Learning device, autoencoding device, learning method, autoencoding method, and program

By integrating noise addition within a (K-1)-dimensional sphere and scalar quantization, the self-encoding process addresses inefficiencies in vector quantization, reducing computational burden and resource needs for high-dimensional data processing.

JP7583338B2Active Publication Date: 2024-11-14NIPPON TELEGRAPH & TELEPHONE CORP
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2023563383
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2021-11-24
Publication Date
2024-11-14
Estimated Expiration
2041-11-24

AI Technical Summary

Technical Problem

Conventional vector quantization techniques face stability issues, high computational burden, and resource constraints, particularly when encoding high-dimensional data, leading to inefficiencies in self-encoding processing.

Method used

A self-encoding process using vector quantization that incorporates noise addition by randomly sampling within a (K-1)-dimensional sphere circumscribing the Voronoi region in a K-dimensional vector space, combined with scalar quantization of auxiliary features, to update encoding and decoding processes through learning.

Benefits of technology

This approach reduces the computational burden and resource requirements for self-encoding, enabling efficient processing of high-dimensional data without memory shortages.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007583338000013
    Figure 0007583338000013
  • Figure 0007583338000014
    Figure 0007583338000014
  • Figure 0007583338000015
    Figure 0007583338000015
Patent Text Reader

Abstract

This learning device comprises a learning unit that executes learning to update coding and decoding processes in autoencoding involving using a primary data feature that is a feature to be subjected to autoencoding and an auxiliary feature that is a feature of the primary data feature, and vector-quantizing the primary data feature and scalar-quantizing the auxiliary feature. The learning unit assigns a noise to the primary data feature through learning, and the assignment of the noise is either one of: a process for assigning, as the noise, samples in a voronoi region of a K-dimensional vector space among samples generated at random in a (K-1)-dimensional sphere circumscribing the voronoi region; a process for assigning, as the noise, samples generated uniformly in a (K-1)-dimensional sphere with a volume that approximates the volume of the voronoi region; and a process for assigning, as the noise, samples generated uniformly in a region that is obtained by dividing a vector space in which representative vectors are arranged in a lattice pattern and is a hyper-rectangular region including one lattice point of the vector space.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a learning device, an autoencoding device, a learning method, an autoencoding method, and a program. [Background technology]

[0002] In data compression using deep learning, there are two types of quantization processing, scalar quantization and vector quantization, with vector quantization having higher compression performance. Attempts at vector quantization include those described in the following Non-Patent Document 1 and Non-Patent Document 2. The technology described in Non-Patent Document 1 is a technology called VQVAE. In the technology described in Non-Patent Document 1, end-to-end learning is not performed, and the distribution of feature vectors is assumed to be uniform, and optimization of encoding and decoding and optimization of representative vectors are performed separately. The technology in Non-Patent Document 2 is a technology called Soft to Hard. In the technology in Non-Patent Document 2, the quantization processing is approximated using a softmax function, and the occurrence probability of the representative vector is approximated by a histogram to perform end-to-end learning. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Ali Razavi, et.al.,"Generating Diverse High-Fidelity Images with VQ-VAE-2,",2019. [Non-Patent Document 2] Eirikur Agustsson, et.al.,"Soft-to-Hard Vector Quantization for End-to-End Learning Compressible Representations,",2017. Summary of the Invention [Problem to be solved by the invention]

[0004] However, conventional techniques such as the technique described in Non-Patent Document 1 and the technique described in Non-Patent Document 2 have problems with obtaining a self-encoding process using vector quantization. The problems include, for example, stability problems, processing performance problems, and problems such as an explosive increase in the amount of calculations or memory usage when the dimension of the vector is large or when encoding at a high rate. Thus, in conventional techniques, the burden required to obtain the self-encoding process using vector quantization may be large. As a result, the burden required for self-encoding using vector quantization may be large, such as not being able to realize self-encoding using vector quantization due to a lack of computer resources.

[0005] In view of the above circumstances, an object of the present invention is to provide a technique for reducing the burden required for self-encoding using vector quantization. [Means for solving the problem]

[0006] One aspect of the present invention is an auto-encoding process using vector quantization, which uses main data features that are features to be auto-encoded and auxiliary features that are features of the main data features, and includes a learning unit that updates encoding and decoding processes in the auto-encoding process by learning, the learning unit executing a process of adding noise to the main data features in the learning, and the process of adding noise is performed by applying a vector quantization noise to the main data features in a K-dimensional (K is an integer equal to or greater than 2) vector space. The learning device performs one of the following processes: of samples randomly generated within a (K-1)-dimensional sphere circumscribing a Voronoi region, adding as noise samples within the Voronoi region; of samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space; and of samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, that includes one lattice point of the vector space, and that has a shape of a hypercuboid.

