_complexity_info_present_flag equal to 0 can specify that no syntax elements indicating the complexity of the NNPF associated with nnpfc_id are present.

[0237] nnpfc_parameter_type_idc equal to 0 can indicate that the neural network uses only integer parameters. nnpfc_parameter_type_flag equal to 1 can indicate that the neural network can use floating-point or integer parameters. nnpfc_parameter_type_idc equal to 2 can indicate that the neural network uses only binary parameters. nnpfc_parameter_type_idc equal to 3 can be reserved for future use and shall not be present in the bitstream. Decoders shall ignore NNPFC SEI messages with nnpfc_parameter_type_idc equal to 3.

[0238] nnpfc_log2_parameter_bit_length_minus3 equal to 0, 1, 2, and 3 indicate that the neural network does not use parameters with bit length greater than 8, 16, 32, and 64, respectively. When nnpfc_parameter_type_idc is present and nnpfc_log2_parameter_bit_length_minus3 is not present, the neural network can not use parameters with bit length greater than 1.

[0239] nnpfc_num_parameters_idc can indicate the maximum number of neural network parameters for the NNPF in units of power of 2048. nnpfc_num_parameters_idc equal to 0 can indicate that the maximum number of neural network parameters is unknown. The value nnpfc_num_parameters_idc shall be in the range of 0 to 52, inclusive. Values of nnpfc_num_parameters_idc greater than 52 are reserved for future use and shall not appear in the bitstream. Decoders shall ignore NNPFC SEI messages with nnpfc_num_parameters_idc greater than 52.

[0240] If the value of nnpfc_num_parameters_idc is greater than zero, the variable maxNumParameters is derived as shown in Equation 9.

[0241] [Equation 9]

[0242] It is a requirement of bitstream conformance that the number of neural network parameters for the NNPF shall be less than or equal to maxNumParameters.

[0243] nnpfc_num_kmac_operations_idc greater than 0 can indicate that the maximum number of multiply-accumulate operations per sample of the NNPF is less than or equal to nnpfc_num_kmac_operations_idc 1000. nnpfc_num_kmac_operations_idc equal to 0 can indicate that the maximum number of multiply-accumulate operations of the network is unknown. The value of nnpfc_num_kmac_operations_idc shall be in the range of 0 to 2 32 - 2, inclusive. 32

[0244] ​nnpfc_total_kilobyte_size greater than 0 can indicate the total size in kilobytes required to store uncompressed parameters for a neural network. The total size in bits is equal to or greater than the total number of bits used to store each parameter. nnpfc_total_kilobyte_size is the total size in bits divided by 8000, rounded up. nnpfc_total_kilobyte_size equal to 0 can indicate that the total size required to store the parameters for a neural network is unknown. The value of nnpfc_total_kilobyte_size shall be in the range of 0 to 2 32 - 2, inclusive. 32 - 2, inclusive.

[0245] nnpfc_metadata_extension_num_bits equal to 0 can specify that nnpfc_reserved_metadata_extension is not present. nnpfc_metadata_extension_num_bits greater than 0 can specify the length in bits of nnpfc_reserved_metadata_extension. nnpfc_metadata_extension_num_bits shall be equal to 0. Values for nnpfc_metadata_extension_num_bits in the range of 1 to 2048, inclusive, are reserved for future use and shall not occur in the bitstream. Decoders shall allow any value of nnpfc_metadata_extension_num_bits in the range of 0 to 2048, inclusive. Values of nnpfc_metadata_extension_num_bits greater than 2048 shall not occur in the bitstream and can not be reserved for future use.

[0246] nnpfc_reserved_metadata_extension shall not occur in the bitstream. However, decoders shall ignore the presence of nnpfc_reserved_metadata_extension and the value of nnpfc_reserved_metadata_extension. When present, the length in bits of nnpfc_reserved_metadata_extension can be equal to nnpfc_metadata_extension_num_bits.

[0247] In the bitstream, nnpfc_reserved_zero_bit_b shall be equal to 0. Decoders shall ignore NNPFC SEI messages in which nnpfc_reserved_zero_bit_b is not equal to 0.

