KR20220063063A - A method and an apparatus for performing artificial intelligence encoding and artificial intelligence decoding - Google Patents

Abstract

Suggested is an AI encoding apparatus including a memory and a processor. The memory stores instructions executable by the processor. The processor acquires an original image, previously encoded frame information, and network environment information, acquires deblocking filter setting information based on the original image, the previously encoded frame information and the network environment information, applies deblocking filtering to the original image based on the deblocking filter setting information and acquires a first image AI-downscaled through a DNN for downscaling, generates image data by performing a first encoding operation on the first image, and transmits the image data and AI data including the deblocking filter setting information, and information related with the AI downscaling. Therefore, the present invention is capable of effectively encoding and decoding a high-resolution image based on AI.

Description

A method and an apparatus for performing artificial intelligence encoding and artificial intelligence decoding

The present disclosure relates to a method and apparatus for processing an image, and more particularly, the present disclosure relates to a method and apparatus for performing artificial intelligence (AI) encoding and artificial intelligence decoding.

An image is encoded by a codec conforming to a predetermined data compression standard, for example, a Moving Picture Expert Group (MPEG) standard, and then stored in a recording medium in the form of a bitstream or transmitted through a communication channel.

With the development and dissemination of hardware capable of reproducing and storing high-resolution/high-definition images, the need for a codec capable of effectively encoding and decoding high-resolution/high-definition images is increasing.

AI encoding and AI decoding method and apparatus of an image according to an embodiment predicts the occurrence of blocking artifacts in a variable and limited network environment in a real-time streaming service, further performs deblocking filtering, and downscaling and upscaling based on AI It is a technical task to scale and encode and decode an image.

In order to solve the above technical problem, the AI decoding apparatus proposed by the present disclosure includes: a memory for storing one or more instructions; and a processor executing the one or more instructions stored in the memory, wherein the processor comprises image data generated as a result of a first encoding of a first image and AI data related to AI downscaling from an original image to the first image , and deblocking filter setting information, first decoding the image data to obtain a second image corresponding to the first image, and DNN for AI upscaling of the second image among a plurality of DNN setting information Setting information is obtained based on the AI data, and AI is upscaled from the second image through a Deep Neural Network (DNN) for upscaling that operates with the obtained DNN setting information and based on the deblocking filter setting information Thus, a reconstructed image to which deblocking filtering is applied may be generated.

In order to solve the above technical problem, the AI decoding method proposed by the present disclosure includes image data generated as a result of a first encoding of a first image, AI data related to AI downscaling from an original image to the first image, and deblocking obtaining filter setting information; obtaining a second image corresponding to the first image by first decoding the image data; obtaining DNN setting information for AI upscaling of the second image from among a plurality of DNN setting information, based on the AI data; and generating a reconstructed image to which AI is upscaled from the second image through a deep neural network (DNN) for upscaling that operates with the obtained DNN setting information and to which deblocking filtering is applied based on the deblocking filter setting information. may include

In order to solve the above technical problem, an AI encoding apparatus proposed by the present disclosure includes: a memory for storing one or more instructions; a processor executing the one or more instructions stored in the memory, wherein the processor obtains an original image, previously coded frame information, and network environment information, the original image, the previously coded frame information, and acquiring deblocking filter setting information based on the network environment information, applying deblocking filtering to the original image based on the deblocking filter setting information, and AI downscaled first image through DNN for downscaling acquisition, generating image data by first encoding the first image, and transmitting the deblocking filter setting information, AI data including information related to the AI downscaling, and the image data.

In order to solve the above technical problem, the AI encoding method proposed by the present disclosure includes: obtaining an original image, previously encoded frame information, and network environment information; obtaining deblocking filter setting information based on the original image, the previously encoded frame information, and the network environment information; applying deblocking filtering to the original image based on the deblocking filter setting information and obtaining an AI downscaled first image through a DNN for downscaling; generating image data by first encoding the first image; The method may include transmitting the deblocking filter setting information, AI data including information related to the AI downscaling, and the image data.

During encoding, deblocking filter setting information is obtained based on the original image, previous frame information, and network environment information, deblocking filtering is applied to the original image based on the deblocking filter setting information, and deblocking-filtered original image AI downscales to transmit the downscaled image and deblocking filter setting information, AI upscales the downscaled image transmitted during decoding to obtain an upscaled image, and based on the transmitted deblocking filter setting information By applying deblocking filtering to the upscaled image to obtain a reconstructed image, a high-resolution image can be effectively encoded and decoded based on AI even when the network situation suddenly changes.

In order to more fully understand the drawings cited herein, a brief description of each drawing is provided.
1 is a diagram for describing an AI encoding process and an AI decoding process according to an embodiment.
2 is a block diagram illustrating a configuration of an AI decoding apparatus according to an embodiment.
3 is an exemplary diagram illustrating a second DNN for AI upscaling of a second image.
4 is a diagram for explaining a convolution operation using a convolution layer.
5 is an exemplary diagram illustrating a mapping relationship between various pieces of image-related information and various pieces of DNN configuration information.
6 is a diagram illustrating a second image composed of a plurality of frames.
7 is a block diagram illustrating a configuration of an AI encoding apparatus according to an embodiment.
8 is an exemplary diagram illustrating a first DNN for AI downscaling of an original video.
9 is a diagram illustrating a structure of AI encoded data according to an embodiment.
10 is a diagram illustrating a structure of AI encoded data according to another embodiment.
11 is a diagram for explaining a method of training a first DNN and a second DNN.
12 is a diagram for explaining a training process of a first DNN and a second DNN by a training apparatus.
13 is a diagram for explaining an AI encoding process and an AI decoding process for adding deblocking filtering according to an embodiment.
14 is a block diagram illustrating a configuration of an AI decoding apparatus according to an embodiment.
15 is a block diagram illustrating a configuration of an AI encoding apparatus according to an embodiment.
16 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.
17 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.
18 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.
19 is an exemplary diagram illustrating a third DNN and a fourth DNN for a deblocking filter based on AI.
20 is a diagram for explaining a method of training the third DNN 2010 and the fourth DNN 2020. Referring to FIG.
21 is a diagram for explaining a training process of the third DNN 2010 and the fourth DNN 2020 by the training apparatus 2100 .
22 is a flowchart illustrating an AI encoding method according to an embodiment.
23 is a flowchart illustrating an AI decoding method according to an embodiment.

Since the present disclosure can make various changes and can have various embodiments, specific embodiments are illustrated in the drawings and will be described in detail through the detailed description. However, this is not intended to limit the embodiments of the present disclosure, and it should be understood that the present disclosure includes all modifications, equivalents, and substitutes included in the spirit and scope of various embodiments.

In describing the embodiment, if it is determined that a detailed description of a related known technology may unnecessarily obscure the subject matter of the present disclosure, the detailed description thereof will be omitted. In addition, numbers (eg, first, second, etc.) used in the description process of the specification are only identification symbols for distinguishing one component from other components.

In addition, in this specification, when a component is referred to as “connected” or “connected” with another component, the component may be directly connected or directly connected to the other component, but in particular It should be understood that, unless there is a description to the contrary, it may be connected or connected through another element in the middle.

In addition, in the present specification, components expressed as '~ part (unit)', 'module', etc. are two or more components combined into one component, or two or more components for each more subdivided function. may be differentiated into In addition, each of the components to be described below may additionally perform some or all of the functions of other components in addition to the main functions they are responsible for, and some of the main functions of each component may have different functions. It goes without saying that it may be performed exclusively by the component.

In addition, in this specification, an 'image' or 'picture' may indicate a still image, a moving picture composed of a plurality of continuous still images (or frames), or a video.

In addition, in the present specification, 'deep neural network (DNN)' is a representative example of an artificial neural network model simulating a brain nerve, and is not limited to an artificial neural network model using a specific algorithm.

Also, in the present specification, a 'parameter' is a value used in a calculation process of each layer constituting a neural network, and may include, for example, a weight used when an input value is applied to a predetermined calculation expression. Also, the parameter may be expressed in a matrix form. A parameter is a value set as a result of training, and may be updated through separate training data if necessary.

In addition, in this specification, 'first DNN' means a DNN used for AI downscaling of an image, and 'second DNN' means a DNN used for AI upscaling of an image.

Also, in the present specification, 'DNN configuration information' includes the aforementioned parameters as information related to elements constituting the DNN. The first DNN or the second DNN may be configured using the DNN configuration information.

In addition, in this specification, the 'original image' refers to an image to be subjected to AI encoding, and the 'first image' refers to an image obtained as a result of AI downscaling of the original image in the AI encoding process. In addition, the 'second image' refers to an image obtained through the first decoding in the AI decoding process, and the 'third image' refers to an image obtained by AI upscaling the second image in the AI decoding process.

In addition, in this specification, 'AI downscaling' refers to processing of reducing the resolution of an image based on AI, and 'first encoding' refers to encoding processing by a frequency transform-based image compression method. In addition, 'first decoding' means a decoding process by a frequency conversion-based image restoration method, and 'AI upscaling' means a process of increasing the resolution of an image based on AI.

1 is a diagram for describing an artificial intelligence (AI) encoding process and an AI decoding process according to an embodiment.

As described above, as the resolution of an image rapidly increases, the amount of information processing for encoding/decoding increases. Accordingly, a method for improving encoding and decoding efficiency of an image is required.

As shown in FIG. 1 , according to an embodiment of the present disclosure, a first image 115 is acquired by AI downscaling 110 of an original image 105 having a high resolution. In addition, since the first encoding 120 and the first decoding 130 are performed on the first image 115 of a relatively small resolution, the first encoding 120 and Compared to the case of performing the first decoding 130, the bit rate may be greatly reduced.

1 , in an embodiment, in the AI encoding process, the original image 105 is AI downscaled 110 to obtain a first image 115 , and the first image 115 is first Encoding (120). In the AI decoding process, AI encoding data including AI data and image data obtained as a result of AI encoding is received, a second image 135 is obtained through the first decoding 130, and the second image 135 is A third image 145 is obtained by AI upscaling 140 .

Looking at the AI encoding process in more detail, when the original image 105 is input, the original image 105 is AI downscaled 110 to obtain the first image 115 of a predetermined resolution and/or a predetermined image quality. At this time, the AI downscaling 110 is performed based on AI, the AI for the AI downscaling 110 must be trained in connection with the AI for the AI upscaling 140 of the second image 135 (joint trained) do. Because, when the AI for the AI downscaling 110 and the AI for the AI upscaling 140 are separately trained, between the original image 105 that is the AI encoding target and the third image 145 restored through AI decoding because the difference between

In an embodiment of the present disclosure, AI data may be used in order to maintain this linkage in the AI encoding process and the AI decoding process. Therefore, the AI data obtained through the AI encoding process should include information indicating the upscale target, and in the AI decoding process, the second image 135 is AI upscaled ( 140) should be done.

AI for AI downscaling 110 and AI for AI upscaling 140 may be implemented as a deep neural network (DNN). As will be described later with reference to FIG. 11 , since the first DNN and the second DNN are jointly trained through sharing loss information under a predetermined target, the AI encoding apparatus uses target information used when the first DNN and the second DNN are jointly trained. is provided to the AI decoding apparatus, and the AI decoding apparatus may upscale the AI to the image quality and/or resolution targeting the second image 135 based on the received target information.

When the first encoding 120 and the first decoding 130 shown in FIG. 1 are described in detail, the first image 115 AI downscaled 110 from the original image 105 is the first encoding 120 . can reduce the amount of information. The first encoding 120 includes a process of predicting the first image 115 to generate prediction data, a process of generating residual data corresponding to a difference between the first image 115 and the prediction data, and a spatial domain component. It may include a process of transforming the residual data into a frequency domain component, a process of quantizing the residual data transformed into a frequency domain component, and a process of entropy encoding the quantized residual data. Such a first encoding process 120 is MPEG-2, H.264 Advanced Video Coding (AVC), MPEG-4, High Efficiency Video Coding (HEVC), VC-1, VP8, VP9 and AV1 (AOMedia Video 1). It may be implemented through one of the image compression methods using frequency transformation, etc.

The second image 135 corresponding to the first image 115 may be restored through the first decoding 130 of image data. The first decoding 130 includes a process of generating quantized residual data by entropy-decoding image data, a process of inverse quantizing the quantized residual data, a process of transforming the residual data of a frequency domain component into a spatial domain component, a process of predicting data It may include a process of generating , and a process of reconstructing the second image 135 using prediction data and residual data. The first decoding 130 process as described above is an image compression using frequency conversion such as MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9 and AV1 used in the first encoding 120 process. It may be implemented through an image restoration method corresponding to one of the methods.

The AI encoded data obtained through the AI encoding process may include image data obtained as a result of the first encoding 120 of the first image 115 and AI data related to the AI downscale 110 of the original image 105. can The image data may be used in the first decoding 130 process, and the AI data may be used in the AI upscaling 140 process.

The image data may be transmitted in the form of a bitstream. The image data may include data obtained based on pixel values in the first image 115 , for example, residual data that is a difference between the first image 115 and prediction data of the first image 115 . . Also, the image data includes information used in the process of the first encoding 120 of the first image 115 . For example, the image data includes prediction mode information used for the first encoding 120 of the first image 115 , motion information, and quantization parameter related information used in the first encoding 120 , etc. can do. Video data is a video compression method used in the first encoding 120 process among video compression methods using frequency conversion such as MPEG-2, H.264 AVC, MPEG-4, HEVC, VC-1, VP8, VP9 and AV1. It may be generated according to a rule, for example, a syntax.

AI data is used for AI upscaling 140 based on the second DNN. As described above, since the first DNN and the second DNN are jointly trained, the AI data includes information enabling accurate AI upscaling 140 of the second image 135 through the second DNN to be performed. . In the AI decoding process, the AI upscaling 140 may be performed to a resolution and/or image quality targeting the second image 135 based on the AI data.

AI data may be transmitted together with image data in the form of a bitstream. According to an embodiment, AI data may be transmitted separately from image data in the form of frames or packets.

Alternatively, according to an embodiment, AI data may be transmitted while being included in image data.

The image data and AI data may be transmitted through the same network or different networks.

2 is a block diagram showing the configuration of the AI decoding apparatus 200 according to an embodiment.

Referring to FIG. 2 , the AI decoding apparatus 200 according to an embodiment includes a receiving unit 210 and an AI decoding unit 230 . The AI decoding unit 230 may include a parsing unit 232 , a first decoding unit 234 , an AI upscaling unit 236 , and an AI setting unit 238 .

Although the receiver 210 and the AI decoder 230 are illustrated as separate devices in FIG. 2 , the receiver 210 and the AI decoder 230 may be implemented through one processor. In this case, the receiving unit 210 and the AI decoding unit 230 may be implemented as a dedicated processor, and a general-purpose processor such as an application processor (AP), a central processing unit (CPU), or a graphic processing unit (GPU) and S/W It may be implemented through a combination of Also, in the case of a dedicated processor, a memory for implementing an embodiment of the present disclosure or a memory processing unit for using an external memory may be included.

The receiver 210 and the AI decoder 230 may include a plurality of processors. In this case, it may be implemented as a combination of dedicated processors, or may be implemented through a combination of S/W with a plurality of general-purpose processors such as an AP, CPU, or GPU. In one embodiment, the receiving unit 210 is implemented as a first processor, the first decoding unit 234 is implemented as a second processor different from the first processor, the parsing unit 232, the AI upscaling unit 236 and the AI setting unit 238 may be implemented as a third processor different from the first processor and the second processor.

The receiving unit 210 receives AI encoded data obtained as a result of AI encoding. As an example, the AI-encoded data may be a video file having a file format such as mp4 or mov.

The receiver 210 may receive AI-encoded data transmitted through a network. The receiving unit 210 outputs the AI encoded data to the AI decoding unit 230 .

In one embodiment, AI-encoded data is stored in a hard disk, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floppy disks. It may be obtained from a data storage medium including an optical medium).

The parsing unit 232 parses the AI encoded data and transmits image data generated as a result of the first encoding of the first image 115 to the first decoder 234 , and transmits the AI data to the AI setting unit 238 . transmit

In an embodiment, the parsing unit 232 may parse the image data and the AI data separately included in the AI encoded data. The parsing unit 232 may read the header in the AI-encoded data to distinguish the AI data and the image data included in the AI-encoded data. In one example, AI data may be included in a Vendor Specific InfoFrame (VSIF) in the HDMI stream.