[0007] One aspect of the present invention includes an auto-encoding target acquisition unit that acquires an object of auto-encoding, and a learning unit that updates, by learning, encoding and decoding processes in the auto-encoding process using vector quantization, the auto-encoding process using main data features that are features of the object of auto-encoding and auxiliary features that are features of the main data features, and that performs vector quantization of the main data features and scalar quantization of the auxiliary features. In the learning, the learning unit executes a process of adding noise to the main data features, and the process of adding noise is performed by applying noise to samples randomly generated within a (K-1)-dimensional sphere circumscribing a Voronoi region in a K-dimensional (K is an integer equal to or greater than 2) vector space. and a self-encoding execution unit that performs self-encoding by vector quantization of the target acquired by the self-encoding target acquisition unit using a learned encoding process and a learned decoding process obtained using a learning device, the self-encoding device being one of the following processes: a process of adding samples within the Voronoi region as noise; a process of adding samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximating the volume of a Voronoi region in a K-dimensional vector space as noise; and a process of adding samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid as noise.

[0008] One aspect of the present invention is an auto-encoding process using vector quantization, which uses main data features that are features to be auto-encoded and auxiliary features that are features of the main data features, and the auto-encoding process performs vector quantization of the main data features and scalar quantization of the auxiliary features, and includes a learning step of updating encoding and decoding processes by learning, wherein the learning step executes a process of adding noise to the main data features, and the process of adding noise is carried out in a K-dimensional (K is an integer equal to or greater than 2) vector space. the learning method is one of the following: of samples randomly generated within a (K-1)-dimensional sphere circumscribing a Voronoi region in a K-dimensional vector space, samples within the Voronoi region are assigned as noise; samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximating the volume of a Voronoi region in a K-dimensional vector space are assigned as noise; and samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid are assigned as noise.

[0009] One aspect of the present invention includes an auto-encoding target acquisition step of acquiring an auto-encoding target, and an auto-encoding process using vector quantization, the auto-encoding process using main data features which are features of the auto-encoding target and auxiliary features which are features of the main data features, and performing vector quantization of the main data features and scalar quantization of the auxiliary features. In the learning, the learning unit executes a process of adding noise to the main data features, and the process of adding noise includes: a process of adding, as noise, samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space; or a process of adding, as noise, samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid; and an auto-encoding execution step of performing auto-encoding by vector quantization of the target acquired in the auto-encoding target acquisition step using a learned encoding process and a learned decoding process obtained using a learning device, the process being one of the following processes.

[0010] One aspect of the present invention is a program for causing a computer to function as the above-mentioned learning device.

[0011] One aspect of the present invention is a program for causing a computer to function as the above-described self-encoding device. Effect of the Invention

[0012] According to the present invention, it is possible to reduce the burden required for auto-encoding using vector quantization. [Brief description of the drawings]

[0013] [Figure 1] FIG. 1 is an explanatory diagram illustrating an overview of a learning device according to an embodiment. [Diagram 2] FIG. 2 is an explanatory diagram for explaining LatticeVQ according to an embodiment. [Diagram 3] FIG. 4 is a first explanatory diagram illustrating an example of adding noise according to the embodiment. [Figure 4] FIG. 2 is a second explanatory diagram illustrating an example of adding noise in the embodiment. [Diagram 5] 5A and 5B are explanatory diagrams illustrating the relationship between a cumulative distribution function and occurrence probability in the embodiment. [Figure 6] FIG. 4 is an explanatory diagram for explaining hyper-rectangular parallelepiped division in the embodiment. [Figure 7] FIG. 4 is a diagram showing an example of a processing flow executed by a learning unit in the embodiment. [Figure 8] FIG. 1 is a first explanatory diagram illustrating an overview of an auto-encoding device according to an embodiment. [Figure 9] FIG. 2 is a second explanatory diagram illustrating an overview of the self-encoding device according to the embodiment. [Figure 10] 10 is a flowchart showing an example of a flow of a process executed by an encoder in the embodiment. [Figure 11] 10 is a flowchart showing an example of a flow of a process executed by a decoder in the embodiment. [Figure 12] FIG. 2 is a diagram showing an example of a hardware configuration of a learning device according to an embodiment. [Figure 13] FIG. 2 is a diagram showing an example of the configuration of a control unit included in the learning device according to the embodiment. [Figure 14] FIG. 2 is a diagram showing an example of a hardware configuration of a self-encoding device according to an embodiment. [Figure 15] FIG. 2 is a diagram showing an example of the configuration of a control unit included in the self-encoding device according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] (Embodiment) FIG. 1 is an explanatory diagram for explaining an overview of a learning device 1 according to an embodiment. The learning device 1 includes a learning unit 10. The learning unit 10 performs learning so as to improve the performance of auto-encoding using vector quantization of data expressed as a tensor. The performance of auto-encoding is evaluated by the smallness of the RD cost (D+λR), which is the sum of the error D between the original data and the restored data and the code amount R of the data weighted by the Lagrangian constant λ. The smaller the RD cost, the better the RD performance. The learning means machine learning. The learning is, for example, deep learning. As is well known, the auto-encoding of data means data compression.