[0248] nnpfc_payload_byte[ i ] can contain the i-th byte of the bitstream. The byte sequence nnpfc_payload_byte[ i ] for all existing values of i shall be a conforming bitstream according to ISO / IEC 15938-17.

[0249] Neural network post filter activation (NNFPA)

[0250] The syntax structure for NNPFA is as shown in Table 20.

[0251] [Table 20]

[0252] The NNPFA syntax structure of Table 20 can be signaled in the form of an SEI message. The SEI message that signals the NNPFA syntax structure of Table 20 can be referred to as the NNPFA SEI message.

[0253] The neural network post-filter activation (NNPFA) SEI message can activate or deactivate the possible use of the target neural network post-filter (NNPF) identified by nnpfa_target_id for post-filtering of a set of pictures. For a particular picture for which the NNPF is activated, the target NNPF can be derived as follows: - If nnpfa_target_base_flag is equal to 1, the target NNPF is the base NNPF with nnpfc_id equal to nnpfa_target_id.

[0254] - Otherwise (nnpfa_target_base_flag is equal to 0), the target NNPF is the NNPF specified by the last NNPFC SEI message with nnpfc_id equal to nnpfa_target_id before the first VCL NAL unit of the current picture in decoding order and is not a repetition of the NNPFC SEI message containing the base NNPF.

[0255] For example, there can be several NNPFA SEI messages present for the same picture when the NNPF is a filter for a different purpose or for a different color component.

[0256] nnpfa_target_id can indicate the target NNPF, which is specified by one or more NNPFC SEI messages with nnpfc_id equal to nnpfa_target_id, with respect to the current picture. The value of nnpfa_target_id shall be in the range of 0 to 2 32 - 2 (inclusive 0 and 2 32 - 2).

[0257] Unless one or both of the following conditions are true, there shall not be an NNPFA SEI message with the special value of nnpfa_target_id in the current PU: - There is an NNPFC SEI message with the particular value of nnpfc_id equal to nnpfa_target_id present in a PU that precedes the current PU in decoding order within the current CLVS.

[0258] - There is an NNPFC SEI message with the particular value of nnpfc_id equal to nnpfa_target_id in the current PU.

[0259] When a PU contains both an NNPFC SEI message with the particular value of nnpfc_id and an NNPFA SEI message with nnpfa_target_id equal to the particular value of nnpfc_id, the NNPFC SEI message shall precede the NNPFA SEI message in decoding order.

[0260] nnpfa_cancel_flag equal to 1 can indicate that the persistence of the target NNPF established by any previous NNPFA SEI message with the same nnpfa_target_id as the current SEI message is cancelled, i.e., the target NNPF is no longer used unless it is activated by another NNPFA SEI message with the same nnpfa_target_id as the current SEI message and nnpfa_cancel_flag equal to 0. nnpfa_cancel_flag equal to 0 can indicate that nnpfa_target_base_flag, nnpfa_persistence_flag, and nnpfa_num_output_entries follow.

[0261] nnpfa_target_base_flag equal to 1 can specify that the target NNPF is the base NNPF with nnpfc id equal to nnpfa target id. nnpfa target base flag equal to 0 can specify that the target NNPF is the last NNPF specified by the last NNPFC SEI message with nnpfc id equal to nnpfa target id before the first VCL NAL unit of the current picture in decoding order and is not a repetition of the NNPFC SEI message containing the base NNPF.

[0262] nnpfa_persistence_flag can specify the persistence of the target NNPF for the current layer. nnpfa_persistence_flag equal to 0 can specify that the target NNPF can only be used for post-processing filtering of the current picture. nnpfa_persistence_flag equal to 1 can specify that the target NNPF can be used for post-processing filtering of the current picture and all subsequent pictures of the current layer in output order until one or more of the following conditions are true: - a new CLVS of the current layer starts.

[0263] - the bitstream ends.

[0264] - a picture in the current layer associated with an NNPFA SEI message with the same nnpfa target id as the current SEI message and nnpfa cancel flag equal to 1 is output that follows the current picture in output order.

[0265] - the target NNPF is not applied for the subsequent pictures in the current layer associated with an NNPFA SEI message with the same nnpfa target id as the current SEI message and nnpfa cancel flag equal to 1.