The structure of AI-encoded data including separated AI data and image data will be described later with reference to FIG. 9 .

In another embodiment, the parsing unit 232 parses the image data from the AI encoded data, extracts the AI data from the image data, transmits the AI data to the AI setting unit 238, and first decodes the remaining image data. may be transmitted to the unit 234 . That is, AI data may be included in image data. For example, AI data may be included in supplemental enhancement information (SEI), which is an additional information area of a bitstream corresponding to image data. The structure of AI-encoded data including image data including AI data will be described later with reference to FIG. 10 .

In another embodiment, the parsing unit 232 divides a bitstream corresponding to image data into a bitstream to be processed by the first decoding unit 234 and a bitstream corresponding to AI data, and divides each of the divided bitstreams into a bitstream corresponding to AI data. It may output to the first decoding unit 234 and the AI setting unit 238 .

The parsing unit 232 obtains image data included in the AI encoded data through a predetermined codec (eg, MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9, or AV1). It can also be confirmed that it is the image data that has been processed. In this case, the corresponding information may be transmitted to the first decoder 234 so that the image data can be processed by the identified codec.

The first decoder 234 reconstructs the second image 135 corresponding to the first image 115 based on the image data received from the parser 232 . The second image 135 obtained by the first decoding unit 234 is provided to the AI upscaling unit 236 .

According to an embodiment, first decoding-related information such as prediction mode information, motion information, and quantization parameter information may be provided from the first decoding unit 234 to the AI setting unit 238 . The first decoding-related information may be used to obtain DNN configuration information.

The AI data provided to the AI setting unit 238 includes information enabling AI upscaling of the second image 135 . In this case, the upscale target of the second image 135 should correspond to the downscale target of the first DNN. Therefore, the AI data should include information that can confirm the downscale target of the first DNN.

Specifically exemplifying the information included in the AI data, there are information about a difference between the resolution of the original image 105 and the resolution of the first image 115 , and information related to the first image 115 .

The difference information may be expressed as information about a resolution conversion degree of the first image 115 compared to the original image 105 (eg, resolution conversion rate information). In addition, since the resolution of the first image 115 is known through the resolution of the reconstructed second image 135 and the degree of resolution conversion can be checked through this, the difference information can be expressed only with the resolution information of the original image 105 . may be Here, the resolution information may be expressed as a horizontal/vertical screen size, or may be expressed as a ratio (16:9, 4:3, etc.) and a size of one axis. Also, when there is preset resolution information, it may be expressed in the form of an index or a flag.

In addition, the information related to the first image 115 includes the resolution of the first image 115 , the bit rate of image data obtained as a result of the first encoding of the first image 115 , and the first encoding of the first image 115 . It may include information on at least one of the codec types used at the time of operation.

The AI setting unit 238 may determine an upscale target of the second image 135 based on at least one of difference information included in the AI data and information related to the first image 115 . The upscaling target may indicate, for example, to which resolution the second image 135 should be upscaled. When the upscaling target is determined, the AI upscaling unit 236 AI upscales the second image 135 through the second DNN to obtain a third image 145 corresponding to the upscaling target.

Prior to explaining a method for the AI setting unit 238 to determine an upscaling target based on AI data, an AI upscaling process through the second DNN will be described with reference to FIGS. 3 and 4 .

3 is an exemplary diagram illustrating a second DNN 300 for AI upscaling of the second image 135, and FIG. 4 is a convolution operation in the first convolution layer 310 shown in FIG. is showing

3 , the second image 135 is input to the first convolutional layer 310 . 3X3X4 displayed in the first convolution layer 310 shown in FIG. 3 exemplifies convolution processing on one input image using four filter kernels having a size of 3×3. As a result of the convolution process, four feature maps are generated by four filter kernels. Each feature map represents unique characteristics of the second image 135 . For example, each feature map may indicate a vertical direction characteristic, a horizontal direction characteristic, or an edge characteristic of the second image 135 .

A convolution operation in the first convolution layer 310 will be described in detail with reference to FIG. 4 .

Through the multiplication operation and addition operation between the parameters of the filter kernel 430 having a size of 3 X 3 used in the first convolution layer 310 and the pixel values in the second image 135 corresponding thereto, one A feature map 450 may be generated. Since four filter kernels are used in the first convolution layer 310 , four feature maps may be generated through a convolution operation process using the four filter kernels.

In FIG. 4 , I1 to I49 displayed on the second image 135 indicate pixels of the second image 135 , and F1 to F9 displayed on the filter kernel 430 indicate parameters of the filter kernel 430 . Also, M1 to M9 displayed in the feature map 450 indicate samples of the feature map 450 .

4 illustrates that the second image 135 includes 49 pixels, but this is only an example, and when the second image 135 has a resolution of 4K, for example, 3840 X 2160 pixels It may contain pixels.

In the convolution operation process, each of the pixel values of I1, I2, I3, I8, I9, I10, I15, I16, and I17 of the second image 135 and F1, F2, F3, F4, and F5 of the filter kernel 430 is performed. , F6, F7, F8, and F9 may each be multiplied, and a value obtained by combining (eg, addition operation) result values of the multiplication operation may be assigned as the value of M1 of the feature map 450 . If the stride of the convolution operation is 2, pixel values of I3, I4, I5, I10, I11, I12, I17, I18, and I19 of the second image 135 and F1 and F2 of the filter kernel 430, respectively , F3, F4, F5, F6, F7, F8, and F9 may each be multiplied, and a value obtained by combining result values of the multiplication operation may be assigned as the value of M2 of the feature map 450 .

A convolution operation between the pixel values in the second image 135 and the parameters of the filter kernel 430 is performed while the filter kernel 430 moves along the stride until the last pixel of the second image 135 is reached. By being performed, a feature map 450 having a predetermined size may be obtained.

According to the present disclosure, parameters of the second DNN through joint training of the first DNN and the second DNN, for example, parameters of the filter kernel used in convolutional layers of the second DNN (eg, filter The values of F1, F2, F3, F4, F5, F6, F7, F8, and F9 of the kernel 430 may be optimized. The AI setting unit 238 determines an upscale target corresponding to the downscale target of the first DNN based on the AI data, and uses parameters corresponding to the determined upscale target in the convolution layers of the second DNN. It can be determined by parameters of the filter kernel.

The convolution layers included in the first DNN and the second DNN may be processed according to the convolution operation process described in relation to FIG. 4 , but the convolution operation process described in FIG. 4 is only an example, and is limited thereto it is not

Referring back to FIG. 3 , the feature maps output from the first convolutional layer 310 are input to the first activation layer 320 .

The first activation layer 320 may provide a non-linear characteristic to each feature map. The first activation layer 320 may include, but is not limited to, a sigmoid function, a Tanh function, a Rectified Linear Unit (ReLU) function, and the like.

Giving the nonlinear characteristic in the first activation layer 320 means changing and outputting some sample values of the feature map, which is an output of the first convolution layer 310 . At this time, the change is performed by applying a non-linear characteristic.

The first activation layer 320 determines whether to transfer sample values of the feature maps output from the first convolution layer 310 to the second convolution layer 330 . For example, some sample values among the sample values of the feature maps are activated by the first activation layer 320 and transmitted to the second convolution layer 330 , and some sample values are activated by the first activation layer 320 . It is deactivated and is not transmitted to the second convolutional layer 330 . The intrinsic characteristic of the second image 135 indicated by the feature maps is emphasized by the first activation layer 320 .

The feature maps 325 output from the first activation layer 320 are input to the second convolution layer 330 . Any one of the feature maps 325 shown in FIG. 3 is a result of processing the feature map 450 described with reference to FIG. 4 in the first activation layer 320 .

3X3X4 displayed in the second convolution layer 330 exemplifies convolution processing on the input feature maps 325 using four filter kernels having a size of 3×3. The output of the second convolutional layer 330 is input to the second activation layer 340 . The second activation layer 340 may impart a nonlinear characteristic to the input data.

The feature maps 345 output from the second activation layer 340 are input to the third convolution layer 350 . 3X3X1 displayed in the third convolution layer 350 shown in FIG. 3 exemplifies convolution processing to generate one output image using one filter kernel having a size of 3×3. The third convolution layer 350 is a layer for outputting a final image and generates one output using one filter kernel. According to the example of the present disclosure, the third convolution layer 350 may output the third image 145 through a convolution operation.

DNN setting information indicating the number of filter kernels and parameters of the filter kernels of the first convolutional layer 310, the second convolutional layer 330, and the third convolutional layer 350 of the second DNN 300 is As will be described later, there may be a plurality of pieces, and a plurality of pieces of DNN configuration information should be associated with a plurality of pieces of DNN configuration information of the first DNN. The association between the plurality of DNN configuration information of the second DNN and the plurality of DNN configuration information of the first DNN may be implemented through association learning of the first DNN and the second DNN.

3 shows that the second DNN 300 includes three convolutional layers 310 , 330 , 350 and two activation layers 320 , 340 , but this is only an example, and the implementation Accordingly, the number of convolutional layers and activation layers may be variously changed. Also, according to an embodiment, the second DNN 300 may be implemented through a recurrent neural network (RNN). In this case, it means changing the CNN structure of the second DNN 300 according to the example of the present disclosure to the RNN structure.

In an embodiment, the AI upscaling unit 236 may include at least one arithmetic logic unit (ALU) for the above-described convolution operation and operation of the activation layer. The ALU may be implemented as a processor. For the convolution operation, the ALU includes a multiplier that performs a multiplication operation between the sample values of the feature map output from the second image 135 or the previous layer and the sample values of the filter kernel, and an adder that adds the result values of the multiplication. can In addition, for the operation of the activation layer, the ALU is a multiplier that multiplies an input sample value by a weight used in a predetermined sigmoid function, a Tanh function, or a ReLU function, and compares the multiplication result with a predetermined value to calculate the input sample value It may include a comparator that determines whether to transfer to the next layer.

Hereinafter, a method in which the AI setting unit 238 determines the upscaling target and the AI upscaling unit 236 AI upscales the second image 135 according to the upscaling target will be described.

In an embodiment, the AI setting unit 238 may store a plurality of DNN setting information settable in the second DNN.

Here, the DNN configuration information may include information on at least one of the number of convolutional layers included in the second DNN, the number of filter kernels for each convolutional layer, and parameters of each filter kernel. The plurality of DNN configuration information may respectively correspond to various upscale targets, and the second DNN may operate based on the DNN configuration information corresponding to a specific upscale target. The second DNN may have a different structure according to the DNN configuration information. For example, the second DNN may include three convolutional layers according to certain DNN configuration information, and the second DNN may include four convolutional layers according to other DNN configuration information.

In an embodiment, the DNN configuration information may include only parameters of a filter kernel used in the second DNN. In this case, the structure of the second DNN is not changed, but only the parameters of the internal filter kernel may be changed according to the DNN configuration information.

The AI setting unit 238 may acquire DNN setting information for AI upscaling of the second image 135 among a plurality of DNN setting information. Each of the plurality of DNN configuration information used herein is information for acquiring the third image 145 of a predetermined resolution and/or predetermined image quality, and is trained in association with the first DNN.

For example, the DNN setting information of any one of the plurality of DNN setting information is the third image 145 having a resolution twice that of the second image 135, for example, the second image of 2K (2048 * 1080). It may include information for acquiring the third image 145 of 4K (4096*2160) that is twice as large as the second image 135 , and the other DNN setting information is 4 more than the resolution of the second image 135 . Information for obtaining a third image 145 of twice the resolution, for example, a third image 145 of 8K (8192 * 4320) that is 4 times larger than the second image 135 of 2K (2048 * 1080) may include

Each of the plurality of DNN setting information is created in association with the DNN setting information of the first DNN of the AI encoding device 700, and the AI setting unit 238 is set at an enlargement ratio corresponding to the reduction rate of the DNN setting information of the first DNN. Accordingly, one piece of DNN configuration information among a plurality of DNN configuration information is acquired. To this end, the AI setting unit 238 should check the information of the first DNN. In order for the AI setting unit 238 to check the information of the first DNN, the AI decoding apparatus 200 according to an embodiment receives AI data including the information of the first DNN from the AI encoding apparatus 700 .

In other words, the AI setting unit 238 uses the information received from the AI encoding device 700 to identify information targeted by the DNN setting information of the first DNN used to obtain the first image 115 , and , it is possible to obtain the DNN setting information of the second DNN trained in association with it.

When DNN setting information for AI upscaling of the second image 135 is obtained among a plurality of DNN setting information, the DNN setting information is transmitted to the AI upscaling unit 236, and a second DNN operating according to the DNN setting information. Based on the input data may be processed.

For example, when any one DNN configuration information is obtained, the AI upscaling unit 236 includes the first convolutional layer 310 and the second convolutional layer 330 of the second DNN 300 shown in FIG. 3 . ) and the third convolutional layer 350, the number of filter kernels included in each layer and parameters of the filter kernels are set to values included in the obtained DNN configuration information.

Specifically, the AI upscaling unit 236 determines that the parameters of the 3 X 3 filter kernel used in any one convolutional layer of the second DNN shown in FIG. 4 are {1, 1, 1, 1, 1, 1 , 1, 1, 1}, and when there is a change in DNN configuration information, the parameters of the filter kernel are changed to {2, 2, 2, 2, 2, 2, 2, 2, 2, which are parameters included in the changed DNN configuration information. } can be replaced with

The AI setting unit 238 may acquire DNN setting information for upscaling the second image 135 among a plurality of DNN setting information based on the information included in the AI data, which is used to obtain the DNN setting information. AI data will be explained in detail.

In an embodiment, the AI setting unit 238 may acquire DNN setting information for upscaling the second image 135 among a plurality of DNN setting information based on the difference information included in the AI data. For example, based on the difference information, the resolution of the original image 105 (eg, 4K (4096 * 2160)) is higher than the resolution of the first image 115 (eg, 2K (2048 * 1080)). When it is confirmed that it is twice as large, the AI setting unit 238 may acquire DNN setting information capable of increasing the resolution of the second image 135 by two times.

In another embodiment, the AI setting unit 238 is based on the information related to the first image 115 included in the AI data, DNN setting information for AI upscaling the second image 135 among a plurality of DNN setting information. can be obtained. The AI setting unit 238 may determine a mapping relationship between the image related information and the DNN setting information in advance, and obtain DNN setting information mapped to the first image 115 related information.

5 is an exemplary diagram illustrating a mapping relationship between various pieces of image-related information and various pieces of DNN configuration information.

5, it can be seen that the AI encoding/AI decoding process according to an embodiment of the present disclosure does not only consider a change in resolution. As shown in FIG. 5, DNN setting information considering resolutions such as SD, HD, Full HD, bitrates such as 10Mbps, 15Mbps, and 20Mbps, and codec information such as AV1, H.264, and HEVC individually or all selection can be made. For this consideration, training in consideration of each element in the AI training process should be performed in connection with the encoding and decoding processes (see FIG. 11 ).

Accordingly, as shown in FIG. 5 according to the training content, when a plurality of DNN setting information is provided based on image-related information including a codec type and image resolution, the first image ( 115) DNN setting information for AI upscaling of the second image 135 may be acquired based on the related information.

That is, the AI setting unit 238 matches the image-related information shown on the left side of the table shown in FIG. 5 with the DNN setting information on the right side of the table, so that the DNN setting information according to the image-related information can be used.

5 , from the information related to the first image 115 , the resolution of the first image 115 is SD, the bit rate of image data obtained as a result of the first encoding of the first image 115 is 10Mbps, , when it is confirmed that the first image 115 is first coded with the AV1 codec, the AI setting unit 238 may obtain A DNN setting information among a plurality of DNN setting information.

In addition, from the information related to the first image 115, the resolution of the first image 115 is HD, the bit rate of image data obtained as a result of the first encoding is 15Mbps, and the first image 115 is converted to H.264 codec. When it is confirmed that the first encoding is performed, the AI setting unit 238 may acquire B DNN setting information among a plurality of DNN setting information.