[0015] The learning unit 10 includes a learning network 100 and an optimization unit 113. The learning network 100 is a neural network. The learning network 100 includes a main data acquisition unit 101, a main data side encoding unit 102, an auxiliary data side encoding unit 103, a main data side noise adding unit 104, an auxiliary data side noise adding unit 105, an auxiliary data side probability estimating unit 106, an auxiliary data side decoding unit 107, an auxiliary entropy acquisition unit 108, a main data side probability estimating unit 109, a main entropy acquisition unit 110, a main data side decoding unit 111, and a reconstruction error calculation unit 112. The optimization unit 113 updates the learning network 100 based on the output of the learning network 100, as will be described in detail later.

[0016] The main data acquisition unit 101 acquires data expressed by tensors as main data. The data expressed by tensors is, for example, image data. The data expressed by tensors may be, for example, time-series data of audio of one or more channels. Hereinafter, the data acquired by the main data acquisition unit 101 is referred to as main data.

[0017] The main data side encoding unit 102 executes a main data feature acquisition process. The main data feature acquisition process is a process of encoding the main data. Since encoding is a process of acquiring information indicating the features of the object of encoding, encoding is a process of acquiring information indicating the features.

[0018] Therefore, the encoding of the main data by the main data side encoding unit 102 is a process of acquiring the features of the main data. Therefore, the main data feature acquisition process is a process of acquiring the main data features. The main data features are encoded main data. The contents of the main data feature acquisition process are updated by learning. In other words, the contents of the process executed by the main data side encoding unit 102 are updated by learning.

[0019] The auxiliary data side encoding unit 103 further encodes the main data features. As described above, encoding is a process of acquiring information indicating features, so encoding of the main data features by the auxiliary data side encoding unit 103 is a process of acquiring the features of the main data features. Therefore, hereinafter, information obtained by further encoding the encoded main data is referred to as auxiliary features. In other words, the auxiliary features are information indicating the features of the main data features. Since the auxiliary features are information obtained by encoding the main data features, the entropy of the auxiliary features is information with a smaller entropy than the main data features.

[0020] Hereinafter, the process of encoding the main data features will be referred to as the auxiliary feature acquisition process. The content of the auxiliary feature acquisition process is updated through learning. That is, the content of the process executed by the auxiliary data side encoding unit 103 is updated through learning.

[0021] The contents of the auxiliary feature acquisition process are updated by learning so that the auxiliary features include a larger amount of information in the main data statistics information. The main data statistics information is information indicating the statistics of each probability distribution to which the values ​​of each element of the tensor expressing the main data follow. The statistics of the probability distribution are, for example, the scatter. The statistics of the probability distribution may be not only the scatter but also a pair of the scatter and a representative value.

[0022] As explained in the fields of information processing and information theory, data generally follows a probability distribution, such as the probability distribution of each Roman character that appears in an English sentence. The values ​​of each element of a tensor that represents the main data also follow a probability distribution.

[0023] The main data-side noise adding unit 104 executes a noise-added main data feature acquisition process. The noise-added main data feature acquisition process is a process of executing a vector noise addition process on the main data feature. The vector noise addition process is a process of processing a K-dimensional vector (hereinafter referred to as a "noise-added vector") having a predetermined number of elements K (K is an integer equal to or greater than 2) as a processing object, and is a process of adding noise to the processing object. Therefore, the noise-added main data feature acquisition process is a process of acquiring information in which noise has been added to the main data feature (hereinafter referred to as a "noise-added main data feature").

[0024] Specifically, the noise-added vector is a K-dimensional vector included in a tensor that represents the main data feature amount. A K-dimensional vector included in a tensor means a vector whose k-th element (k is an integer between 1 and K) is the k-th element among K consecutive elements of all elements of the tensor.

[0025] The number of elements K of the noise-added vector is the number of elements mapped to one code by vector quantization by a device that performs vector quantization using the results of learning by the learning unit 10. Hereinafter, the results of learning by the learning unit 10 are referred to as network learning results. Therefore, when K elements are collectively mapped to one code by vector quantization using the results of network learning, the number of elements of the noise-added vector is K. Note that when K elements are collectively mapped to one code by quantization, the obtained code is an index indicating a K-dimensional vector.