[0266] nnpfcTargetPictures can be the set of pictures that the last NNPFC SEI message with nnpfc id equal to nnpfa target id before the current NNPFA SEI message in decoding order is about. nnpfaTargetPictures can be the set of pictures for which the target NNPF is activated by the current NNPFA SEI message. The requirement for bitstream conformance is that any picture included in nnpfaTargetPictures shall also be included in nnpfcTargetPictures.

[0267] nnpfa_num_output_entries can specify the number of nnpfa_output_flag[ i ] syntax elements present in the NNPFA SEI message. The value of nnpfa_num_output_entries shall be in the range of 0 to NumInpPicsInOutputTensor, inclusive.

[0268] nnpfa_output_flag[ i ] equal to 1 can specify that a NNPF-generated picture corresponding to the input picture with index InpIdx[ i ] is output by the NNPF process activated by this NNPFA SEI message, where the NNPF process can be specified in the semantics of the NNPFC SEI message. nnpfa_output_flag[ i ] equal to 0 can specify that no NNPF-generated picture corresponding to the input picture with index InpIdx[ i ] is output by the NNPF process activated by this NNPFA SEI message. When nnpfa_num_output_entries is less than NumInpPicsInOutputTensor, for each value of i in the range of nnpfa_num_output_entries to NumInpPicsInOutputTensor - 1, inclusive, nnpfa_output_flag[ i ] can be inferred to be equal to 1.

[0269] Post filter hint

[0270] The syntax structure for the post-filter hint is shown in Table 21.

[0271] [Table 21]

[0272] The post-filter hint syntax structure of Table 21 can be signaled in the form of an SEI message. The SEI message that signals the post-filter hint syntax structure of Table 21 can be referred to as the post-filter hint SEI message.

[0273] This SEI message provides the coefficients of the post-filter or the correlation information for the design of the post-filter for potential use in the post-processing of the set of pictures after they are decoded and output for improved display quality.

[0274] filter_hint_cancel_flag equal to 1 can indicate that the SEI message cancels the persistence of any previous post-filter hint SEI messages applied to the current layer in output order. filter_hint_cancel_flag equal to 0 can indicate that the post-filter hint information follows.

[0275] filter_hint_persistence_flag can specify the persistence of the post-filter hint SEI message for the current layer. filter_hint_persistence_flag equal to 0 can specify that the post-filter hint applies only to the currently decoded picture. filter_hint_persistence_flag equal to 1 can specify that the post-filter hint SEI message applies to the currently decoded picture and persists for all subsequent pictures of the current layer in output order until one or more of the following conditions are true: - a new CLVS of the current layer starts.

[0276] - the end of the bitstream.

[0277] - the picture in the current layer in the AU associated with the post-filter hint SEI message is output that follows the current picture in output order.

[0278] filter_hint_size_y can specify the vertical size of the filter coefficients or related array. The value of filter_hint_size_y shall be in the range of 1 to 15, inclusive.

[0279] filter_hint_size_x can specify the horizontal size of the filter coefficients or related array. The value of filter_hint_size_x shall be in the range of 1 to 15, inclusive.

[0280] filter_hint_type can identify the type of filter hint sent as specified in Table 22. The value of filter_hint_type shall be in the range of 0 to 2, inclusive. The value of filter_hint_type equal to 3 can be reserved for future use and shall not appear in the bitstream. Decoders shall ignore post-filter hint SEI messages with filter_hint_type equal to 3.

[0281] [Table 22]

[0282] filter_hint_chroma_coeff_present_flag equal to 1 can specify that filter coefficients for chroma are present. filter_hint_chroma_coeff_present_flag equal to 0 can specify that filter coefficients for chroma are not present.

[0283] filter_hint_value[ cIdx ][ cy ][ cx ] can specify filter coefficients or elements of the cross-correlation matrix between the original signal and the decoded signal with a precision of 16 bits. The value of filter_hint_value[ cIdx ][ cy ][ cx ] shall be in the range of -2 31 + 1 to 2 31 - 1, including -2 31 + 1 and 2 31 - 1. cIdx can specify the related color component, cy denotes the counter in the vertical direction, and cx denotes the counter in the horizontal direction. Depending on the value of filter_hint_type, the following applies: - If filter_hint_type is equal to 0, coefficients of a two-dimensional Finite Impulse Response (FIR) filter with a size of filter_hint_size_y filter_hint_size_x can be transmitted.