Also, from the information related to the first image 115 , the resolution of the first image 115 is Full HD, the bit rate of image data obtained as a result of the first encoding of the first image 115 is 20 Mbps, and the first image ( 115) is confirmed to be first encoded with the HEVC codec, the AI setting unit 238 obtains C DNN setting information among a plurality of DNN setting information, the resolution of the first image 115 is Full HD, and the first When the bit rate of the image data obtained as a result of the first encoding of the image 115 is 15 Mbps and it is confirmed that the first image 115 is first encoded with the HEVC codec, the AI setting unit 238 sets a plurality of DNN setting information Among them, DNN configuration information can be obtained. Either one of the C DNN setting information and the D DNN setting information is selected according to whether the bit rate of the image data obtained as a result of the first encoding of the first image 115 is 20 Mbps or 15 Mbps. When the first image 115 of the same resolution is first encoded with the same codec, the fact that the bit rates of the image data are different from each other means that the image quality of the reconstructed images is different from each other. Accordingly, the first DNN and the second DNN may be jointly trained based on a predetermined image quality, and accordingly, the AI setting unit 238 sets the DNN according to the bit rate of the image data indicating the image quality of the second image 135 . information can be obtained.

In another embodiment, the AI setting unit 238 relates to the information (prediction mode information, motion information, quantization parameter information, etc.) provided from the first decoder 234 and the first image 115 included in the AI data. DNN configuration information for AI upscaling of the second image 135 among a plurality of pieces of DNN configuration information may be acquired in consideration of all information. For example, the AI setting unit 238 receives quantization parameter information used in the first encoding process of the first image 115 from the first decoder 234 , and receives the quantization parameter information from the AI data of the first image 115 . The bit rate of the image data obtained as a result of encoding may be checked, and DNN setting information corresponding to the quantization parameter and the bit rate may be obtained. Even at the same bit rate, there may be a difference in the quality of the reconstructed image depending on the complexity of the image. The bit rate is a value representative of the entire first image 115 to be first encoded, and the bit rate of each frame in the first image 115 is The picture quality may be different. Accordingly, when the prediction mode information, motion information, and/or quantization parameters obtainable from the first decoder 234 for each frame are considered together, the DNN setting more suitable for the second image 135 than using only AI data information can be obtained.

Also, depending on the implementation, the AI data may include an identifier of mutually agreed DNN configuration information. The identifier of the DNN setting information is an upscale target corresponding to the downscale target of the first DNN, and the DNN setting information trained in association between the first DNN and the second DNN so that the second image 135 can be AI upscaled. Information for distinguishing pairs of The AI setting unit 238 obtains the identifier of the DNN setting information included in the AI data, and then obtains the DNN setting information corresponding to the identifier of the DNN setting information, and the AI upscaling unit 236 receives the corresponding DNN setting information. It can be used to upscale the second image 135 by AI. For example, an identifier indicating each of a plurality of pieces of DNN configuration information configurable in the first DNN and an identifier indicating each of a plurality of pieces of DNN configuration information configurable in the second DNN may be predefined. In this case, the same identifier may be assigned to a pair of DNN configuration information configurable to each of the first DNN and the second DNN. The AI data may include an identifier of DNN setting information set in the first DNN for AI downscaling of the original image 105 . The AI setting unit 238 that has received the AI data acquires DNN setting information indicated by an identifier included in the AI data among a plurality of DNN setting information, and the AI upscaling unit 236 uses the corresponding DNN setting information to obtain the second The image 135 may be upscaled by AI.

Also, according to implementation, AI data may include DNN configuration information. The AI setting unit 238 may obtain DNN setting information included in AI data, and the AI upscaling unit 236 may AI upscale the second image 135 using the corresponding DNN setting information.

According to the embodiment, when information constituting the DNN setting information (eg, the number of convolution layers, the number of filter kernels for each convolutional layer, parameters of each filter kernel, etc.) is stored in the form of a lookup table, The AI setting unit 238 obtains DNN setting information by combining some selected among the lookup table values based on information included in the AI data, and the AI upscaling unit 236 uses the corresponding DNN setting information to obtain the second image (135) may be AI upscaled.

According to an embodiment, when the structure of the DNN corresponding to the upscale target is determined, the AI setting unit 238 may obtain DNN setting information corresponding to the determined structure of the DNN, for example, parameters of the filter kernel.

As described above, the AI setting unit 238 obtains DNN setting information of the second DNN through AI data including information related to the first DNN, and the AI upscaling unit 236 sets the DNN setting information to the corresponding DNN setting information. AI upscales the second image 135 through the second DNN, which can reduce memory usage and computational amount compared to upscaling by directly analyzing the characteristics of the second image 135 .

In an embodiment, when the second image 135 is composed of a plurality of frames, the AI setting unit 238 may independently acquire DNN setting information for each frame of a predetermined number, or a common DNN for all frames. It is also possible to obtain setting information.

6 is a diagram illustrating a second image 135 composed of a plurality of frames.

As shown in FIG. 6 , the second image 135 may include frames corresponding to t0 to tn.

In one example, the AI setting unit 238 obtains DNN setting information of the second DNN through AI data, and the AI upscaling unit 236 AI frames corresponding to t0 to tn based on the corresponding DNN setting information. can be upscaled. That is, frames corresponding to t0 to tn may be AI upscaled based on common DNN configuration information.

In another example, the AI setting unit 238 obtains 'A' DNN setting information from AI data for some of the frames corresponding to t0 to tn, for example, frames corresponding to t0 to ta. and 'B' DNN configuration information can be obtained from AI data for frames corresponding to ta+1 to tb. Also, the AI setting unit 238 may acquire 'C' DNN setting information from AI data for frames corresponding to tb+1 to tn. In other words, the AI setting unit 238 independently obtains DNN setting information for each group including a predetermined number of frames among the plurality of frames, and the AI upscaling unit 236 independently selects the frames included in each group. AI upscaling is possible with the obtained DNN setting information.

In another example, the AI setting unit 238 may independently obtain DNN setting information for each frame constituting the second image 135 . For example, when the second image 135 consists of three frames, the AI setting unit 238 obtains DNN setting information in relation to the first frame, and DNN setting information in relation to the second frame. and may acquire DNN configuration information in relation to the third frame. That is, DNN configuration information may be independently obtained for each of the first frame, the second frame, and the third frame. DNN setting information is obtained based on information (prediction mode information, motion information, quantization parameter information, etc.) provided from the above-described first decoder 234 and information related to the first image 115 included in AI data According to the method, DNN setting information may be independently obtained for each frame constituting the second image 135 . This is because mode information, quantization parameter information, and the like can be independently determined for each frame constituting the second image 135 .

In another example, the AI data may include information indicating to which frame the DNN setting information obtained based on the AI data is valid. For example, if the AI data includes information that the DNN setting information is valid until the ta frame, the AI setting unit 238 acquires the DNN setting information based on the AI data, and the AI upscaling unit 236 is AI upscales t0 to ta frames with corresponding DNN configuration information. And, when the information that the DNN setting information is valid until the tn frame is included in the other AI data, the AI setting unit 238 acquires the DNN setting information based on the other AI data, and the AI upscaling unit 236 may AI upscale ta+1 to tn frames with the obtained DNN configuration information.

Hereinafter, an AI encoding apparatus 700 for AI encoding of the original image 105 will be described with reference to FIG. 7 .

7 is a block diagram illustrating a configuration of an AI encoding apparatus 700 according to an embodiment.

Referring to FIG. 7 , the AI encoding apparatus 700 may include an AI encoding unit 710 and a transmission unit 730 . The AI encoder 710 may include an AI downscaler 712 , a first encoder 714 , a data processor 716 , and an AI setting unit 718 .

7 illustrates the AI encoder 710 and the transmitter 730 as separate devices, the AI encoder 710 and the transmitter 730 may be implemented through one processor. In this case, it may be implemented as a dedicated processor, or may be implemented through a combination of an AP or a general-purpose processor such as a CPU or GPU and S/W. Also, in the case of a dedicated processor, a memory for implementing an embodiment of the present disclosure or a memory processing unit for using an external memory may be included.

Also, the AI encoder 710 and the transmitter 730 may include a plurality of processors. In this case, it may be implemented as a combination of dedicated processors, or may be implemented through a combination of S/W with a plurality of general-purpose processors such as an AP, CPU, or GPU.

In an embodiment, the first encoding unit 714 is configured as a first processor, and the AI downscale unit 712 , the data processing unit 716 and the AI setting unit 718 are configured as a second processor different from the first processor. The transmitter 730 may be implemented as a third processor different from the first processor and the second processor. The AI encoder 710 performs AI downscale of the original image 105 and the first image 115 . performs the first encoding of , and transmits the AI encoded data to the transmission unit 730 . The transmitter 730 transmits the AI encoded data to the AI decoding apparatus 200 .

The image data includes data obtained as a result of the first encoding of the first image 115 . The image data may include data obtained based on pixel values in the first image 115 , for example, residual data that is a difference between the first image 115 and prediction data of the first image 115 . . In addition, the image data includes information used in the first encoding process of the first image 115 . For example, the image data may include prediction mode information used for first encoding the first image 115 , motion information, and quantization parameter related information used for first encoding the first image 115 , etc. .

The AI data includes information that enables the AI upscaling unit 236 to AI upscale the second image 135 to an upscale target corresponding to the downscale target of the first DNN. In one example, the AI data may include difference information between the original image 105 and the first image 115 . In an example, the AI data may include information related to the first image 115 . The information related to the first image 115 is used for the resolution of the first image 115 , the bit rate of image data obtained as a result of the first encoding of the first image 115 , and the first encoding of the first image 115 . information on at least one of the specified codec types may be included.

In an embodiment, the AI data may include an identifier of mutually agreed DNN configuration information so that the second image 135 can be AI upscaled to an upscale target corresponding to the downscale target of the first DNN. .

Also, in an embodiment, the AI data may include DNN configuration information settable in the second DNN.

The AI downscaler 712 may acquire the AI downscaled first image 115 from the original image 105 through the first DNN. The AI downscaling unit 712 may AI downscale the original image 105 using the DNN setting information provided from the AI setting unit 718 . The AI setting unit 718 may determine a downscale target of the original image 105 based on a predetermined criterion.

In order to acquire the first image 115 matching the downscale target, the AI setting unit 718 may store a plurality of settable DNN setting information in the first DNN. The AI setting unit 718 obtains DNN setting information corresponding to the downscale target among a plurality of DNN setting information, and provides the obtained DNN setting information to the AI downscale unit 712 .

Each of the plurality of pieces of DNN configuration information may be trained to acquire the first image 115 having a predetermined resolution and/or a predetermined image quality. For example, the DNN setting information of any one of the plurality of DNN setting information is the first image 115 having a resolution that is 1/2 times smaller than the resolution of the original image 105, for example, 4K (4096 * 2160) may include information for obtaining the first image 115 of 2K (2048 * 1080) that is 1/2 times smaller than the original image 105 of A first image 115 of a resolution of 1/4 times smaller than that of a first image 115, for example, a first image 115 of 2K (2048*1080) which is 1/4 times smaller than an original image 105 of 8K (8192*4320). ) may include information for obtaining

According to the embodiment, when information constituting the DNN setting information (eg, the number of convolution layers, the number of filter kernels for each convolutional layer, parameters of each filter kernel, etc.) is stored in the form of a lookup table, The AI setting unit 718 may provide the DNN setting information obtained by combining some selected among the lookup table values to the AI downscaling unit 712 according to the downscaling target.

According to an embodiment, the AI setting unit 718 may determine the structure of the DNN corresponding to the downscale target, and obtain DNN setting information corresponding to the determined structure of the DNN, for example, parameters of the filter kernel.

The plurality of DNN configuration information for AI downscaling of the original image 105 may have an optimized value by performing joint training of the first DNN and the second DNN. Here, each DNN configuration information includes at least one of the number of convolutional layers included in the first DNN, the number of filter kernels for each convolutional layer, and parameters of each filter kernel.

The AI downscaling unit 712 sets the first DNN as the DNN setting information determined for the AI downscaling of the original image 105, and converts the first image 115 of a predetermined resolution and/or a predetermined quality to the first DNN. can be obtained through When DNN setting information for AI downscaling of the original image 105 is obtained among a plurality of DNN setting information, each layer in the first DNN may process input data based on information included in the DNN setting information. .

Hereinafter, a method for the AI setting unit 718 to determine the downscale target will be described. The downscale target may indicate, for example, how much the first image 115 with reduced resolution should be obtained from the original image 105 .

The AI setting unit 718 acquires one or more pieces of input information. In an embodiment, the input information includes a target resolution of the first image 115, a target bitrate of the image data, a bitrate type of the image data (eg, a variable bitrate type, a constant bitrate type, or an average bitrate type, etc.); Color format to which AI downscale is applied (luminance component, chrominance component, red component, green component, or blue component, etc.), codec type for the first encoding of the first image 115, compression history information, and the original image 105 may include at least one of a resolution of , and a type of the original image 105 .

The one or more pieces of input information may include information previously stored in the AI encoding apparatus 700 or received from a user.

The AI setting unit 718 controls the operation of the AI downscale unit 712 based on the input information. In an embodiment, the AI setting unit 718 may determine a downscale target according to the input information, and provide DNN setting information corresponding to the determined downscale target to the AI downscale unit 712 .

In an embodiment, the AI setting unit 718 transmits at least a portion of the input information to the first encoder 714 so that the first encoder 714 can select a specific value of a bitrate, a specific type of bitrate, and a specific codec. may cause the first image 115 to be first encoded.

In an embodiment, the AI setting unit 718 sets the compression rate (eg, the difference in resolution between the original image 105 and the first image 115 , the target bit rate), and the compression quality (eg, the bit rate type). ), compression history information, and a downscale target may be determined based on at least one of the type of the original image 105 .

In one example, the AI setting unit 718 may determine a downscale target based on a preset or a compression rate or compression quality input from a user.

As another example, the AI setting unit 718 may determine the downscale target by using the compression history information stored in the AI encoding apparatus 700 . For example, according to the compression history information available to the AI encoding apparatus 700, the encoding quality or compression rate preferred by the user may be determined, and the downscale target may be determined according to the encoding quality determined based on the compression history information. can For example, the resolution and quality of the first image 115 may be determined according to the encoding quality that has been most frequently used according to the compression history information.

As another example, the AI setting unit 718 may set the encoding quality that has been used more than a predetermined threshold value (eg, average quality of encoding qualities that have been used more than a predetermined threshold value) according to the compression history information. Based on the downscale target may be determined.

As another example, the AI setting unit 718 may determine the downscale target based on the resolution and type (eg, file format) of the original image 105 .

In one embodiment, when the original image 105 is composed of a plurality of frames, the AI setting unit 718 independently acquires DNN setting information for each predetermined number of frames, and uses the independently acquired DNN setting information to AI downscaling. A portion 712 may be provided.

In one example, the AI setting unit 718 may divide the frames constituting the original image 105 into a predetermined number of groups, and independently obtain DNN setting information for each group. The same or different DNN configuration information may be obtained for each group. The number of frames included in the groups may be the same or different for each group.

In another example, the AI setting unit 718 may independently determine DNN setting information for each frame constituting the original image 105 . The same or different DNN configuration information may be obtained for each frame.

Hereinafter, an exemplary structure of the first DNN 800 as a basis for AI downscaling will be described.

FIG. 8 is an exemplary diagram illustrating a first DNN 800 for AI downscaling of an original image 105 .

As shown in FIG. 8 , the original image 105 is input to the first convolutional layer 810 . The first convolutional layer 810 performs convolution processing on the original image 105 using 32 filter kernels having a size of 5×5. The 32 feature maps generated as a result of the convolution process are input to the first activation layer 820 . The first activation layer 820 may provide non-linear characteristics to 32 feature maps.

The first activation layer 820 determines whether to transfer sample values of the feature maps output from the first convolution layer 810 to the second convolution layer 830 . For example, some sample values among the sample values of the feature maps are activated by the first activation layer 820 and transmitted to the second convolution layer 830 , and some sample values are activated by the first activation layer 820 . It is deactivated and is not transmitted to the second convolution layer 830 . Information indicated by the feature maps output from the first convolutional layer 810 is emphasized by the first activation layer 820 .

The output 825 of the first activation layer 820 is input to the second convolution layer 830 . The second convolution layer 830 performs convolution processing on input data using 32 filter kernels having a size of 5×5. The 32 feature maps output as a result of the convolution process are input to the second activation layer 840 , and the second activation layer 840 may provide nonlinear characteristics to the 32 feature maps.