[0026] Therefore, the number of elements K of the noise-added vector is a predetermined number. Note that, since vector quantization of data is encoding of data, vector quantization using the network learning result means encoding of data using the network learning result.

[0027] Here, a method for adding noise will be described, but prior to the explanation of adding noise, LatticeVQ will be described.

[0028] <latticevq> Vector quantization requires a representative vector. A lattice is a set of lattice points in a vector space. LatticeVQ is vector quantization in which representative vectors are arranged in a lattice pattern in a vector space. In other words, LatticeVQ is vector quantization that satisfies the condition that representative vectors are arranged in a lattice pattern in a vector space. LatticeVQ is known to have better RD performance than scalar quantization, except under certain conditions.

[0029] Fig. 2 is an explanatory diagram illustrating LatticeVQ in an embodiment. More specifically, Fig. 2 is an explanatory diagram illustrating LatticeVQ when the number of elements of a noise-added vector is two.

[0030] Figure 2 shows an example of the A2 lattice. In the case of two dimensions, the A2 lattice is one of the types of lattices. In the case of eight dimensions, the E8 lattice is one of the types of lattices. In the case of 24 dimensions, the Reach lattice is one of the types of lattices. These are LatticeVQ lattices that maximize RD performance for a uniform distribution.

[0031] Hereinafter, a space in which the representative vectors are arranged in a lattice pattern will be referred to as a lattice space. In other words, the representative vectors are located at each lattice point of the lattice space. Figure 2 is also a diagram showing an example of a lattice space. In the example of Figure 2, the lattice space is two-dimensional, but if the representative vectors are K-dimensional, the lattice space is also K-dimensional.

[0032] <Adding noise> An example of adding noise will be described with reference to Fig. 3 and Fig. 4, taking the case where the number of elements is 2 as an example. Fig. 3 is a first explanatory diagram for explaining an example of adding noise in the embodiment. Fig. 4 is a second explanatory diagram for explaining an example of adding noise in the embodiment.

[0033] When adding noise, first, multiple random K-dimensional vectors that are uniformly distributed within a (K-1)-dimensional sphere circumscribing the lattice unit area of ​​the origin lattice (hereinafter referred to as the "target lattice unit area") are generated and placed in the lattice space. For example, in the example of Figure 4, when adding noise, first, they are generated until the number of internal points in Figure 4 becomes at least 1. The lattice unit areas are each area resulting from dividing the lattice space into multiple areas so that the division conditions are satisfied.

[0034] The division conditions are that the size and shape of each region are the same, and each region contains one lattice point of the lattice space. Therefore, a lattice unit region is a region that divides a vector space in which representative vectors are arranged in a lattice pattern, and contains one lattice point of the vector space.

[0035] In addition, the (K-1)-dimensional sphere is a circle if K = 2, a sphere if K = 3, and a hypersphere if K is 4 or more. Region B1 in Figure 3 is an example of a target lattice unit region.

[0036] 3 and 4, the lattice unit region is a Voronoi region, which is each of the regions obtained as a result of dividing a metric space such as a lattice space into a plurality of regions by Voronoi division.

[0037] The noise points in Fig. 3 are an example of samples arranged within the circumscribing circle of a target lattice unit area. Specifically, the process of arranging samples within the lattice space is a process of acquiring coordinates within the lattice space. Since a lattice unit area in the lattice space is an area within the lattice space, the process of arranging samples within the circumscribing circle of any one lattice unit area is a process of acquiring coordinates within the circumscribing circle.

[0038] Next, of the placed noise points, noise points located within the target lattice unit area (hereinafter referred to as "internal points") and noise points located outside the target lattice unit area (hereinafter referred to as "exclusion points") are distinguished as shown in the example of Figure 4. Specifically, the process of distinguishing between internal points and exclusion points is a process of determining, for each noise point, whether the noise point has coordinates within the target lattice unit area or coordinates outside the target lattice unit area, based on its coordinates. Specifically, the determination process is a process in which the noise points are vector quantized using equation (11), and points that match the origin lattice are determined to have coordinates within the area.

[0039] Next, one of the multiple noise points located in the target lattice unit area is randomly selected. The selected noise point is then added to one noise-added target vector. The adding process specifically means a process of adding the noise-added target vector and a position vector indicating the selected noise point.

[0040] In this way, the position vector indicating the selected noise point is a quantity determined by a random number, and is therefore a type of noise. And since the position vector indicating the selected noise point is expressed as a vector, the position vector indicating the selected noise point is noise expressed as a vector. Therefore, hereinafter, the position vector indicating the selected noise point is referred to as a noise vector. In this way, noise is added to the noise-adding target vector.