[0284] - Otherwise, if filter_hint_type is equal to 1, filter coefficients of two one-dimensional FIR filters can be transmitted. In this case, filter_hint_size_y shall be equal to 2. The index cy equal to 0 specifies the filter coefficients of the horizontal filter, and cy equal to 1 specifies the filter coefficients of the vertical filter. In the filtering process, the horizontal filter is applied first, and the result is filtered by the vertical filter.

[0285] - Otherwise (filter_hint_type is equal to 2), the transmitted hint can specify the cross-correlation matrix between the original signal s and the decoded signal s’.

[0286] The normalized cross-correlation matrix for the related color component identified by cIdx with a size of filter_hint_size_y filter_hint_size_x can be defined as in Equation 10.

[0287] [Equation 10]

[0288] In Equation 10, s represents the sample array of the color component cIdx of the original image, s' represents the corresponding array of the decoded image, h represents the vertical height of the correlated color component, w represents the horizontal width of the correlated color component, bitDepth represents the bit depth of the color component, OffsetY equals (filter_hint_size_y >> 1), OffsetX equals (filter_hint_size_x >> 1), cy is in the range of 0 <= cy < filter_hint_size_y, and cx is in the range of 0 <= cx. <filter_hint_size_x。

[0289] The decoder can derive the Wiener post-filter based on the cross-correlation matrix of the original signal and the decoded signal, as well as the autocorrelation matrix of the decoded signal.

[0290] Problems with the related art

[0291] In the current design for NNPFC SEI and NNPFA SEI messages described above, the output image and its association with the input images are first signaled via the NNPFC SEI message. Then, the output image and its association with the input images are updated via the NNPFA SEI message that activates NNPFC. For example, nnpfc_num_input_pics_minus1, nnpfc_input_pic_output_flag[i], nnpfc_absent_input_pic_zero_flag, and nnpfc_interpolated_pics[i], and each of their semantics, can correspond to the current design for NNPFC SEI messages. Similarly, nnpfa_num_output_entries and nnpfa_output_flag[i], and each of their semantics, can correspond to the current design for NNPFA SEI messages.

[0292] In the signal transmission associated with the output image of the input image, there is a constraint established to ensure that at least one output image should be produced at the NNPF. However, when updating the signal transmission associated with the output image of the input image, there may be a problem where the activated NNPFA does not output any image.

[0293] Embodiments

[0294] This disclosure aims to solve the above-mentioned problems. Various embodiments proposed by this disclosure can be summarized as follows. The following embodiments can be performed individually, or two or more embodiments can be combined.

[0295] 1. When pictureRateUpsamplingFlag is equal to 0 and nnpfa_num_output_entries is equal to NumInpPicsInOutputTensor, nnpfa_output_flag[ i ] shall be equal to 1 for at least one value of i in the range of 0 to NumInpPicsInOutputTensor - 1, inclusive. Alternatively, when pictureRateUpsamplingFlag is equal to 0 and nnpfa_num_output_entries is equal to NumInpPicsInOutputTensor, nnpfa_output_flag[ i ] shall be equal to 1 for at least one value of i in the range of 0 to nnpfa_num_output_entries - 1, inclusive.

[0296] 2. When pictureRateUpsamplingFlag is equal to 0, nnpfa_output_flag[ i ] shall be equal to 1 for at least one value of i in the range of 0 to NumInpPicsInOutputTensor - 1, inclusive.

[0297] 3. When nnpfc_num_input_pics_minus1 is equal to 0, nnpfa_num_output_entries shall be equal to 0. In addition, when pictureRateUpsamplingFlag is equal to 0 and nnpfc_num_input_pics_minus1 is greater than 0, nnpfa_output_flag[ i ] shall be equal to 1 for at least one value of i in the range of 0 to NumInpPicsInOutputTensor - 1, inclusive.

[0298] 4. When nnpfc_num_input_pics_minus1 is equal to 0 and nnpfa_num_output_entries is equal to 1, nnpfa_output_flag[0] shall be equal to 1. In addition, when pictureRateUpsamplingFlag is equal to 0 and nnpfc_num_input_pics_minus1 is greater than 0, for at least one value of i in the range of 0 to NumInpPicsInOutputTensor - 1, inclusive, nnpfa_output_flag[i] shall be equal to 1.