An output 845 of the second activation layer 840 is input to the third convolutional layer 850 . The third convolution layer 850 performs convolution processing on input data using one filter kernel having a size of 5×5. As a result of the convolution processing, one image may be output from the third convolutional layer 850 . The third convolutional layer 850 is a layer for outputting a final image and obtains one output by using one filter kernel. According to the example of the present disclosure, the third convolution layer 850 may output the first image 115 through the convolution operation result.

DNN setting information indicating the number of filter kernels and parameters of the filter kernels of the first convolutional layer 810, the second convolutional layer 830, and the third convolutional layer 850 of the first DNN 800 is There may be a plurality, and the plurality of DNN configuration information should be associated with the plurality of DNN configuration information of the second DNN. The association between the plurality of DNN configuration information of the first DNN and the plurality of DNN configuration information of the second DNN may be implemented through association learning of the first DNN and the second DNN.

8 illustrates that the first DNN 800 includes three convolutional layers 810 , 830 , 750 and two activation layers 820 , 840 , but this is only an example, and the implementation Accordingly, the number of convolutional layers and activation layers may be variously changed. Also, according to an embodiment, the first DNN 800 may be implemented through a recurrent neural network (RNN). In this case, it means changing the CNN structure of the first DNN 800 according to the example of the present disclosure to the RNN structure.

In an embodiment, the AI downscaling unit 712 may include at least one ALU for a convolution operation and an activation layer operation. The ALU may be implemented as a processor. For the convolution operation, the ALU may include a multiplier that performs a multiplication operation between the sample values of the feature map output from the original image 105 or the previous layer and the sample values of the filter kernel, and an adder that adds the result values of the multiplication. there is. In addition, for the operation of the activation layer, the ALU is a multiplier that multiplies an input sample value by a weight used in a predetermined sigmoid function, a Tanh function, or a ReLU function, and compares the multiplication result with a predetermined value to calculate the input sample value It may include a comparator that determines whether to transfer to the next layer.

Referring back to FIG. 7 , the AI setting unit 718 transmits AI data to the data processing unit 716 . The AI data includes information that enables the AI upscaling unit 236 to AI upscale the second image 135 to an upscale target corresponding to the downscale target of the first DNN. The first encoder 714 receiving the first image 115 from the AI downscaler 712 first encodes the first image 115 according to the frequency transformation-based image compression method to obtain the first image 115 . ) can reduce the amount of information it has. As a result of the first encoding through a predetermined codec (eg, MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9, or AV1), image data is obtained. The image data is obtained according to a rule of a predetermined codec, that is, a syntax. For example, the image data includes residual data that is a difference between the first image 115 and the prediction data of the first image 115 , prediction mode information used to first encode the first image 115 , motion information, and Information related to a quantization parameter used to first encode the first image 115 may be included. The image data obtained as a result of the first encoding by the first encoder 714 is provided to the data processor 716 .

The data processing unit 716 generates AI encoded data including the image data received from the first encoding unit 714 and the AI data received from the AI setting unit 718 .

In an embodiment, the data processing unit 716 may generate AI-encoded data including image data and AI data in a separate state. As an example, AI data may be included in a Vendor Specific InfoFrame (VSIF) in the HDMI stream.

In another embodiment, the data processing unit 716 may include AI data in image data obtained as a result of the first encoding by the first encoding unit 714 and generate AI encoded data including the corresponding image data. . For example, the data processing unit 716 may generate image data in the form of one bitstream by combining a bitstream corresponding to image data and a bitstream corresponding to AI data. To this end, the data processing unit 716 may express AI data as bits having a value of 0 or 1, that is, a bitstream. In an embodiment, the data processing unit 716 may include a bitstream corresponding to AI data in supplemental enhancement information (SEI), which is an additional information area of a bitstream obtained as a result of the first encoding.

AI-encoded data is transmitted to the transmitter 730 . The transmission unit 730 transmits the AI-encoded data obtained as a result of the AI-encoding through a network. In an embodiment, the AI-encoded data is a hard disk, a magnetic medium such as a floppy disk and a magnetic tape, and a CD-ROM and a DVD. It may be stored in a data storage medium including an optical recording medium, a magneto-optical medium such as a floptical disk, and the like.

9 is a diagram illustrating a structure of AI-encoded data 900 according to an embodiment.

As described above, the AI data 912 and the image data 932 may be separately included in the AI encoded data 900 . Here, the AI-encoded data 900 may be in a container format such as MP4, AVI, MKV, or FLV. The AI-encoded data 900 may include a metadata box 910 and a media data box 930 .

The metadata box 910 includes information about the image data 932 included in the media data box 930 . For example, the metadata box 910 may include information about the type of the first image 115 , the type of codec used to encode the first image 115 , and the playback time of the first image 115 . there is. Also, AI data 912 may be included in the metadata box 910 . The AI data 912 may be encoded according to an encoding method provided by a predetermined container format and stored in the metadata box 910 .

The media data box 930 may include image data 932 generated according to the syntax of a predetermined image compression method.

10 is a diagram illustrating a structure of AI-encoded data 1000 according to another embodiment.

Referring to FIG. 10 , AI data 1034 may be included in image data 1032 . The AI-encoded data 1000 may include a metadata box 1010 and a media data box 1030 . When the AI data 1034 is included in the image data 1032 , the AI data box 1010 contains the AI data. Data 1034 may not be included.

The media data box 1030 includes image data 1032 including AI data 1034 . For example, the AI data 1034 may be included in the additional information area of the image data 1032 .

Hereinafter, a method for jointly training the first DNN 800 and the second DNN 300 will be described with reference to FIG. 11 .

11 is a diagram for explaining a method of training the first DNN 800 and the second DNN 300 .

In an embodiment, the AI-encoded original image 105 through the AI encoding process is restored to the third image 145 through the AI decoding process. The third image 145 and the original image 105 obtained as a result of AI decoding In order to maintain the similarity with the AI encoding process and the AI decoding process, correlation is required. That is, information lost in the AI encoding process should be able to be restored in the AI decoding process, and for this purpose, joint training between the first DNN 800 and the second DNN 300 is required.

For accurate AI decoding, it is ultimately necessary to reduce the quality loss information 1130 corresponding to the comparison result between the third training image 1104 and the original training image 1101 shown in FIG. 11 . Accordingly, the quality loss information 1130 is used for both training of the first DNN 800 and the second DNN 300 .

First, the training process shown in FIG. 11 will be described.

In FIG. 11 , an original training image 1101 is an image subject to AI downscale, and a first training image 1102 is an AI downscaled image from the original training image 1101 . it's a video Also, a third training image 1104 is an AI upscaled image from the first training image 1102 .

The original training image 1101 includes a still image or a moving image composed of a plurality of frames. In an embodiment, the original training image 1101 may include a still image or a luminance image extracted from a moving picture including a plurality of frames. Also, according to an embodiment, the original training image 1101 may include a still image or a patch image extracted from a moving picture including a plurality of frames. When the original training image 1101 consists of a plurality of frames, the first training image 1102, the second training image, and the third training image 1104 also include a plurality of frames. When a plurality of frames of the original training image 1101 are sequentially input to the first DNN 800 , the first training image 1102 and the second training image are performed through the first DNN 800 and the second DNN 300 . and a plurality of frames of the third training image 1104 may be sequentially acquired.

For joint training of the first DNN 800 and the second DNN 300 , an original training image 1101 is input to the first DNN 800 . The original training image 1101 input to the first DNN 800 is AI downscaled and output as the first training image 1102 , and the first training image 1102 is input to the second DNN 300 . A third training image 1104 is output as a result of AI upscaling for the first training image 1102 .

Referring to FIG. 11 , a first training image 1102 is being input to the second DNN 300 . According to an embodiment, the first training image 1102 is obtained through a first encoding and a first decoding process. A second training image may be input to the second DNN 300 . In order to input the second training image to the second DNN, any one of MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9, and AV1 codecs may be used. Specifically, in the first encoding of the first training image 1102 and the first decoding of image data corresponding to the first training image 1102, MPEG-2, H.264, MPEG-4, HEVC, VC-1 , VP8, VP9, and any one codec of AV1 may be used.

Referring to FIG. 11 , the legacy downscaled reduced training image 1103 is obtained from the original training image 1101 separately from the output of the first training image 1102 through the first DNN 800 . Here, the legacy downscale may include at least one of a bilinear scale, a bicubic scale, a lanczos scale, and a stair step scale.

In order to prevent the structural features of the first image 115 from greatly deviating based on the structural features of the original image 105, a reduced training image 1103 that preserves the structural features of the original training image 1101 is acquired. .

Before the training proceeds, the first DNN 800 and the second DNN 300 may be set with predetermined DNN configuration information. As training progresses, structural loss information 1110 , complexity loss information 1120 , and quality loss information 1130 may be determined.

The structural loss information 1110 may be determined based on a comparison result between the reduced training image 1103 and the first training image 1102 . In one example, the structural loss information 1110 may correspond to a difference between the structural information of the reduced training image 1103 and the structural information of the first training image 1102 . The structural information may include various features that can be extracted from the image, such as luminance, contrast, and histogram of the image. The structural loss information 1110 indicates to what extent the structural information of the original training image 1101 is maintained in the first training image 1102 . As the structural loss information 1110 is smaller, the structural information of the first training image 1102 is similar to the structural information of the original training image 1101 .

The complexity loss information 1120 may be determined based on the spatial complexity of the first training image 1102 . In an example, as the spatial complexity, a total variance value of the first training image 1102 may be used. The complexity loss information 1120 is related to a bit rate of image data obtained by first encoding the first training image 1102 . The smaller the complexity loss information 1120 is, the smaller the bit rate of image data is defined.

The quality loss information 1130 may be determined based on a comparison result between the original training image 1101 and the third training image 1104 . The quality loss information 1130 includes an L1-norm value, an L2-norm value, a Structural Similarity (SSIM) value, and a Peak Signal-To (PSNR-HVS) value for the difference between the original training image 1101 and the third training image 1104 . It may include at least one of a Noise Ratio-Human Vision System) value, a Multiscale SSIM (MS-SSIM) value, a Variance Inflation Factor (VIF) value, and a Video Multimethod Assessment Fusion (VMAF) value. The quality loss information 1130 indicates how similar the third training image 1104 is to the original training image 1101 . As the quality loss information 1130 is smaller, the third training image 1104 is more similar to the original training image 1101 .

11, structural loss information 1110, complexity loss information 1120, and quality loss information 1130 are used for training of the first DNN 800, and the quality loss information 1130 is the second DNN ( 300) is used for training. That is, the quality loss information 1130 is used for both training of the first DNN 800 and the second DNN 300 .

The first DNN 800 may update parameters such that the final loss information determined based on the structural loss information 1110 , the complexity loss information 1120 , and the quality loss information 1130 is reduced or minimized. Also, the second DNN 300 may update the parameter so that the quality loss information 1130 is reduced or minimized.

Final loss information for training of the first DNN 800 and the second DNN 300 may be determined as in Equation 1 below.

[Equation 1]

In Equation 1, LossDS represents final loss information to be reduced or minimized for training of the first DNN 800, and LossUS represents final loss information to be reduced or minimized for training of the second DNN 300. indicates. Also, a, b, c, and d may correspond to predetermined weights.

That is, the first DNN 800 updates parameters in a direction in which LossDS of Equation 1 is decreased, and the second DNN 300 updates parameters in a direction in which LossUS is decreased. When the parameters of the first DNN 800 are updated according to the LossDS derived in the training process, the first training image 1102 obtained based on the updated parameters is different from the first training image 1102 in the previous training process. and, accordingly, the third training image 1104 is also different from the third training image 1104 in the previous training process. When the third training image 1104 is different from the third training image 1104 in the previous training process, the quality loss information 1130 is also newly determined, and accordingly, the second DNN 300 updates the parameters. When the quality loss information 1130 is newly determined, since LossDS is also newly determined, the first DNN 800 updates parameters according to the newly determined LossDS. That is, the parameter update of the first DNN 800 causes the parameter update of the second DNN 300 , and the parameter update of the second DNN 300 causes the parameter update of the first DNN 800 . In other words, since the first DNN 800 and the second DNN 300 are jointly trained through sharing of the quality loss information 1130 , the parameters of the first DNN 800 and the parameters of the second DNN 300 are They can be optimized in relation to each other.

Referring to Equation 1, it can be seen that LossUS is determined according to quality loss information 1130, but this is an example, and LossUS is at least one of structural loss information 1110 and complexity loss information 1120, It may be determined based on the quality loss information 1130 .

Previously, it has been described that the AI setting unit 238 of the AI decoding apparatus 200 and the AI setting unit 718 of the AI encoding apparatus 700 store a plurality of DNN setting information, the AI setting unit 238 and the AI A method of training each of the plurality of DNN configuration information stored in the setting unit 718 will be described.

As described in relation to Equation 1, in the case of the first DNN 800, the degree of similarity between the structural information of the first training image 1102 and the structural information of the original training image 1101 (structural loss information 1110) ), the bit rate (complexity loss information 1120) of the image data obtained as a result of the first encoding of the first training image 1102 and the difference between the third training image 1104 and the original training image 1101 (quality loss) The parameter is updated in consideration of the information 1130).

More specifically, while making it possible to obtain a first training image 1102 similar to the structural information of the original training image 1101 and having a small bit rate of image data obtained when the first encoding is performed, the first training image The parameters of the first DNN 800 may be updated so that the second DNN 300 that AI upscales 1102 may acquire a third training image 1104 similar to the original training image 1101 .

By adjusting the weights of a, b, and c in Equation 1, the directions in which the parameters of the first DNN 800 are optimized are different. For example, when the weight of b is determined to be high, the parameter of the first DNN 800 may be updated with more importance on lowering the bit rate than the quality of the third training image 1104 . In addition, when the weight of c is determined to be high, it is more important to increase the quality of the third training image 1104 than to increase the bit rate or to maintain the structural information of the original training image 1101. 1 A parameter of the DNN 800 may be updated.

Also, the direction in which parameters of the first DNN 800 are optimized may be different according to a type of a codec used to first encode the first training image 1102 . This is because, depending on the type of codec, the second training image to be input to the second DNN 300 may vary.

That is, the parameters of the first DNN 800 and the parameters of the second DNN 300 are linked based on the weight a, the weight b, the weight c, and the type of the codec for the first encoding of the first training image 1102 . so it can be updated. Therefore, when each of the weight a, the weight b, and the weight c is determined as a predetermined value and the type of the codec is determined as a predetermined type, the first DNN 800 and the second DNN 300 are trained, they are linked to each other. Optimized parameters of the first DNN 800 and parameters of the second DNN 300 may be determined.

Then, after changing the weight a, the weight b, the weight c, and the type of codec, if the first DNN 800 and the second DNN 300 are trained, the parameters of the first DNN 800 that are optimized in connection with each other and parameters of the second DNN 300 may be determined. In other words, when the first DNN 800 and the second DNN 300 are trained while changing the respective values of the weight a, the weight b, the weight c, and the codec type, the plurality of DNN setting information that are connected to each other and trained is the first It may be determined from the DNN 800 and the second DNN 300 .

As described above with reference to FIG. 5 , the plurality of DNN configuration information of the first DNN 800 and the second DNN 300 may be mapped to the first image related information. To establish such a mapping relationship, a first training image 1102 output from the first DNN 800 is first encoded with a specific codec according to a specific bit rate, and a bitstream obtained as a result of the first encoding is first decoded. The obtained second training image may be input to the second DNN 300 . That is, after setting the environment so that the first training image 1102 of a specific resolution is first encoded at a specific bit rate by a specific codec, by training the first DNN 800 and the second DNN 300 , the first DNN setting mapped to the resolution of the training image 1102, the type of codec used for the first encoding of the first training image 1102, and the bitrate of the bitstream obtained as a result of the first encoding of the first training image 1102 The information pair can be determined. The resolution of the first training image 1102, the type of codec used for the first encoding of the first training image 1102, and the bitrate of the bitstream obtained according to the first encoding of the first training image 1102 are varied. , a mapping relationship between a plurality of DNN configuration information of the first DNN 800 and the second DNN 300 and the first image related information may be determined.

12 is a diagram for explaining a training process of the first DNN 800 and the second DNN 300 by the training apparatus 1200 .

The training of the first DNN 800 and the second DNN 300 described with reference to FIG. 11 may be performed by the training apparatus 1200 . The training device 1200 includes a first DNN 800 and a second DNN 300 . The training device 1200 may be, for example, the AI encoding device 700 or a separate server. DNN setting information of the second DNN 300 obtained as a result of training is stored in the AI decoding apparatus 200 .