[0041] Returning to the description of FIG. 1, the auxiliary data-side noise adding unit 105 executes noise-added auxiliary feature acquisition processing. The noise-added auxiliary feature acquisition processing is processing for adding noise to the auxiliary feature. In other words, the noise-added auxiliary feature acquisition processing is processing for acquiring information in which scalar noise has been added to the auxiliary feature (hereinafter referred to as "noise-added auxiliary feature"). More specifically, the auxiliary data-side noise adding unit 105 adds scalar noise to each element of a tensor indicating the auxiliary feature. The scalar noise is, for example, uniform noise between -1 / 2 and 1 / 2.

[0042] The auxiliary data side probability estimation unit 106 executes auxiliary data side probability estimation processing. The auxiliary data side probability estimation processing is processing for estimating auxiliary data side probability based on noise-added auxiliary features. The auxiliary data side probability is information indicating the occurrence probability of each element of a tensor indicating the auxiliary features. Information on the probability distribution of each element of the tensor indicating the auxiliary features is used to estimate the occurrence probability of each element of the tensor indicating the auxiliary features.

[0043] When the probability distribution of each element of the tensor indicating the auxiliary feature is a predetermined probability distribution, the auxiliary data side probability estimation unit 106 acquires the predetermined probability distribution by, for example, reading it from a predetermined storage device such as the storage unit 14 described below. Then, the auxiliary data side probability estimation unit 106 estimates the occurrence probability of each element of the tensor indicating the auxiliary feature based on the acquired probability distribution. Note that the predetermined probability distribution is, for example, a probability distribution expressed using a cumulative distribution function.

[0044] When the probability distribution of each element of the tensor indicating the auxiliary feature is expressed by a parameterized auxiliary feature cumulative distribution function, the auxiliary data side probability estimation unit 106 estimates the occurrence probability of each element of the tensor indicating the auxiliary feature based on the parameterized auxiliary feature cumulative distribution function. The parameterized auxiliary feature cumulative distribution function is a parameterized function that indicates the probability distribution of each element of the tensor indicating the auxiliary feature. Specifically, the parameters of the parameterized auxiliary feature cumulative distribution function are parameters that change according to statistics that indicate the probability distribution of each element of the tensor indicating the auxiliary feature.

[0045] The parameter values ​​of the parameterized auxiliary feature cumulative distribution function are updated through learning, i.e., the content of the process executed by the auxiliary data side probability estimator 106 is updated through learning.

[0046] The parameterized cumulative distribution function is, for example, a parameterized sigmoid function or a softplus function. As described above, the parameter values ​​of the parameterized auxiliary feature cumulative distribution function are updated by learning, so that the parameter values ​​of the parameterized cumulative distribution function are updated by learning.

[0047] Here, the relationship between the cumulative distribution function and the occurrence probability will be described by taking the case where the random variable is one-dimensional as an example. FIG. 5 is an explanatory diagram for explaining the relationship between the cumulative distribution function and the occurrence probability in the embodiment. Image G1 is an image showing an example of a probability density function showing the probability density of a random variable q. Δ indicates the quantization step size. Image G1 shows that the value obtained by integrating the probability density function in a range Δ within a domain centered on the value q is the occurrence probability p(q) of the value q. In the one-dimensional case as in FIG. 5, the step size Δ is, for example, the size of the closed interval [-1 / 2, 1 / 2] (i.e., 1).

[0048] Image G2 is a cumulative distribution function cdf(q) obtained as a result of integrating the probability density function of image G1. In this way, the result of integrating the probability density function is expressed by a monotonically increasing function such as a sigmoid function. Note that the result of integrating the probability density function is expressed by a monotonically increasing function such as a sigmoid function regardless of the shape of the probability density function. More specifically, the cumulative distribution function is a function that satisfies the conditions cdf(-∞)=0, cdf(∞)=1, and the differential value of cdf(q) with respect to q is 0 or more.

[0049] Image G2 shows that the probability of occurrence p(q) is equal to cdf(q+Δ / 2)-cdf(q-Δ / 2). Returning to the description of FIG.

[0050] The auxiliary data side decoding unit 107 executes auxiliary feature decoding processing. The auxiliary feature decoding processing is processing for information obtained based on the auxiliary feature, and is processing for decoding the processing object. The processing object of the auxiliary feature decoding processing executed by the auxiliary data side decoding unit 107 is the noise-added auxiliary feature. Hereinafter, information obtained by decode the noise-added auxiliary feature is referred to as auxiliary data. Therefore, the auxiliary data side decoding unit 107 obtains the auxiliary data by decoding the noise-added auxiliary feature.

[0051] As described above, the more the content of the auxiliary feature acquisition process is updated through learning, the smaller the RD cost becomes. Therefore, the more the content of the auxiliary feature acquisition process is updated, the larger the maximum amount of information of the main data statistics information that the auxiliary data can contain becomes.