[0299] Figure 5 is a flowchart illustrating an image encoding method according to an embodiment of the disclosure, and Figure 6 is a flowchart illustrating an image decoding method according to an embodiment of the disclosure. Figure 5 The image encoding method of Figure 6 The image decoding method of

[0300] Referring to Figure 5 , the first information can be encoded in an NNPFC SEI message (S510). The first information can indicate whether at least one output picture corresponding to at least one input picture is generated. The first information can be an nnpfc_input_pic_output_flag[i] syntax element.

[0301] The second information can be encoded in an NNPFA SEI message (S520). The second information can indicate whether the generated output picture is output. The second information can be an nnpfa_output_flag[i] syntax element.

[0302] The value of the second information (for example, nnpfa_output_flag[i]) can be determined based on at least one of whether the purpose of the NNPF is picture rate up-sampling (pictureRateUpsamplingFlag) or whether the number of input pictures is multiple (for example, nnpfc_num_input_pics_minus1). Here, "the value of the second information is determined" can mean "the value of the second information should be determined as a predetermined value".

[0303] Referring to Figure 6The first information can be obtained from the NNPFC SEI message (S610). The first information can indicate whether at least one output picture corresponding to at least one input picture is generated. The first information can be an nnpfc_input_pic_output_flag[i] syntax element.

[0304] The second information can be obtained from the NNPFA SEI message (S620). The second information can indicate whether the generated output picture is output. The second information can be an nnpfa_output_flag[i] syntax element.

[0305] The value of the second information (e.g., nnpfa_output_flag[i]) can be determined based on at least one of whether the purpose of the NNPF is picture rate up-sampling (pictureRateUpsamplingFlag) or whether the number of input pictures is multiple (e.g., nnpfc_num_input_pics_minus1). Here, "determining the value of the second information" can mean "the value of the second information should be determined as a predetermined value."

[0306] Hereinafter, various embodiments proposed by the disclosure will be specifically described. The embodiments proposed by the disclosure can be performed alone, or two or more embodiments can be combined and performed. In addition, the embodiments corresponding to the image encoding method can be performed by the image encoding apparatus 100, and the embodiments corresponding to the image decoding method can be performed by the image decoding apparatus 200.

[0307] Embodiment 1

[0308] Figure 7 is a flowchart for explaining an example of the image encoding / decoding method according to Embodiment 1.

[0309] Referring to Figure 7 It can be determined whether the purpose of the NNPF for the associated NNPFC is picture rate up-sampling and whether the number of pictures existing in the output tensor when having the corresponding input picture is equal to the number of second information in the NNPFA SEI message (S710). Here, whether the purpose of the NNPF is picture rate up-sampling can be determined based on the value of nnpfc_purpose & 0x08 or the value of pictureRateUpsamplingFlag. In addition, "the number of pictures existing in the output tensor when having the corresponding input picture" can be determined by a variable NumInpPicsInOutputTensor, and "the number of second information in the NNPFA SEI message" can be determined by an nnpfa_num_output_entries syntax element.

[0310] If the purpose of the NNPF for the associated NNPFC is not picture rate up-sampling (pictureRateUpsamplingFlag = 0) and there are as many pictures present in the output tensor as the number of second information in the NNPFA SEI message (nnpfa_num_output_entries = NumInpPicsInOutputTensor) when there are corresponding input pictures, the value of the second information (e.g., nnpfa_output_flag[i]) shall be equal to the first value (e.g., 1) indicating that the generated output picture is output (S720). Here, i can be at least one value in the range of 0 to nnpfa_num_output_entries - 1.

[0311] Embodiment 2

[0312] Figure 8 is a flowchart for explaining an example of the image encoding / decoding method according to Embodiment 2.

[0313] Referring to Figure 8 It can be determined whether the purpose of the NNPF for the associated NNPFC is picture rate up-sampling (S810). Here, it can be determined whether the purpose of the NNPF is picture rate up-sampling based on the value of nnpfc_purpose & 0x08 or the value of pictureRateUpsamplingFlag.