Referring to FIG. 12 , the training apparatus 1200 initially sets DNN configuration information of the first DNN 800 and the second DNN 300 ( S1240 , S1245 ). Accordingly, the first DNN 800 and the second DNN 300 may operate according to predetermined DNN configuration information. DNN setting information includes at least the number of convolution layers included in the first DNN 800 and the second DNN 300, the number of filter kernels for each convolutional layer, the size of filter kernels for each convolutional layer, and parameters of each filter kernel. It can contain information about one.

The training apparatus 1200 inputs the original training image 1101 into the first DNN 800 (S1250). The original training image 1101 may include at least one frame constituting a still image or a moving image.

The first DNN 800 processes the original training image 1101 according to the initially set DNN setting information, and outputs the AI downscaled first training image 1102 from the original training image 1101 ( S1255 ). 12 illustrates that the first training image 1102 output from the first DNN 800 is directly input to the second DNN 300 , but the first training image 1102 output from the first DNN 800 . ) may be input to the second DNN 300 by the training device 1200 . Also, the training apparatus 1200 may first encode and first decode the first training image 1102 using a predetermined codec, and then input the second training image to the second DNN 300 .

The second DNN 300 processes the first training image 1102 or the second training image according to the initially set DNN setting information, and the AI upscaled third from the first training image 1102 or the second training image. A training image 1104 is output (S1260).

The training apparatus 1200 calculates complexity loss information 1120 based on the first training image 1102 (S1265).

The training apparatus 1200 calculates structural loss information 1110 by comparing the reduced training image 1103 with the first training image 1102 ( S1270 ).

The training apparatus 1200 calculates quality loss information 1130 by comparing the original training image 1101 and the third training image 1104 ( S1275 ).

The first DNN 800 updates the initially set DNN configuration information through a back propagation process based on the final loss information (S1280). The training apparatus 1200 may calculate final loss information for training the first DNN 800 based on the complexity loss information 1120 , the structural loss information 1110 , and the quality loss information 1130 .

The second DNN 300 updates the initially set DNN configuration information through a reverse transcription process based on the quality loss information or the final loss information (S1285). The training apparatus 1200 may calculate final loss information for training the second DNN 300 based on the quality loss information 1130 .

Thereafter, the training apparatus 1200, the first DNN 800, and the second DNN 300 update the DNN configuration information while repeating processes S1250 to S1285 until the final loss information is minimized. In this case, during each iteration process, the first DNN 800 and the second DNN operate according to the DNN configuration information updated in the previous process.

Table 1 below shows the effects of AI encoding and AI decoding of the original image 105 and encoding and decoding of the original image 105 using HEVC according to an embodiment of the present disclosure.

[Table 1]

As can be seen from Table 1, the subjective picture quality when AI encoding and AI decoding of contents consisting of 300 frames of 8K resolution according to an embodiment of the present disclosure is higher than the subjective picture quality when encoding and decoding using HEVC. , it can be seen that the bit rate is reduced by more than 50%.

Hereinafter, according to the AI-based upscaling and downscaling described above in FIGS. 1 to 12, the video is streamed in real time using the AI upscale model and the AI downscale model determined according to the resolution and bitrate at the time of video transmission. A method of transmitting an image by adding a deblocking filter in order to solve a blocking phenomenon caused by a decrease in bandwidth due to a change in the environment of a streaming channel will be described later.

13 is a diagram for explaining an AI encoding process and an AI decoding process for adding deblocking filtering according to an embodiment.

As described above in FIG. 1, when the streaming channel environment is normal, the original image 105 is AI downscaled 110 based on the AI downscale model and the AI upscale model suitable according to the streaming channel environment to the first image ( 115), transmits the first image 115 by first encoding 120, receives AI encoded data including AI data and image data obtained as a result of AI encoding, and performs first decoding 130 A second image 135 is acquired through the AI upscaling 140 of the second image 135 and a third image 145 is acquired.

On the other hand, when a change in the environment of the streaming channel occurs, for example, when a large number of users suddenly gather or the communication environment of the terminal becomes bad outdoors, the bandwidth is reduced, so the quantization parameter is increased to lower the bit rate. A blocking phenomenon may occur depending on the resolution and screen characteristics of the image. When a blocking phenomenon, that is, a blocking artifact, occurs, the viewer thinks that the image quality is bad. In order to prevent such blocking, the image quality can be improved by adding a deblocking filter to AI downscaling and AI upscaling.

Referring to FIG. 13 , in the deblocking prediction 1370 , the network environment information of the current streaming channel, the original image 1305, and information on the previously coded frame are acquired to predict whether a blocking phenomenon will occur in the input original image. and transmits filter setting information such as a region map for a region in which a block is generated when AI encoding and AI decoding are performed. In the AI encoding process, a deblocking-filtered original image 1315 is obtained by applying the deblocking filter 1310 to the original image 1305 based on the filter setting information for the deblocking filter, and the deblocking-filtered image 1315 is AI downscaled (1320) to obtain a first image 1325, and the first image 1325 is first encoded (1330) and transmitted. In the AI decoding process, AI data obtained as a result of AI encoding and AI-encoded data including image data is received, a second image 1345 is obtained through a first decoding 1340, and an AI-upscaled image by AI upscaling 1350 of the second image 1345 1355 is obtained, and a reconstructed image 1365 is obtained by applying the deblocking filter 1360 according to the filter setting information transmitted in the deblocking prediction 1370 to the AI upscaled image 1355 .

Network environment information is a real-time protocol (eg, a real-time transport protocol (RTP), a real-time transport Control Protocol, RTCP)) is information indicating changes in the network environment.

If the bandwidth is reduced due to a change in the network environment while streaming is performed at a predetermined resolution and bit rate, a blocking phenomenon may occur. For example, if the resolution is 4K and the bitrate is 20Mbps, if the resolution is 4K and the bitrate is 15Mbps, if the resolution is FHD and the bitrate is 10Mbps, if the resolution is FHD and the bitrate is 7.5Mbps, if the resolution is is HD and the bit rate is 5Mbps, suitable AI downscaling and AI upscaling models are selected for each case, and changes in the network environment (a sharp increase in users, In case the communication environment deteriorates), deblocking filtering is performed according to the change in the network environment, the characteristics of the original video, and whether blocking artifacts of the previous frame are detected in AI encoding and AI decoding through the deblocking prediction process. The image quality can be improved by predicting and recognizing the occurrence of blocking artifacts in a limited network bandwidth and removing the blocking artifacts. In general, in video encoding and decoding, there is a deblocking filter that removes blocking artifacts on a block-by-block basis as an in-loop filter. Deblocking filtering may be performed through the filter to improve image quality.

The deblocking prediction 1370 obtains whether there is a region where blocking is likely to occur from the original image 1305, obtains whether a blocking phenomenon of the previous frame of the current frame is detected from the previous frame information, and from the network environment information Information on changes in bandwidth, bitrate, etc. is acquired, and filter setting information including a blocking region map for a region in which blocks may be generated is transmitted when AI decoding and AI encoding are performed. Deblocking filtering is additionally performed when each of the AI decoding and AI encoding processes are performed based on the transmitted filter setting information.

The deblocking filter 1310 in the AI encoding process is a pre-filter, and when the total complexity is more complex than the allocated bits, the image quality of the complex region and the non-complex region is redistributed. For example, if you are in a simple area such as a tree standing still in the image, or in a complex and fast-moving area such as a water stream coming out of a fountain, simply blur the complex area to reduce the complexity to reduce the bitrate and reduce the complexity. The bit rate according to is distributed to a simple area during encoding. Since the blocking phenomenon occurs when the image quality is not properly distributed, the probability of occurrence of the blocking phenomenon can be reduced by redistributing the image quality. Complex and fast-moving areas are blurred to make blocking artifacts less noticeable, and the bitrate reduced through blurring of complex and fast-moving areas is allocated to simple areas, resulting in a higher bitrate than before. Prevents the occurrence of a blocking phenomenon in the area.

The deblocking filter 1360 in the AI decoding process is a post-filter, and applies deblocking filtering according to the blocking area map included in the filter setting information, so that the boundary is clear for the blurred part. It creates a sharpen effect that makes it look sharp, and restores it by removing blocking artifacts.

In the deblocking prediction 1370, filter setting information including a blocking area map for the blocking artifact occurrence probability of the current frame and the occurrence probability for each region may be generated. In the original image, blocks with fast movement and complex and flat areas are relatively more visible. Based on this point, a blocking area map indicating the occurrence probability of the blocking artifact may be generated.

In addition, by generating a score map according to various criteria, a blocking area map may be generated according to a weighted average.

For example, the A score map can be generated by calculating the complexity for each block or pixel from the original image, finding a high-complexity area, and giving a high-score to the high-complexity area, and the complexity is calculated based on the total variance. can The B-score map may be generated by checking the motion from the N-1 th frame and an optical flow, and giving a high score to an area where there is a lot of motion change.

In addition, if there are many blocking artifacts in the previous image, the current image may also have many blocking artifacts. Accordingly, the C score map may be generated by giving a high score to a region having many blocking artifacts in the previous region, and the D score map may be generated by giving a high score to a region having a large difference in pixel boundary values at the block boundary.

Finally, during encoding, blocking artifacts may be generated according to the complexity of the current frame, the target bit rate, and the quantization parameter (QP). For example, the E score map may be generated based on a delta quantization parameter (DQP) map indicating a difference between an average QP of pixels of a block of a current frame and a QP of each pixel. The F score map is generated based on the coding complexity, that is, the coding score, the coding score is determined as the sum of the intra coding score and the inter coding score, and the intra coding score is the original image and the intra coded score with the current image as input. is determined based on the difference between the image and the prediction bits used to code the image, and the inter-coding score is used to code the difference between the original image and the inter-coded image and the image with the current image and at least one previous image as inputs. It may be determined based on the prediction bit. The G-score map may be generated based on a value obtained by dividing the coding score of the F-score map by the target bitrate (channel information, bandwidth). Accordingly, the G score map may be generated based on a rate control in which the encoder estimates the bit rate of the encoded bitstream based on available bandwidth on the network. Even if the coding score is large, if the target bitrate is relatively large, the score may not be large in the G-score map.

As described above, the weighted average of the A to G score maps may be the blocking area map. However, the A to G score map is an example and is not limited thereto.

In the deblocking prediction process, deblocking filtering may be performed according to filter setting information including such a blocking region map.

According to an embodiment, the deblocking filter may be a legacy deblocking filter.

According to another embodiment, the deblocking filter may be a trained DNN. Accordingly, in the deblocking prediction, filter setting information transmitted for the deblocking filter may include a blocking region map and DNN setting information. In FIGS. 20 and 21, when the deblocking filter is a DNN, the third DNN for deblocking filtering for the pre-processing filter performed before AI downscaling and the fourth DNN for deblocking filtering for the post-processing filter performed after the AI upscaling A method and training process for training will be described later.

Accordingly, even if a change in the communication environment occurs, the AI downscale model does not change the initially selected AI downscale model and AI scale model, but performs deblocking filtering according to the network environment, original image, and previous frame information. And the quality of the image can be adjusted while maintaining the AI scale model.

According to an embodiment, information on a DQP map or rate control is additionally first encoded ( 1330) and may be used to perform the first encoding.

14 is a block diagram illustrating a configuration of an AI decoding apparatus 1400 according to an embodiment.

Referring to FIG. 14 , the AI decoding apparatus 1400 according to an embodiment includes a receiving unit 1410 , an AI decoding unit 1420 , and a deblocking prediction unit 1430 . The AI decoding unit 230 may include a parsing unit 1421 , a first decoding unit 1422 , an AI upscaling unit 1423 , a deblocking filter unit 1424 , and an AI setting unit 1425 .

Although the receiver 1410 , the AI decoder 1420 , and the deblocking predictor 1430 are shown as separate devices in FIG. 14 , the receiver 1410 , the AI decoder 1420 , and the deblocking predictor 1430 . ) can be implemented through one processor. In this case, the receiving unit 1410 , the AI decoding unit 1420 , and the deblocking prediction unit 1430 may be implemented as a dedicated processor, and may include an application processor (AP), a central processing unit (CPU), or a graphic processing unit (GPU). ) can also be implemented through a combination of a general-purpose processor and S/W. Also, in the case of a dedicated processor, a memory for implementing an embodiment of the present disclosure or a memory processing unit for using an external memory may be included.

The receiver 1410 , the AI decoder 1420 , and the deblocking predictor 1430 may include a plurality of processors. In this case, it may be implemented as a combination of dedicated processors, or may be implemented through a combination of S/W with a plurality of general-purpose processors such as an AP, CPU, or GPU. In an embodiment, the receiving unit 1410 is implemented as a first processor, the first decoding unit 1422 is implemented as a second processor different from the first processor, and the parsing unit 1421 and the AI upscaling unit 1423 are implemented. , the deblocking filter unit 1424 and the AI setting unit 1425 are implemented as a first processor and a third processor different from the second processor, and the deblocking prediction unit 1430 includes the first processor, the second processor, and the second processor. It may be implemented with a third processor different from the three processors.

The receiving unit 1410 receives the AI encoded data obtained as a result of the AI encoding and filter setting information transmitted from the deblocking prediction unit of the AI encoding apparatus. As an example, the AI-encoded data may be a video file having a file format such as mp4 or mov.

The receiver 1410 may receive AI-encoded data and filter setting information transmitted through a network. The receiving unit 1410 outputs the AI encoded data and filter setting information to the AI decoding unit 1420 .

In another embodiment, the receiver 1410 may receive AI-encoded data including filter setting information.

In one embodiment, AI-encoded data is stored in a hard disk, magnetic media such as floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magneto-optical media such as floppy disks. It may be obtained from a data storage medium including an optical medium).

The parsing unit 1421 parses the AI encoded data, and transmits image data generated as a result of the first encoding of the first image 1325 to the first decoder 1422 , and transmits the AI data to the AI setting unit 1425 . and transfers the filter setting information to the deblocking prediction unit 1430 .

In an embodiment, the deblocking prediction unit 1430 transmits filter setting information to the deblocking filter unit 1424 . The filter setting information may include a blocking area map indicating an area in which block artifacts are generated.

In another embodiment, the filter setting information may include a blocking region map indicating a region in which block artifacts are generated and DNN setting information of a DNN for deblocking filtering.

In an embodiment, the parsing unit 1421 may parse the image data and the AI data separately included in the AI encoded data. The parsing unit 1421 may read the header in the AI-encoded data to distinguish between AI data and image data included in the AI-encoded data. In one example, AI data may be included in a Vendor Specific InfoFrame (VSIF) in the HDMI stream. The structure of AI-encoded data including separated AI data and image data will be omitted as described above with reference to FIG. 9 .

In another embodiment, the parsing unit 1421 parses the image data from the AI encoded data, extracts the AI data from the image data, transmits the AI data to the AI setting unit 1425, and first decodes the remaining image data. may be passed to unit 1422 . That is, AI data may be included in image data. For example, AI data may be included in supplemental enhancement information (SEI), which is an additional information area of a bitstream corresponding to image data. The structure of AI-encoded data including image data including AI data is omitted as described above with reference to FIG. 10 .

In another embodiment, the parsing unit 1421 parses image data, AI data, and filter setting information separately included in the AI encoded data, and image data generated as a result of the first encoding of the first image 1325 . may be output to the first decoder 1422 , AI data may be output to the AI setting unit 1425 , and filter setting information may be outputted to the deblocking prediction unit 1430 .

In another embodiment, the parsing unit 1421 divides a bitstream corresponding to image data into a bitstream to be processed by the first decoding unit 1422 and a bitstream corresponding to AI data, and divides each of the divided bitstreams into a bitstream corresponding to AI data. It may output to the first decoding unit 1422 and the AI setting unit 1425 .

The parsing unit 1421 obtains image data included in the AI encoded data through a predetermined codec (eg, MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9, or AV1). It can also be confirmed that it is the image data that has been processed. In this case, the corresponding information may be transmitted to the first decoder 1422 so that the image data can be processed by the identified codec.

The first decoder 1422 reconstructs the second image 1345 corresponding to the first image 1325 based on the image data received from the parser 1421 . The second image 1345 obtained by the first decoder 1422 is provided to the AI upscaler 1423 .