[0052] The content of the auxiliary feature decoding process is updated through learning so as to reduce the RD cost, that is, the content of the process executed by the auxiliary data side decoding unit 107 is updated through learning.

[0053] The auxiliary entropy acquisition unit 108 acquires auxiliary entropy based on the auxiliary data-side probability that is the result of estimation by the auxiliary data-side probability estimation unit 106. The auxiliary entropy is the entropy of the auxiliary feature.

[0054] The main data side probability estimation unit 109 executes a main data side probability estimation process. The main data side probability estimation process is a process for estimating a main data side probability based on the noise-added main data feature and auxiliary data. The main data side probability is information indicating the occurrence probability of each element of a tensor indicating the main data feature. Information on the probability distribution of each element of the tensor indicating the main data feature is used to estimate the occurrence probability of each element of the tensor indicating the main data feature.

[0055] When the probability distribution of each element of a tensor indicating the main data features is expressed using a parameterized main data feature cumulative distribution function, the main data side probability estimation unit 109 estimates the occurrence probability of each element of the tensor indicating the main data features based on the parameterized main data feature cumulative distribution function.

[0056] The parameterized main data feature cumulative distribution function is a parameterized cumulative distribution function that indicates the probability distribution of each element of a tensor that indicates the main data feature. The probability distribution is, for example, a Gaussian distribution.

[0057] Specifically, the parameters of the parametrized main data feature cumulative distribution function are statistics that represent the probability distribution of each element of the tensor that represents the main data feature. Therefore, in learning, values ​​indicated by the auxiliary data are used as the parameter values ​​of the parametrized auxiliary feature cumulative distribution function.

[0058] For example, if the probability distribution is a Gaussian distribution, the auxiliary data obtained in the trained learning network 100 indicates the representative value and the scatter of the Gaussian distribution.

[0059] <Estimation of occurrence probability> An example of the main data side probability estimation process will be described. More specifically, an example of the main data side probability estimation process when the above-mentioned LatticeVQ is used will be described. When vector quantization is performed, each vector of the main data is represented by one of the representative vectors. Therefore, the description of the main data side probability estimation process when LatticeVQ is used is, more specifically, a description of a process for estimating the occurrence probability of the representative vector when LatticeVQ is used. Hereinafter, the process for estimating the occurrence probability of the representative vector will be referred to as the representative vector occurrence probability estimation process. The representative vector occurrence probability estimation process is an example of the main data side probability estimation process.

[0060] <Representative vector occurrence probability estimation process> As described above, the main data side probability estimation process uses a parameterized main data feature cumulative distribution function. As described above, the parameterized main data feature cumulative distribution function is specifically a parameterized cumulative distribution function. As described above, a cumulative distribution function is a result of integrating a probability density function. Therefore, the representative vector occurrence probability estimation process is a process that uses a cumulative distribution function. And, the use of a cumulative distribution function in the representative vector occurrence probability estimation process means that the result of integrating a parametrized probability density function in a lattice unit domain is used.

[0061] As mentioned above, a Voronoi region is one of the lattice unit regions. Since the Voronoi tessellation is a well-known method of domain division, it is often used to obtain lattice unit regions. However, when the lattice space is two-dimensional, the shape of the Voronoi region is a hexagon, as mentioned above.

[0062] However, in the case of integrating a function on a manifold, if the shape of the region indicating the integration region is hexagonal, it is not necessarily easy to execute the integration. Note that "easy to execute the integration" means that the amount of calculation required to obtain the integration result with a precision equal to or higher than a predetermined precision is small.

[0063] Furthermore, when a manifold has dimensions higher than two, integrating a function in a region obtained as a result of the Voronoi tessellation is even less easy than in the two-dimensional case. For example, in the case of four dimensions, it is often difficult to draw the shape of a Voronoi region in two-dimensional space, and it is not easy to integrate a function on such a manifold. Thus, it is not necessarily easy to integrate a function using a region obtained as a result of the Voronoi tessellation as an integral region.

[0064] Therefore, it may not be easy to obtain the parameterized cumulative distribution function used in the representative vector occurrence probability estimation process.

[0065] Therefore, in the representative vector occurrence probability estimation process, a parametrized cumulative distribution function obtained using the results of hypercuboid division is used instead of Voronoi division. Hypercuboid division is a process that divides a lattice space into lattice unit regions whose shape is a hypercuboid. Note that a two-dimensional hypercuboid is a rectangle, and a three-dimensional hypercuboid is a rectangular prism.