[0314] If the purpose of the NNPF for the associated NNPFC is not picture rate up-sampling (pictureRateUpsamplingFlag = 0), the value of the second information (e.g., nnpfa_output_flag[i]) shall be equal to the first value (e.g., 1) indicating that the generated output picture is output (S820). Here, i can be at least one value in the range of 0 to NumInpPicsInOutputTensor - 1.

[0315] Embodiment 3

[0316] Figure 9 and Figure 10 is a flowchart for explaining an example of the image encoding / decoding method according to Embodiment 3.

[0317] Referring to Figure 9The number of input pictures can be determined (S910). The number of input pictures can be determined by the nnpfc_num_input_pics_minus1 syntax element. If the number of input pictures is 1 (nnpfc_num_input_pics_minus1 = 0), the number of second information (e.g., nnpfa_output_flag[i]) should be 0 (S920, nnpfa_num_output_entries = 0). That is, if the number of input pictures is 1 (nnpfc_num_input_pics_minus1 = 0), the second information (e.g., nnpfa_output_flag[i]) can not be encoded in the NNPFA SEI message and can not be obtained from the NNPFA SEI message.

[0318] Referring to Figure 10 , whether the purpose of the NNPF is picture rate up-sampling and the number of input pictures can be determined (S1010). Here, whether the purpose of the NNPF is picture rate up-sampling can be determined based on the value of nnpfc_purpose & 0x08 or the value of pictureRateUpsamplingFlag. In addition, the "number of input pictures" can be determined by the nnpfc_num_input_pics_minus1 syntax element.

[0319] If the purpose of the NNPF is not picture rate up-sampling (pictureRateUpsamplingFlag = 0) and the number of input pictures is greater than 0 (nnpfc_num_input_pics_minus1 > 0), the value of the second information (e.g., nnpfa_output_flag[i]) should be equal to a first value (e.g., 1) indicating that the generated output picture is output (S1020). Here, i can be at least one value in the range of 0 to NumInpPicsInOutputTensor - 1.

[0320] Embodiment 4

[0321] Figure 11 and Figure 12 is a flowchart for explaining an example of an image encoding / decoding method according to Embodiment 4.

[0322] Referring to Figure 11The number of input pictures in the NNPFA SEI message and the number of second information can be determined (S1110). The number of input pictures can be determined by the nnpfc num input pics minusl syntax element. The number of second information in the NNPFA SEI message can be determined by the nnpfa num output entries syntax element.

[0323] When the number of input pictures is 1 (nnpfc num input pics minusl = 0) and the number of second information in the NNPFA SEI message is 1 (nnpfa num output entries = 1), the value of the first (i = 0) second information (e.g., nnpfa output flag[0]) should be equal to the first value (e.g., 1) indicating that the generated output picture is output (S1120). That is, the first output picture should be output.

[0324] Reference Figure 12 , whether the purpose of the NNPF is picture rate up-sampling and the number of input pictures can be determined (S1210). Here, whether the purpose of the NNPF is picture rate up-sampling can be determined based on the value of nnpfc purpose & 0x08 or the value of pictureRateUpsamplingFlag. In addition, the number of input pictures can be determined by the nnpfc num input pics minusl syntax element.

[0325] If the purpose of the NNPF is not picture rate up-sampling (pictureRateUpsamplingFlag = 0) and the number of input pictures is greater than 0 (nnpfc num input pics minusl > 0), the value of the second information (e.g., nnpfa output flag[i]) should be equal to the first value (e.g., 1) indicating that the generated output picture is output (S1220). Here, i can be at least one value in the range of 0 to NumInpPicsInOutputTensor - 1.

[0326] Figure 13 is a view showing a content streaming system to which embodiments of the present disclosure are applicable.

[0327] As Figure 13 indicated in the above-described embodiments, the content streaming system to which embodiments of the present disclosure are applied can mainly include an encoding server, a streaming server, a web server, a media store, a user device, and a multimedia input device.

[0328] The encoding server compresses content input from a multimedia input device such as a smartphone, a camera, a camcorder, etc. into digital data to generate a bitstream and transmits the bitstream to the streaming server. As another example, when a multimedia input device such as a smartphone, a camera, a camcorder, etc. directly generates a bitstream, the encoding server can be omitted.