According to an embodiment, first decoding-related information such as prediction mode information, motion information, and quantization parameter information may be provided from the first decoding unit 1422 to the AI setting unit 1425 . The first decoding-related information may be used to obtain DNN configuration information.

The AI data provided to the AI setting unit 1425 includes information enabling AI upscaling of the second image 1345 . In this case, the upscale target of the second image 1345 must correspond to the downscale target of the first DNN. Therefore, the AI data should include information that can confirm the downscale target of the first DNN.

Specifically, the information included in the AI data includes information on a difference between the resolution of the original image 1305 and the resolution of the first image 1325 and information related to the first image 132 .

The difference information may be expressed as information about a resolution conversion degree of the first image 1325 compared to the original image 1305 (eg, resolution conversion rate information). In addition, since the resolution of the first image 1325 is known through the resolution of the reconstructed second image 1345 and the degree of resolution conversion can be checked through this, the difference information can be expressed only with the resolution information of the original image 1305 . may be Here, the resolution information may be expressed as a horizontal/vertical screen size, or a ratio (16:9, 4:3, etc.) and a size of one axis. In addition, when there is preset resolution information, it may be expressed in the form of an index or a flag.

The information related to the first image 1325 is used for the resolution of the first image 1325 , the bit rate of image data obtained as a result of the first encoding of the first image 1325 , and the first encoding of the first image 1325 . information on at least one of the specified codec types may be included.

The AI setting unit 1425 may determine an upscaling target of the second image 1345 based on at least one of difference information included in the AI data and information related to the first image 1325 . The upscaling target may indicate, for example, to what resolution the second image 1345 should be upscaled.

When the upscale target is determined, the AI upscale unit 1423 AI upscales the second image 1345 through the second DNN to obtain an upscaled image 1355 corresponding to the upscale target.

The deblocking filter unit 1424 obtains the third image 1465 by applying deblocking filtering to the upscaled second image 1355 based on the filter setting information obtained from the deblocking prediction unit 1430 .

The AI upscaling process through the second DNN is omitted as described above in FIGS. 3 and 4 .

Hereinafter, a method in which the AI setting unit 1425 determines an upscaling target based on AI data and the AI upscaling unit 1423 AI upscales the second image 1345 according to the upscaling target will be described. .

In an embodiment, the AI setting unit 1425 may store a plurality of DNN setting information settable in the second DNN.

Here, the DNN configuration information may include information on at least one of the number of convolutional layers included in the second DNN, the number of filter kernels for each convolutional layer, and parameters of each filter kernel.

The plurality of DNN configuration information may respectively correspond to various upscale targets, and the second DNN may operate based on the DNN configuration information corresponding to a specific upscale target. The second DNN may have a different structure according to the DNN configuration information. For example, the second DNN may include three convolutional layers according to certain DNN configuration information, and the second DNN may include four convolutional layers according to other DNN configuration information.

In an embodiment, the DNN configuration information may include only parameters of a filter kernel used in the second DNN. In this case, the structure of the second DNN is not changed, but only the parameters of the internal filter kernel may be changed according to the DNN configuration information.

The AI setting unit 1425 may acquire DNN setting information for AI upscaling of the second image 1345 among a plurality of DNN setting information. Each of the plurality of DNN configuration information used here is information for acquiring the AI upscaled second image 1355 of a predetermined resolution and/or predetermined image quality, and is trained in association with the first DNN.

For example, the DNN setting information of any one of the plurality of DNN setting information is the third image 1365 having a resolution twice that of the second image 1345, for example, the second image of 2K (2048 * 1080). It may include information for acquiring the third image 1365 of 4K (4096 * 2160) that is twice as large as the 2 images 1345 , and the other DNN setting information is 4 more than the resolution of the second image 1345 . Information for obtaining a third image 1365 with a resolution that is twice as large, for example, a third image 1365 of 8K (8192 * 4320) that is four times larger than the second image 1345 of 2K (2048 * 1080) may include

Each of the plurality of DNN setting information is created in association with the DNN setting information of the first DNN of the AI encoding device 1500, and the AI setting unit 1425 is set at an enlargement ratio corresponding to the reduction rate of the DNN setting information of the first DNN. Accordingly, one piece of DNN configuration information among a plurality of DNN configuration information is acquired. To this end, the AI setting unit 1425 must check the information of the first DNN. In order for the AI setting unit 1425 to check the information of the first DNN, the AI decoding apparatus 1400 according to an embodiment receives AI data including the information of the first DNN from the AI encoding apparatus 1500 .

In other words, the AI setting unit 1425 uses the information received from the AI encoding apparatus 1500 to identify information targeted by the DNN setting information of the first DNN used to obtain the first image 1325 and , it is possible to obtain the DNN setting information of the second DNN trained in association with it.

When DNN setting information for AI upscaling of the second image 1345 is obtained among a plurality of DNN setting information, the DNN setting information is transmitted to the AI upscaling unit 1423, and a second DNN operating according to the DNN setting information. Based on the input data may be processed.

For example, when any one DNN configuration information is obtained, the AI upscaling unit 1423 includes the first convolutional layer 310 and the second convolutional layer 330 of the second DNN 300 shown in FIG. 3 . ) and the third convolutional layer 350, the number of filter kernels included in each layer and parameters of the filter kernels are set to values included in the obtained DNN configuration information.

Specifically, the AI upscaling unit 1423 determines that the parameters of the 3 X 3 filter kernel used in any one convolution layer of the second DNN shown in FIG. 4 are {1, 1, 1, 1, 1, 1 , 1, 1, 1}, and when there is a change in DNN configuration information, the parameters of the filter kernel are changed to {2, 2, 2, 2, 2, 2, 2, 2, 2, which are parameters included in the changed DNN configuration information. } can be replaced.

The AI setting unit 1425 may acquire DNN setting information for upscaling the second image 1345 among a plurality of DNN setting information based on information included in the AI data, which is used to obtain the DNN setting information. AI data will be explained in detail.

In an embodiment, the AI setting unit 1425 may obtain DNN setting information for upscaling the second image 1345 from among a plurality of DNN setting information, based on the difference information included in the AI data. For example, based on the difference information, the resolution of the original image 1305 (eg, 4K (4096 * 2160)) is higher than the resolution of the first image 1325 (eg, 2K (2048 * 1080)). When it is confirmed that the size is 2 times larger, the AI setting unit 238 may obtain DNN setting information capable of increasing the resolution of the second image 1345 by 2 times.

In another embodiment, the AI setting unit 1425 is DNN setting information for AI upscaling the second image 1345 among a plurality of DNN setting information based on the information related to the first image 1325 included in the AI data. can be obtained. The AI setting unit 1425 may determine a mapping relationship between the image related information and the DNN setting information in advance, and obtain DNN setting information mapped to the first image 1325 related information.

15 is a block diagram illustrating a configuration of an AI encoding apparatus 1500 according to an embodiment.

Referring to FIG. 15 , the AI encoding apparatus 1500 may include an AI encoding unit 1510 , a transmission unit 1520 , and a deblocking prediction unit 1530 . The AI encoder 1510 may include a deblocking filter unit 1511 , an AI downscale unit 1512 , a first encoder 1513 , a data processing unit 1514 , and an AI setting unit 1515 .

15 shows the AI encoder 1510, the transmitter 1520, and the deblocking predictor 1530 as separate devices, the AI encoder 1510, the transmitter 1520, and the deblocking predictor 1530 may be implemented through one processor. In this case, it may be implemented as a dedicated processor, or may be implemented through a combination of an AP or a general-purpose processor such as a CPU or GPU and S/W. Also, in the case of a dedicated processor, a memory for implementing an embodiment of the present disclosure or a memory processing unit for using an external memory may be included.

Also, the AI encoder 1510 , the transmitter 1520 , and the deblocking predictor 1530 may include a plurality of processors. In this case, it may be implemented as a combination of dedicated processors, or may be implemented through a combination of S/W with a plurality of general-purpose processors such as an AP, CPU, or GPU. In an embodiment, the first encoding unit 1513 is configured as a first processor, and the deblocking filter unit 1511 , the AI downscale unit 1512 , the data processing unit 1514 , and the AI setting unit 1515 are the first It is implemented as a second processor different from the first processor, the transmission unit 1520 is implemented as a first processor and a third processor different from the second processor, and the deblocking prediction unit 1530 includes the first processor, the second processor, and It may be implemented with a fourth processor different from the third processor.

The deblocking prediction unit 1530 obtains the original image 1305, previous frame information, and network environment information, and converts the filter setting information to the deblocking filter based on the original image 1305, the previous frame information, and the network environment information. sent to the unit 1511.

The deblocking prediction unit 1530 obtains previous frame information from the first encoder 1513 , and obtains network environment information from the AI decoding apparatus 1400 according to a real-time protocol.

In an embodiment, the deblocking prediction unit 1530 transmits filter setting information to the deblocking filter unit 1511 . The filter setting information may include a blocking area map indicating an area in which block artifacts are generated.

In another embodiment, the filter setting information may include a blocking region map indicating a region in which block artifacts are generated and DNN setting information of a DNN for deblocking filtering.

The deblocking filter unit 1511 obtains the deblocking-filtered original image 1315 by applying deblocking filtering to the original image 1305 based on the filter setting information obtained from the deblocking prediction unit 1530 .

The AI encoder 1513 performs AI downscaling of the deblocking-filtered original image 1315 and the first encoding of the first image 1325 , and transmits AI-encoded data to the transmitter 1520 . The transmitter 1520 transmits the AI encoding data and filter setting information to the AI decoding apparatus 1400 .

In another embodiment, the transmitter 1520 transmits AI-encoded data including filter setting information, AI data, and image data to the AI decoding apparatus 1400 .

The image data includes data obtained as a result of the first encoding of the first image 1325 . The image data may include data obtained based on pixel values in the first image 1325 , for example, residual data that is a difference between the first image 1325 and prediction data of the first image 1325 . . Also, the image data includes information used in the first encoding process of the first image 1325 . For example, the image data may include prediction mode information used for first encoding the first image 1325 , motion information, and quantization parameter related information used for first encoding the first image 1325 , etc. .

The AI data includes information enabling the AI upscaling unit 1513 to AI upscale the second image 1345 to an upscale target corresponding to the downscale target of the first DNN.

In an example, the AI data may include difference information between the original image 1305 and the first image 1325 .

In one example, the AI data may include information related to the first image 1325 . The information related to the first image 1325 is used for the resolution of the first image 1325 , the bit rate of image data obtained as a result of the first encoding of the first image 1325 , and the first encoding of the first image 1325 . information on at least one of the specified codec types may be included.

In an embodiment, the AI data may include an identifier of mutually agreed DNN configuration information so that the second image 1345 can be AI upscaled to an upscale target corresponding to the downscale target of the first DNN. .

Also, in an embodiment, the AI data may include DNN configuration information settable in the second DNN.

The AI downscaler 1512 may acquire the AI downscaled image 1325 through the first DNN from the deblocking-filtered original image 1315 . The AI downscaling unit 1512 may AI downscale the filtered original image 1315 using the DNN setting information provided from the AI setting unit 1515 .

The AI setting unit 1515 may determine a downscale target of the original image 105 based on a predetermined criterion.

In order to acquire the first image 1325 matching the downscale target, the AI setting unit 1515 may store a plurality of settable DNN setting information in the first DNN. The AI setting unit 1515 obtains DNN setting information corresponding to the downscale target among a plurality of DNN setting information, and provides the obtained DNN setting information to the AI downscale unit 1512 .

Each of the plurality of pieces of DNN configuration information may be trained to acquire the first image 1325 having a predetermined resolution and/or a predetermined image quality. For example, the DNN setting information of any one of the plurality of DNN setting information is the first image 1325 having a resolution that is 1/2 times smaller than the resolution of the original image 1305, for example, 4K (4096 * 2160) may include information for acquiring the first image 1325 of 2K (2048*1080) 1/2 times smaller than the original image 1305 of A first image 1325 with a resolution of 1/4 times smaller than that of the first image 1325, for example, a first image 1325 of 2K (2048*1080) which is 1/4 times smaller than the original image 1305 of 8K (8192*4320). ) may include information for obtaining

According to the embodiment, when information constituting the DNN setting information (eg, the number of convolution layers, the number of filter kernels for each convolutional layer, parameters of each filter kernel, etc.) is stored in the form of a lookup table, The AI setting unit 1515 may obtain DNN setting information by combining some selected among the lookup table values according to the downscale target, and may provide the obtained DNN setting information to the AI downscaling unit 1512 .

According to an embodiment, the AI setting unit 1515 may determine the structure of the DNN corresponding to the downscale target, and obtain DNN setting information corresponding to the determined structure of the DNN, for example, parameters of the filter kernel.

The plurality of DNN configuration information for AI downscaling of the filtered original image 1315 may have an optimized value by jointly training the first DNN and the second DNN. Here, each DNN configuration information includes at least one of the number of convolutional layers included in the first DNN, the number of filter kernels for each convolutional layer, and parameters of each filter kernel.

The AI downscaling unit 1512 sets the first DNN as the DNN setting information determined for AI downscaling of the filtered original image 1315 to convert the first image 1325 of a predetermined resolution and/or a predetermined quality to the first It can be obtained through DNN. When DNN setting information for AI downscaling of the original image 1305 is obtained among a plurality of DNN setting information, each layer in the first DNN may process input data based on information included in the DNN setting information. .

A method for the AI setting unit 1515 to determine the downscale target will be described. The downscale target may indicate, for example, how much the resolution reduced first image 1325 should be obtained from the deblocking-filtered original image 1315 .

The AI setting unit 1515 acquires one or more pieces of input information. In an embodiment, the input information includes a target resolution of the first image 1325, a target bitrate of the image data, a bitrate type of the image data (eg, a variable bitrate type, a constant bitrate type, or an average bitrate type, etc.); Color format to which AI downscale is applied (luminance component, chrominance component, red component, green component, or blue component, etc.), codec type for the first encoding of the first image 1325, compression history information, and the original image 1305 may include at least one of a resolution of , and a type of the original image 1305 .

One or more pieces of input information may include information previously stored in the AI encoding apparatus 1500 or received from a user.

The AI setting unit 1515 controls the operation of the AI downscale unit 1512 based on the input information. In an embodiment, the AI setting unit 1515 may determine a downscale target according to input information, and provide DNN setting information corresponding to the determined downscale target to the AI downscale unit 1512 .

In an embodiment, the AI setting unit 1515 transmits at least a portion of the input information to the first encoder 1513 so that the first encoder 1513 sets the bitrate of a specific value, the bitrate of a specific type, and the specific codec. may cause the first image 1325 to be first encoded.

In another embodiment, the deblocking prediction unit 1530 additionally acquires information on a DQP map or rate control based on the original image 1305, previous frame information, and network environment information, and the DQP map Alternatively, information on rate control may be transmitted to the first encoder 1513 . Also, the first encoder 1513 may use information on a DQP map or rate control to first encode the first image 1325 .

In an embodiment, the AI setting unit 1515 sets the compression rate (eg, the difference in resolution between the original image 1305 and the first image 1325 , the target bitrate), and the compression quality (eg, the bitrate type). ), compression history information, and a downscale target may be determined based on at least one of the type of the original image 1305 .

In one example, the AI setting unit 1515 may determine a downscale target based on a preset or a compression rate or compression quality input from a user.

As another example, the AI setting unit 1515 may determine the downscale target by using the compression history information stored in the AI encoding apparatus 1500 . For example, according to the compression history information available to the AI encoding apparatus 1500, the encoding quality or compression rate preferred by the user may be determined, and the downscale target may be determined according to the encoding quality determined based on the compression history information. can For example, the resolution and quality of the first image 115 may be determined according to the encoding quality that has been most frequently used according to the compression history information.

As another example, the AI setting unit 1515 sets the encoding quality that has been used more than a predetermined threshold value (eg, average quality of encoding qualities that have been used more than a predetermined threshold value) according to the compression history information. Based on the downscale target may be determined.

As another example, the AI setting unit 1515 may determine the downscale target based on the resolution and type (eg, file format) of the original image 1305 .

In one embodiment, when the original image 1305 is composed of a plurality of frames, the AI setting unit 1515 independently acquires DNN setting information for each predetermined number of frames, and uses the independently acquired DNN setting information to AI downscaling. A portion 1512 may be provided.