[0066] Fig. 6 is an explanatory diagram for explaining hyper-rectangle division in an embodiment. More specifically, Fig. 6 is a diagram showing an example of a result of hyper-rectangle division for a two-dimensional lattice space, together with an example of a result of Voronoi division. Both the "true region" and the "approximate region" in Fig. 6 are examples of lattice unit regions.

[0067] The "true region" is the region resulting from the Voronoi tessellation. That is, the "true region" is a Voronoi region. The "true region" is a hexagonal grid unit region. The "approximate region" is a hypercuboid grid unit region resulting from the hypercuboid tessellation.

[0068] The shape of the "approximation region" is a rectangular parallelepiped. Of the two types of division results shown in Fig. 6, one example of the result of the hyper rectangular parallelepiped division executed by the main data side probability estimation unit 109 is the result of division by the "approximation region". S1 in Fig. 6 is the length of a side of the rectangular parallelepiped, which is the shape of the "approximation region", and indicates the length of the side in the first dimension of the two-dimensional lattice space, and S2 indicates the length of the side in the second dimension.

[0069] The length of the first dimension side means the length of the rectangle in one oblique space when a rectangle in a lattice space is projected onto two orthogonal oblique spaces. The length of the second dimension side means the length of the rectangle in the other oblique space. In the following description, the length of the n-th dimension side of a hyper-rectangle in an N-dimensional lattice space means the length of the hyper-rectangle in the n-th oblique space when the hyper-rectangle is projected onto N one-dimensional oblique spaces that are orthogonal to each other. The hyper-rectangle in an oblique space is a straight line.

[0070] Incidentally, integrals with an integration domain of two or more dimensions can be obtained by iterative integration. When integrating a function with an N-dimensional hypercuboid as its integration domain, iterative integration is performed by integrating the function N times. In the case of a hypercuboid, it is possible to perform iterative integration with the integration domain of each of the N integrals being the edges of the hypercuboid.

[0071] When performing iterative integration with the integration domain of each of the N integrals being the edges of a hypercube, each integral is not affected by the results of the other integrals. On the other hand, in the case of a hyperpolyhedron that is not a hypercube, each integral of the iterative integration is affected by the results of the other integrals. Therefore, when the shape of the lattice unit domain is a hypercube, it is easier to perform the integration than when the shape of the lattice unit domain is a hyperpolyhedron that is not a hypercube.

[0072] In the example of Figure 6, the "true domain" is hexagonal in shape and the "approximate domain" is rectangular in shape, so performing integration over the "approximate domain" is easier than integration over the "true domain".

[0073] When obtaining a parameterized cumulative distribution function using the hyperrectangle obtained as a result of the hyperrectangle division as the integration domain, in each integral of the iterative integral, an integration of a one-dimensional probability density function that is not influenced by other dimensions is performed.

[0074] In the learning network 100, the probability density function that is the subject of the iterative integration is a function that indicates a predetermined type of distribution, and is a parametrized function. The predetermined type of distribution is a Gaussian distribution. The parameter values ​​of the probability density function are values ​​that correspond to the auxiliary data. The parametrized cumulative distribution function is obtained by performing iterative integration on the probability density function with the hyperrectangle obtained by the hyperrectangle division as the integration domain. In the learning network 100, the parameter values ​​of the probability density function are values ​​of the auxiliary data, so the parametrized cumulative distribution function obtained in this manner is a function that corresponds to the auxiliary data.

[0075] In the representative vector occurrence probability estimation process, the values ​​of the auxiliary data are substituted into the parameters of the cumulative distribution function thus obtained in advance, and the representative vector occurrence probability estimation process estimates the occurrence probability of the representative vector using the cumulative distribution function thus obtained.

[0076] In the representative vector occurrence probability estimation process, the occurrence probability of the representative vector is estimated by executing the processes shown in the following formulas (1) and (2). In formula (2), the cumulative distribution function cdf represents each cumulative distribution function obtained by each integral of the repeated integral. The result of integrating an m-dimensional (m is a natural number) cumulative distribution function for one dimension is an (m-1)-dimensional cumulative distribution function.

[0077] The right-hand side of equation (1) is an example of an occurrence probability obtained using a parameterized main data feature cumulative distribution function. Equation (1) is a function parameterized via the cumulative distribution function cfd of equation (2). Equation (2) is a cumulative distribution function obtained as a result of integrating the probability density function in the one-dimensional direction of a lattice unit area. Equation (1) is the product of the occurrence probabilities of each dimension obtained by equation (2). Therefore, equation (1) is an occurrence probability expressed using the result of integrating the probability density function over the entire lattice unit area.

[0078]

number

[0079]

number

[0080] i is an identifier that identifies each lattice point. j represents each dimension of the lattice space. The left side of equation (1) indicates the occurrence probability of the representative vector. Since each lattice point represents a representative vector, the occurrence probability p i is the occurrence probability p i It means. Δ j means the length of the dimension represented by the identifier j of the lattice unit area. Since the shape and size of the lattice unit area are predetermined, Δ j is a predetermined length.