[0329] The bitstream can be generated by an image encoding method or an image encoding apparatus to which embodiments of the present disclosure are applied, and the streaming server can temporarily store the bitstream in the process of transmitting or receiving the bitstream.

[0330] The streaming server transmits multimedia data to a user device based on a request of the user through a web server, and the web server serves as a medium for informing a service to the user. When the user requests a desired service from the web server, the web server can deliver it to the streaming server, and the streaming server can transmit multimedia data to the user. In this case, the content streaming system can include a separate control server. In this case, the control server serves to control commands / responses between devices in the content streaming system.

[0331] The streaming server can receive content from a media storage and / or an encoding server. For example, when receiving content from the encoding server, the content can be received in real time. In this case, in order to provide a smooth streaming service, the streaming server can store a bitstream for a predetermined time.

[0332] Examples of the user device can include a mobile phone, a smartphone, a laptop computer, a digital broadcasting terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, a tablet PC, a tablet computer, an ultrabook, a wearable device (e.g., a smartwatch, smart glasses, a head-mounted display), a digital TV, a desktop computer, a digital signage, etc.

[0333] Each server in the content streaming system can operate as a distributed server, in which case data received from each server can be distributed.

[0334] The scope of the present disclosure includes software or machine-executable commands (e.g., an operating system, an application, firmware, a program, etc.) for enabling operations of the methods according to various embodiments to be performed on an apparatus or a computer, or a non-transitory computer-readable medium having such software or commands stored thereon and executable on an apparatus or a computer.

[0335] Industrial applicability

[0336] Embodiments of the present disclosure can be used to encode or decode an image.

Claims

1. An image decoding method performed by an image decoding device, the image decoding method comprising: First information is obtained from the NNPFC (Neural Network Post Filter Feature) SEI (Supplementary Augmentation Information) message, the first information specifying whether at least one output image corresponding to at least one input image has been generated; as well as The second information, specifying whether the generated output image is output, is obtained from the NNPFA (Neural Network Post-Filter Activation) SEI message. The value of the second information is determined based on whether the purpose of NNPF (Neural Network Post-Filter) is image rate upsampling or whether the number of input images is at least one of multiple.

2. The image decoding method according to claim 1, wherein, Since the purpose of the NNPF is not image rate upsampling, the value of the second information is determined as a first value specifying that the generated output image is output.

3. The image decoding method according to claim 2, wherein, Furthermore, based on the fact that the number of images present in the output tensor when there is a corresponding input image is equal to the number of second information in the NNPFA SEI message, the value of the second information is determined to be the first value.

4. The image decoding method according to claim 2, wherein, Furthermore, based on the fact that there are multiple input images, the value of the second information is determined to be the first value.

5. The image decoding method according to claim 1, wherein, Based on the fact that the number of input images is one and the number of second information messages in the NNPFA SEI message is one, the value of the second information message is determined to be a first value specifying that the generated output image is output.

6. The image decoding method according to claim 1, wherein, Since the number of input images is one, the second information is not obtained from the NNPFA SEI message.

7. An image encoding method performed by an image encoding device, the image encoding method comprising: The first information is encoded into the NNPFC (Neural Network Post-Filter Feature) SEI (Supplementary Enhancement Information) message, which specifies whether at least one output image corresponding to at least one input image has been generated; and The second information is encoded into the NNPFA (Neural Network Post-Filter Activation) SEI message, which specifies whether the generated output image is output. The value of the second information is determined based on whether the purpose of NNPF (Neural Network Post-Filter) is image rate upsampling or whether the number of input images is at least one of multiple.

8. A computer-readable recording medium storing a bitstream generated by the image encoding method according to claim 7.

9. A method for transmitting a bitstream generated by an image encoding method, the image encoding method comprising: The first information is encoded into the NNPFC (Neural Network Post-Filter Feature) SEI (Supplementary Enhancement Information) message, which specifies whether at least one output image corresponding to at least one input image has been generated; and The second information is encoded into the NNPFA (Neural Network Post-Filter Activation) SEI message, which specifies whether the generated output image is output. The value of the second information is determined based on whether the purpose of NNPF (Neural Network Post-Filter) is image rate upsampling or whether the number of input images is at least one of multiple.

Images