In one example, the AI setting unit 1515 may classify frames constituting the original image 1305 into a predetermined number of groups, and independently obtain DNN setting information for each group. The same or different DNN configuration information may be obtained for each group. The number of frames included in the groups may be the same or different for each group.

In another example, the AI setting unit 1515 may independently determine DNN setting information for each frame constituting the original image 1305 . The same or different DNN configuration information may be obtained for each frame.

Hereinafter, embodiments in which the position of the deblocking filter is different from that of FIG. 13 will be described in FIGS. 16 to 18 .

16 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.

16 shows an embodiment in which the position of the deblocking filter in the AI encoding process is changed compared with the AI encoding process and the AI decoding process in which the deblocking filtering of FIG. 13 is adaptively added.

Referring to FIG. 16 , in the deblocking prediction process 1670 , the network environment information of the current streaming channel, the original image 1605, and information on the previously encoded frame are acquired to determine whether a blocking phenomenon will occur in the input original image. It predicts and transmits filter setting information such as a blocking region map for a region where blocks are generated when AI encoding and AI decoding are performed. In the AI encoding process, the original image 1605 is AI downscaled 1610 to obtain a first image 1615, and based on the filter setting information obtained in the deblocking prediction process 1670, the first image 1615 The deblocking filter 1620 is applied to obtain a deblocking-filtered first image 1625, and the deblocking-filtered first image 1625 is first encoded 1630 and transmitted. In the AI decoding process, Receives AI encoded data including AI data and image data obtained as a result of AI encoding, obtains a second image 1645 through first decoding 1640, and performs AI upscale (1650) of the second image 1645 ) to acquire an AI upscaled image 1355 , and apply a deblocking filter 1660 according to the filter setting information transmitted in the deblocking prediction 1370 to obtain a third image 1665 .

17 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.

17 is an embodiment in which the positions of the deblocking filter in the AI encoding process and the deblocking filter in the AI decoding process are all changed compared to the AI encoding process and the AI decoding process in which the deblocking filtering of FIG. 13 is adaptively added. indicates.

Referring to FIG. 17, in the deblocking prediction process 1770, the network environment information of the current streaming channel, the original image 1705, and information on the previously encoded frame are acquired to determine whether a blocking phenomenon will occur in the input original image. It predicts and transmits filter setting information such as a blocking region map for a region where blocks are generated when AI encoding and AI decoding are performed. In the AI encoding process, the original image 1705 is AI downscaled (1710) to obtain a first image 1715, and based on the filter setting information obtained in the deblocking prediction process 1770, the first image 1715 The deblocking filter 1720 is applied to obtain a deblocking-filtered first image 1725, and the deblocking-filtered first image 1725 is first encoded 1730 and transmitted. In the AI decoding process, Receives AI encoded data including AI data and image data obtained as a result of AI encoding, obtains a second image 1745 through a first decoding 1740, and performs a deblocking prediction 1370 on the second image 1745 ) by applying the deblocking filter 1750 according to the transmitted filter setting information to obtain a deblocking-filtered second image 1755, and AI upscaling the deblocking-filtered second image 1755 (1760) A third image 1765 is acquired.

18 is a diagram for explaining an AI encoding process and an AI decoding process for adaptively adding deblocking filtering according to another embodiment.

18 shows an embodiment in which the position of the deblocking filter in the AI decoding process is changed compared to the AI encoding process and the AI decoding process in which the deblocking filtering of FIG. 13 is adaptively added.

Referring to FIG. 18 , in the deblocking prediction process 1870 , the network environment information of the current streaming channel, the original image 1805, and information on the previously encoded frame are acquired to determine whether a blocking phenomenon will occur in the input original image. It predicts and transmits filter setting information such as a blocking region map for a region where blocks are generated when AI encoding and AI decoding are performed. In the AI encoding process, the deblocking-filtered original image 1815 by applying the deblocking filter 1810 to the original image 1805 based on the filter setting information obtained in the deblocking prediction process 1870 is AI downscaled ( 1820) to obtain a first image 1825, and transmit the first image 1825 by first encoding 1830. In the AI decoding process, AI encoding including AI data and image data obtained as a result of AI encoding The data is received, the second image 1845 is obtained through the first decoding 1840 , and the second image 1845 is deblocking filter 1850 according to the filter setting information transmitted in the deblocking prediction process 1870 . ) to obtain a deblocking-filtered second image 1855 , and AI upscaling 1860 on the deblocking-filtered second image 1855 to obtain a third image 1865 .

Hereinafter, when the deblocking filtering of FIG. 13 is an AI deblocking filter trained with a DNN rather than a legacy deblocking filter, a DNN for deblocking filtering will be described later.

19 is an exemplary diagram illustrating a third DNN and a fourth DNN for a deblocking filter based on AI.

19 , the original image 1305 or the upscaled second image 1355 is input to the first convolutional layer 1910 . The first convolutional layer 1910 performs convolution processing on the original image 1305 or the upscaled second image 1355 using four filter kernels having a size of 3×3. As a result of the convolution process, four feature maps are generated. Each feature map represents unique characteristics of the original image 1305 or the upscaled second image 1355 . For example, each feature map may indicate a vertical direction characteristic, a horizontal direction characteristic, an edge characteristic, or whether there is a possibility of occurrence of a blocking artifact of the original image 1305 or the upscaled second image 1355 . The generated four feature maps are input to the first activation layer 1920 .

The first activation layer 1920 may provide non-linear characteristics to the four feature maps.

The first activation layer 1920 determines whether to transfer sample values of the feature maps output from the first convolution layer 1910 to the second convolution layer 1930 . For example, some sample values among the sample values of the feature maps are activated by the first activation layer 1920 and transmitted to the second convolution layer 1930 , and some sample values are activated by the first activation layer 1920 . It is deactivated and is not transmitted to the second convolutional layer 1930 . Information indicated by the feature maps output from the first convolutional layer 1910 is emphasized by the first activation layer 1920 .

An output 1925 of the first activation layer 1920 is input to the second convolution layer 1930 . The second convolution layer 1930 performs convolution processing on input data using four filter kernels having a size of 3×3. The four feature maps output as a result of the convolution process are input to the second activation layer 1940 , and the second activation layer 1940 may provide nonlinear characteristics to the four feature maps.

The output 1945 of the second activation layer 1940 is input to the third convolution layer 1950 . The third convolution layer 1950 performs convolution processing on input data using one filter kernel having a size of 3×3. As a result of the convolution process, one image may be output from the third convolutional layer 1950 . The third convolution layer 1950 is a layer for outputting a final image, and uses one filter kernel to obtain one output. According to the example of the present disclosure, the third convolution layer 1950 may output the deblocking-filtered original image 1315 or the third image 1365 through the convolution operation result.

DNN setting information indicating the number of filter kernels and parameters of the filter kernels of the first convolutional layer 1910, the second convolution layer 1930, and the third convolution layer 1950 of the third DNN or the fourth DNN may be plural.

A plurality of DNN configuration information of the 3rd DNN and the 4th DNN should be linked to each other. The association between the plurality of DNN configuration information of the third DNN and the plurality of DNN configuration information of the fourth DNN may be implemented through association learning of the third DNN and the fourth DNN.

19 shows that the third DNN or the fourth DNN includes three convolutional layers 1910, 1930, 1950 and two activation layers 1920 and 1940, but this is only an example, and the implementation According to an example, the number of convolutional layers and activation layers may be variously changed. Also, according to an embodiment, the third DNN or the fourth DNN may be implemented through a recurrent neural network (RNN). In this case, it means changing the CNN structure of the third DNN or the fourth DNN according to the example of the present disclosure to the RNN structure.

In an embodiment, the deblocking filter unit 1424 or the deblocking filter unit 1511 may include at least one ALU for a convolution operation and an activation layer operation. The ALU may be implemented as a processor. For the convolution operation, the ALU performs a multiplication operation between the sample values of the original image 1305, the upscaled second image 1355, or the feature map output from the previous layer and the sample values of the filter kernel. It may include an adder that adds the result values of . In addition, for the operation of the activation layer, the ALU is a multiplier that multiplies an input sample value by a weight used in a predetermined sigmoid function, a Tanh function, or a ReLU function, and compares the multiplication result with a predetermined value to calculate the input sample value It may include a comparator that determines whether to transfer to the next layer.

Hereinafter, a method for jointly training the third DNN 2010 and the fourth DNN 2020 will be described with reference to FIG. 20 . The first DNN 800 and the second DNN 300 used for joint training of the third DNN 2010 and the fourth DNN 2020 are DNNs trained in the manner described above in FIGS. 11 and 12 .

20 is a diagram for explaining a method of training the third DNN 2010 and the fourth DNN 2020. Referring to FIG.

In one embodiment, when there is a change in the streaming environment (when there is a large number of users or when there is a change in the communication environment outdoors), the original image 1305, previous frame information, and network environment information are acquired through the deblocking prediction process and transmits the deblocking filter setting information for AI encoding and AI decoding based on the original image 1305, previous frame information, and network environment information, and through the AI encoding process, the deblocking prediction process transmitted through the deblocking prediction process. Based on the blocking vector setting information, the original image is deblocking filtered, the AI downscaled first image 1325 is first encoded and transmitted, the first decoded second image 1345 is AI upscaled, Based on the vector setting information transmitted through the deblocking prediction process, the deblocking vector is restored to the third image 1365 .

In the AI encoding and AI decoding process described above in FIGS. 1 to 12, when the streaming environment is maintained, the DNN setting information of the first DNN and the second DNN is appropriately selected for the resolution and bit rate according to the initial network environment, so that additional data Although an image is transmitted without a blocking filter, if a problem occurs in the streaming environment, the image may be transmitted while maintaining the image quality through additional deblocking filters without changing the DNN setting information of the first DNN and the second DNN.

Therefore, the third DNN and the fourth DNN for deblocking filtering are AI downscaled and AI upscaled with DNN setting information of the trained first DNN and trained second DNN corresponding to a specific resolution and a specific bitrate, respectively, Assuming a change in the network environment, the bandwidth or bitrate of the network channel is trained under some changed conditions.

Accordingly, in order to maintain the similarity between the original image 1305 and the third image 1365 obtained as a result of AI encoding and AI decoding to which deblocking filtering is added, correlation is required between the AI encoding process and the AI decoding process. That is, information lost in the AI encoding process should be able to be restored in the AI decoding process, and for this purpose, joint training of the third DNN 2010 and the fourth DNN 2020 is required.

For accurate deblocking filtering and AI decoding, it is ultimately necessary to reduce the quality loss information 2050 corresponding to the comparison result between the third training image 2004 and the original training image 2001 shown in FIG. 20 . Accordingly, the quality loss information 2050 is used for both training of the third DNN 2010 and the fourth DNN 2020.

First, the training process shown in FIG. 20 will be described.

In FIG. 20 , an original training image 2001 is an image subjected to deblocking filtering and AI downscaling, and a first training image 2002 is obtained from the original training image 1101 . Deblocking filtering and AI downscaled video. Also, a third training image 2004 is an AI upscaled and deblocking filtered image from the first training image 2002 .

The original training image 2001 includes a still image or a moving image composed of a plurality of frames. In an embodiment, the original training image 2001 may include a still image or a luminance image extracted from a moving picture including a plurality of frames. Also, according to an embodiment, the original training image 2001 may include a still image or a patch image extracted from a moving picture including a plurality of frames. When the original training image 2001 consists of a plurality of frames, the first training image 2002, the second training image, and the third training image 2004 also include a plurality of frames. When a plurality of frames of the original training image 2001 are sequentially input to the third DNN 2010 and the first trained DNN 800, the third DNN 2010, the trained first DNN 800, and the trained A plurality of frames of the first training image 2002 , the second training image, and the third training image 2004 may be sequentially acquired through the second DNN 300 and the fourth DNN 2020 .

For joint training of the third DNN 2010 and the fourth DNN 2020 , an original training image 2001 is input to the third DNN 2010 and the trained first DNN 800 . The original training image 2001 input to the third DNN 2010 and the first trained DNN 800 is deblocking filtered, AI downscaled, and output as the first training image 2002, and the first training image 2002 ) is input to the trained second DNN 300 and fourth DNN 2020. As a result of AI upscaling and deblocking filtering on the first training image 2002, a third training image 2004 is output.

Referring to FIG. 20 , a first training image 2002 is input to the second trained DNN 300 , and according to an embodiment, the first training image 2002 is first encoded and first decoded through a process. The obtained second training image may be input to the trained second DNN 300 . In order to input the second training image to the second DNN, any one of MPEG-2, H.264, MPEG-4, HEVC, VC-1, VP8, VP9, and AV1 codecs may be used. Specifically, in the first encoding of the first training image 2002 and the first decoding of image data corresponding to the first training image 2002, MPEG-2, H.264, MPEG-4, HEVC, VC-1 , VP8, VP9, and any one codec of AV1 may be used.

Referring to FIG. 20 , the legacy downscaled reduction training from the original training image 2001 is separate from the output of the first training image 2002 through the third DNN 2010 and the first trained DNN 800 . An image 2003 is acquired. Here, the legacy downscale may include at least one of a bilinear scale, a bicubic scale, a lanczos scale, and a stair step scale.

In order to prevent the structural features of the first image 1325 from greatly deviating based on the structural features of the original image 1305, a reduced training image 2003 that preserves the structural features of the original training image 2001 is obtained. .

Before the training proceeds, the third DNN 2010 and the fourth DNN 2020 may be set with predetermined DNN configuration information. As training progresses, structural loss information 2030 , complexity loss information 2040 , and quality loss information 2050 may be determined.

The structural loss information 2030 may be determined based on a comparison result between the reduced training image 2003 and the first training image 1102 . In one example, the structural loss information 2030 may correspond to a difference between the structural information of the reduced training image 2003 and the structural information of the first training image 2002 . The structural information may include various features that can be extracted from the image, such as luminance, contrast, and histogram of the image. The structural loss information 2030 indicates to what extent the structural information of the original training image 2001 is maintained in the first training image 2002 . As the structural loss information 2030 is smaller, the structural information of the first training image 2002 is similar to the structural information of the original training image 2001 .

The complexity loss information 2040 may be determined based on the spatial complexity of the first training image 2002 . In an example, as the spatial complexity, a total variance value of the first training image 2002 may be used. The complexity loss information 2040 is related to a bit rate of image data obtained by first encoding the first training image 2002 . The smaller the complexity loss information 2040 is, the smaller the bit rate of image data is defined.

The quality loss information 2050 may be determined based on a comparison result between the original training image 2001 and the third training image 2004 . Quality loss information 2050 includes an L1-norm value, an L2-norm value, a Structural Similarity (SSIM) value, and a Peak Signal-To (PSNR-HVS) value for the difference between the original training image 2001 and the third training image 2004. It may include at least one of a Noise Ratio-Human Vision System) value, a Multiscale SSIM (MS-SSIM) value, a Variance Inflation Factor (VIF) value, and a Video Multimethod Assessment Fusion (VMAF) value. The quality loss information 2050 indicates how similar the third training image 2004 is to the original training image 2001 . As the quality loss information 2050 is smaller, the third training image 2004 is more similar to the original training image 2001 .

Referring to FIG. 20 , structural loss information 2030, complexity loss information 2040, and quality loss information 2050 are used for training of the third DNN 2010, and the quality loss information 2050 is the fourth DNN ( 2020) is used for training. That is, the quality loss information 2050 is used for both training of the third DNN 2010 and the fourth DNN 2020 .

The third DNN 2010 may update parameters such that the final loss information determined based on the structural loss information 2030 , the complexity loss information 2040 , and the quality loss information 2050 is reduced or minimized. Also, the fourth DNN 2020 may update the parameter so that the quality loss information 2050 is reduced or minimized.

Final loss information for training of the third DNN 2010 and the fourth DNN 2020 may be determined as in Equation 2 below.

[Equation 2]

LossDDS = e*structural loss information+f*complexity loss information+g*quality loss information

LossDUS = h*quality loss information

In Equation 2, LossDDS represents final loss information to be reduced or minimized for training of the third DNN 2010, and LossDUS represents final loss information to be reduced or minimized for training of the fourth DNN 2020. indicates. Also, e, f, g, and h may correspond to predetermined weights.

That is, the third DNN 2010 updates parameters in a direction in which LossDDS of Equation 2 is decreased, and the fourth DNN 2020 updates parameters in a direction in which LossDUS decreases. When the parameters of the third DNN 2010 are updated according to the LossDDS derived in the training process, the first training image 2002 obtained based on the updated parameters is different from the first training image 2002 in the previous training process. and, accordingly, the third training image 2004 is also different from the third training image 2004 in the previous training process. When the third training image 2004 is different from the third training image 2004 in the previous training process, quality loss information 2050 is also newly determined, and accordingly, the fourth DNN 2020 updates parameters. When the quality loss information 2050 is newly determined, LossDDS is also newly determined, so the third DNN 2010 updates parameters according to the newly determined LossDDS. That is, the parameter update of the third DNN 2010 causes the parameter update of the fourth DNN 2020 , and the parameter update of the fourth DNN 2020 causes the parameter update of the third DNN 2010 . In other words, since the third DNN 2010 and the fourth DNN 2020 are jointly trained through sharing of quality loss information 2050, the parameters of the third DNN 2010 and the parameters of the fourth DNN 2020 are They can be optimized in relation to each other.

Referring to Equation 2, it can be seen that LossDUS is determined according to quality loss information 2050, but this is an example, and LossDUS is at least one of structural loss information 2030 and complexity loss information 2040, It may be determined based on the quality loss information 2050 .

Previously, it has been described that the deblocking predictor 1430 of the AI decoding apparatus 1400 and the deblocking predictor 1530 of the AI encoding apparatus 1500 store filter setting information, and the filter setting information includes a plurality of DNN settings. It may also contain information. A description will be given of a method for training each of the plurality of DNN configuration information stored in the deblocking prediction unit 1430 and the deblocking prediction unit 1530 .

As described in relation to Equation 2, in the case of the third DNN 2010, the degree of similarity between the structural information of the first training image 2002 and the structural information of the original training image 2001 (structural loss information 2030) ), the bit rate (complexity loss information 2040) of the image data obtained as a result of the first encoding of the first training image 2002 and the difference between the third training image 2004 and the original training image 2001 (quality loss) The parameter is updated in consideration of the information 2050).

In detail, while making it possible to obtain a first training image 2002 similar to the structural information of the original training image 2001 and having a small bit rate of image data obtained when the first encoding is performed, the first training image The parameters of the third DNN 2010 may be updated so that the fourth DNN 2020 that AI upscales 2002 may acquire a third training image 2004 similar to the original training image 2001 .

By adjusting the weights of e, f, and g in Equation 2, the directions in which the parameters of the third DNN 2010 are optimized are different. For example, when the weight of f is determined to be high, the parameter of the third DNN 2010 may be updated with more importance on lowering the bit rate than the quality of the third training image 2004 . In addition, when the weight of g is determined to be high, it is more important to increase the quality of the third training image 2004 than to increase the bit rate or to maintain the structural information of the original training image 2001. 3 A parameter of the DNN 2010 may be updated.

Also, the direction in which parameters of the third DNN 2010 are optimized may be different according to the type of a codec used to first encode the first training image 2002 . This is because, depending on the type of codec, the second training image to be input to the fourth DNN 2020 may vary.

That is, the parameters of the third DNN 2010 and the parameters of the fourth DNN 2020 are linked based on the weight e, the weight f, the weight g, and the type of the codec for the first encoding of the first training image 2002 . so it can be updated. Therefore, when each of the weight e, the weight f, and the weight g is determined as a predetermined value and the type of the codec is determined as a predetermined type, when the third DNN 2010 and the fourth DNN 2020 are trained, they are linked to each other. The optimized parameters of the third DNN 2010 and the parameters of the fourth DNN 2020 may be determined.

And, after changing the weight e, the weight f, the weight g, and the type of codec, if the third DNN 2010 and the fourth DNN 2020 are trained, the parameters of the third DNN 2010 optimized in connection with each other and parameters of the fourth DNN 2020 may be determined. In other words, when the third DNN 2010 and the fourth DNN 2020 are trained while changing the respective values of the weight e, the weight f, the weight g, and the type of codec, the plurality of DNN setting information trained in connection with each other is transmitted to the third DNN It can be determined in the DNN (2010) and the fourth DNN (2020).

21 is a diagram for explaining a training process of the third DNN 2010 and the fourth DNN 2020 by the training apparatus 2100 .

Training of the third DNN 2010 and the fourth DNN 2020 described with reference to FIG. 20 may be performed by the training apparatus 2100 . The training apparatus 2100 includes a third DNN 2010 , a fourth DNN 2020 , a first trained DNN 800 , and a second trained DNN 300 . The training device 2100 may be, for example, the AI encoding device 1500 or a separate server. DNN setting information of the fourth DNN 2020 obtained as a result of training is stored in the AI decoding apparatus 1400 .

Referring to FIG. 21 , the training apparatus 2100 initially sets DNN configuration information of the third DNN 2010 and the fourth DNN 2020 ( S2110 and S2115 ). Accordingly, the third DNN 2010 and the fourth DNN 2020 may operate according to predetermined DNN configuration information. The DNN setting information includes the number of convolution layers included in the third DNN 2010 and the fourth DNN 2020, the number of filter kernels for each convolutional layer, the size of filter kernels for each convolutional layer, and parameters of each filter kernel. It may include information about at least one.

The training apparatus 2100 inputs the original training image 2001 into the third DNN 2010 (S2120). The original training image 1101 may include at least one frame constituting a still image or a moving image. The filtering result is output from the third DNN 2010 and input to the trained first DNN (S2125).

The third DNN 2010 processes the original training image 2001 according to the initially set DNN setting information, and the AI downscaled through the first DNN 800 that is deblocking filtered and trained from the original training image 2001. A first training image 2002 is output (S2130). 21 shows that the first training image 2002 output from the third DNN 2010 and the first trained DNN 800 is directly input to the second trained DNN 300, but the third DNN ( 2010) and the first training image 2002 output from the first trained DNN 800 may be input to the second DNN 300 trained by the training apparatus 2100 . Also, the training apparatus 2100 may first encode and first decode the first training image 2002 using a predetermined codec, and then input the second training image to the trained second DNN 300 .

The second trained DNN 300 AI upscales the first training image 2002 or the second training image (S2135), and the third training is deblocking filtered from the first training image 2002 or the second training image. An image 2004 is output (S2140).

The training apparatus 2100 calculates complexity loss information 2040 based on the first training image 2002 ( S2145 ).

The training apparatus 2100 calculates structural loss information 2030 by comparing the reduced training image 2003 with the first training image 2002 ( S2150 ).

The training apparatus 2100 calculates quality loss information 2050 by comparing the original training image 2001 with the third training image 2004 ( S1255 ).

The third DNN 2010 updates the initially set DNN configuration information through a back propagation process based on the final loss information (S2160). The training apparatus 2100 may calculate final loss information for training the third DNN 2010 based on the complexity loss information 2040 , the structural loss information 2030 , and the quality loss information 2050 .

The fourth DNN 2020 updates the initially set DNN configuration information through a reverse transcription process based on quality loss information or final loss information (S2165). The training apparatus 2100 may calculate final loss information for training of the fourth DNN 2020 based on the quality loss information 2050 .

Thereafter, the training apparatus 2100 , the third DNN 2010 , and the fourth DNN 2020 update the DNN configuration information while repeating processes S2110 to S2165 until the final loss information is minimized. At this time, during each iteration process, the third DNN 2010 and the fourth DNN 2020 operate according to the DNN configuration information updated in the previous process.

22 is a flowchart illustrating an AI encoding method according to an embodiment.

In step S2210, the AI encoding apparatus 1500 acquires the original image, previously encoded frame information, and network environment information.

According to an embodiment, the previous frame information may indicate whether a block artifact of an encoded previous frame is detected.

According to an embodiment, the network environment information may indicate a change in at least one of a bit rate and a bandwidth of a network channel.

In step S2220, the AI encoding apparatus 1500 obtains deblocking filter setting information based on the original image, the previously encoded frame information, and the transmission environment network information.

According to an embodiment, when the network environment information indicates that the network environment does not change, the deblocking filter setting information may not be obtained and deblocking filtering may not be performed. Specifically, in a normal situation, other than when several people connect at the same time or there is a change in a network channel outdoors, a blocking phenomenon in which image quality deteriorates does not occur, so deblocking filtering does not need to be additionally performed.

According to an embodiment, the deblocking filter setting information may include a blocking area map indicating an area in which block artifacts are generated. Specifically, as described above in FIG. 13 , the blocking area map may be determined through a weighted average of the A score maps to the G score maps.

In step S2230, the AI encoding apparatus 1500 applies deblocking filtering to the original image based on the deblocking filter setting information and acquires an AI downscaled first image through a downscaling DNN.

According to an embodiment, the deblocking filtering in AI encoding may be filtering that redistributes an allocation amount of bits based on deblocking filter setting information. Specifically, the deblocking filter reduces the bit amount by blurring the complex and fast-moving area of the image, and distributes the saved bit amount to the simple area to reduce the bit allocation so that the blocking phenomenon in the simple area is less noticeable. can be redistributed.

In step S2240, the AI encoding apparatus 1500 first encodes the first image to generate image data.

In step S2250 , the AI encoding apparatus 1500 transmits the deblocking filter setting information, AI data including information related to the AI downscaling, and the image data.

23 is a flowchart illustrating an AI decoding method according to an embodiment.

In step S2310, the AI decoding apparatus 1400 acquires image data generated as a result of the first encoding of the first image, AI data related to AI downscaling from the original image to the first image, and deblocking filter setting information. .

According to an embodiment, if deblocking filter setting information is not obtained, deblocking filtering may not be performed. Specifically, the fact that the deblocking filter setting information is not obtained means that it is a normal situation, not when a deblocking filter must be applied, that is, when several people connect at the same time or there is a change in a network channel outdoors. This is because it does not need to be performed additionally.

According to an embodiment, the deblocking filter setting information may include a blocking area map indicating an area in which block artifacts are generated. Specifically, as described above in FIG. 13 , the blocking area map may be determined through a weighted average of the A score maps to the G score maps.

In operation S2320, the AI decoding apparatus 1400 first decodes the image data to obtain a second image corresponding to the first image.

In operation S2330, the AI decoding apparatus 1400 obtains DNN setting information for AI upscaling of the second image among a plurality of DNN setting information, based on the AI data.

In step S2340, the AI decoding apparatus 1400 is upscaled AI from the second image through a deep neural network (DNN) for upscaling that operates with the obtained DNN setting information and deblocking filter based on the one deblocking filter. A reconstructed image to which blocking filtering is applied is generated.

According to an embodiment, deblocking filtering in AI decoding may be filtering for blurring a block boundary of an image and removing blocking artifacts based on deblocking filter setting information. Specifically, the deblocking filter may perform deblocking filtering according to each region by blurring the boundary between blocks of an image and removing blocking artifacts.

Meanwhile, the above-described embodiments of the present disclosure can be written as a program or instruction that can be executed in a computer, and the written program or instruction can be stored in a medium.

The medium may continuously store a computer-executable program or instructions, or may be a temporary storage for execution or download. In addition, the medium may be various recording means or storage means in the form of a single or several hardware combined, it is not limited to a medium directly connected to any computer system, and may exist distributed on a network. Examples of the medium include a hard disk, a magnetic medium such as a floppy disk and a magnetic tape, an optical recording medium such as CD-ROM and DVD, a magneto-optical medium such as a floppy disk, and those configured to store program instructions, including ROM, RAM, flash memory, and the like. In addition, examples of other media include recording media or storage media managed by an app store that distributes applications, sites that supply or distribute various other software, and servers.

Meanwhile, the above-described DNN-related model may be implemented as a software module. When implemented as a software module (eg, a program module including instructions), the DNN model may be stored in a computer-readable recording medium.

In addition, the DNN model may be integrated in the form of a hardware chip to become a part of the aforementioned AI decoding apparatus 1400 or AI encoding apparatus 1500 . For example, the DNN model may be fabricated in the form of a dedicated hardware chip for artificial intelligence, or as part of an existing general-purpose processor (eg, CPU or application processor) or graphics-only processor (eg, GPU). it might be

Alternatively, the DNN model may be provided in the form of downloadable software. The computer program product may include a product (eg, a downloadable application) in the form of a software program distributed electronically through a manufacturer or an electronic market. For electronic distribution, at least a portion of the software program may be stored in a storage medium or may be temporarily generated. In this case, the storage medium may be a server of a manufacturer or an electronic market, or a storage medium of a relay server.

In the above, the technical idea of the present disclosure has been described in detail with reference to preferred embodiments, but the technical idea of the present disclosure is not limited to the above embodiments, and those of ordinary skill in the art within the scope of the technical spirit of the present disclosure Various modifications and changes are possible by the person.

Claims (17)

Memory; and
including a processor;
The memory stores instructions executable by the processor and
The processor is
Obtain the original video, previously encoded frame information, and network environment information,
obtaining deblocking filter setting information based on the original image, the previously encoded frame information, and the network environment information;
Applying deblocking filtering to the original image based on the deblocking filter setting information and obtaining an AI downscaled first image through DNN for downscaling,
generating image data by first encoding the first image;
The AI encoding apparatus, characterized in that the deblocking filter setting information, AI data including information related to the AI downscaling, and the image data are transmitted. According to claim 1,
The deblocking filter setting information includes a blocking region map indicating a region in which block artifacts are generated. 3. The method of claim 2,
The blocking region map is determined based on at least one of a complexity and a motion gradient of the original image. According to claim 1,
The deblocking filtering is filtering for redistributing an allocation amount of bits based on the deblocking filter setting information. According to claim 1,
The DNN setting information of the downscaling DNN is obtained through joint training of the downscaling DNN and the upscaling DNN for the AI upscaling of the second image. According to claim 1,
When the network environment information indicates that the network environment does not change, the deblocking filter setting information is not obtained, and the deblocking filtering is not performed. According to claim 1,
The network environment information indicates a change in at least one of a bit rate and a bandwidth of a network channel, AI encoding apparatus. According to claim 1,
The previous frame information indicates whether a block artifact of an encoded previous frame is detected. According to claim 1,
Further obtaining information on a quantization parameter difference value map and rate control based on the original image, the previously encoded frame information, and the network environment information,
and performing the first encoding based on the quantization parameter differential value map and the rate control information. Memory; and
including a processor;
The memory stores instructions executable by the processor and
The processor is
Obtaining image data generated as a result of the first encoding of the first image, AI data related to AI downscaling from the original image to the first image, and deblocking filter setting information,
obtaining a second image corresponding to the first image by first decoding the image data;
Obtaining DNN setting information for AI upscaling of the second image among a plurality of DNN setting information, based on the AI data,
AI upscale from the second image through a deep neural network (DNN) for upscaling that operates with the obtained DNN setting information and generates a reconstructed image to which deblocking filtering is applied based on the deblocking filter setting information An AI decryption device. 11. The method of claim 10,
The deblocking filter setting information includes a blocking region map indicating a region in which block artifacts are generated. 11. The method of claim 10,
The blocking area map is determined based on at least one of a complexity and a motion gradient of an original image. 11. The method of claim 10,
The deblocking filtering is filtering for blurring a block boundary of an image and removing blocking artifacts based on the deblocking filter setting information. 11. The method of claim 10,
The plurality of DNN setting information is obtained through joint training between the upscaling DNN and the downscaling DNN used for the AI downscaling of the original image. 11. The method of claim 10,
If the deblocking filter setting information is not obtained, the deblocking filtering is not performed, AI decoding apparatus. An AI encoding method by an AI encoding device, comprising:
obtaining an original image, previously encoded frame information, and network environment information;
obtaining deblocking filter setting information based on the original image, the previously encoded frame information, and the network environment information;
applying deblocking filtering to the original image based on the deblocking fiter setting information and obtaining an AI downscaled first image through a downscaling DNN;
generating image data by first encoding the first image;
and transmitting the deblocking filter setting information, AI data including information related to the AI downscaling, and the image data. In the AI decoding method by the AI decoding apparatus,
acquiring image data generated as a result of the first encoding of the first image, AI data related to AI downscaling from the original image to the first image, and deblocking filter setting information;
obtaining a second image corresponding to the first image by first decoding the image data;
obtaining DNN setting information for AI upscaling of the second image from among a plurality of DNN setting information, based on the AI data; and
Generating a reconstructed image to which AI is upscaled from the second image through a deep neural network (DNN) for upscaling that operates with the obtained DNN setting information and to which deblocking filtering is applied based on the deblocking filter setting information AI decryption method comprising.

Images