[0081] In addition, y with a hat symbol in the following formula (3) means a noise-added main data feature amount. Hereinafter, the hat symbol A will be written as A^. Therefore, for example, the hat symbol y in the following formula (3) is y^.

[0082]

number

[0083] Equation (1) expresses the probability of occurrence p i is the probability obtained for each dimension p i If the shape of the lattice unit area is not a super-rectangular prism, the occurrence probability p i is the probability obtained for each dimension p i [j]. Therefore, since the shape of the lattice unit area is a super-rectangular parallelepiped, the main data side probability estimator 109 calculates the occurrence probability p i It is easy to obtain.

[0084] Note that the above formulas (1) and (2) are obtained under the assumption that the covariance between dimensions is 0. However, if the shape of the lattice unit area is a hypercube, the covariance between dimensions can be made 0 by rotating the coordinate axes of the lattice space so that they are parallel to each side of the hypercube. Therefore, formulas (1) and (2) are formulas that hold when the coordinate axes are rotated even if the covariance between dimensions is not 0. Note that, as is a well-known fact in linear algebra, the rotation of the coordinate axes is a unitary transformation, and therefore does not change the contents of formulas (1) and (2). Note that the process of making the covariance 0 is the process of diagonalizing the matrix that represents the variance between dimensions.

[0085] In this way, the main data side probability estimation process estimates the occurrence probability of each element of a tensor representing the main data features by using the result of integrating the parameterized probability density function with a lattice unit region whose shape is a hypercuboid as the integration region.

[0086] <Processing for determining the range of a hypercuboid> So far, we have explained the use of a hyper-rectangle as a lattice unit region, but have not touched upon the process of determining the shape and size of the hyper-rectangle. Here, we will explain an example of a process for determining the shape and size of a lattice unit region that at least satisfies the condition that it is a hyper-rectangle (hereinafter referred to as the "hyper-rectangle determination process").

[0087] In the hypercube determination process, first, the adjacent lattices of the origin lattice are calculated. The origin lattice is a lattice point located at the origin of the lattice space. The adjacent lattice is a lattice point next to the origin. Next, in the hypercube determination process, a hypercube that satisfies the hypercube determination conditions is calculated. The hypercube determination conditions include a condition that the lattice unit area of ​​the origin lattice does not overlap with the lattice unit areas of each adjacent lattice, and a condition that the volume of the hypercube is equal to the volume of the Voronoi region.

[0088] As a concrete example of the hypercuboid determination process, the conditions used to determine the shape and size of a hypercuboid in an 8-dimensional E8 lattice space are explained. There are two types of adjacent lattices of the origin lattice. One is [±1 2 , 0 6 ]. The other one is [±(1 / 2) 8 ], where the superscript indicates the number of dimensions. For example, [1 2 , 0 6 ] means [1, 1, 0, 0, 0, 0, 0, 0]. Note that the square brackets [] above represent a vector.

[0089] <Notation for indicating a hypercuboid> Hypercuboid [s1 a , s2 b s3 c The superscript in indicates that the length of each side from the 1st dimension to the ath dimension is s1, and the length of each side from the (a+1)th dimension to the (a+b)th dimension is s2. In this way, the superscript in the notation using superscript indicates that the length of the sides of the hypercuboid is the same length s for a series of dimensions that are the same as the number of superscripts, with s being the character to which the superscript is attached. In other words, s a indicates that the length of the side of the hypercuboid is s for a continuous number of dimensions. Note that information on the order of the dimensions is required to determine whether the dimensions are continuous or not, but the order of the dimensions is a predetermined order.

[0090] In addition, [±1 2 , 0 6 The "±" in [±1] indicates that there are cases of +1 and -1. 2 , 0 6 ] is specifically [1, 1, 0 6 ] and [-1, 1, 0 6 ] and [1, -1, 0 6 ] and [-1, -1, 0 6 ] means the four vectors.

[0091] Returning to the explanation of how to determine the shape and size of a hypercube in an eight-dimensional lattice space, there are two conditions for the hypercube of each adjacent lattice not to overlap with the hypercube of the origin lattice. One is that seven or more of the sides (elements) of the hypercube are 1 or less. The other is that one or more of the sides (elements) of the hypercube are 1 / 2 or less.

[0092] In addition, since the volume of the Voronoi region of the E8 lattice is 1, the hypercuboid satisfies the condition of the following equation (4).

[0093]

number

[0094] In the case of the 8-dimensional E8 lattice space, for example, s=[1 / 2, 1 6 , 2] is obtained.

[0095] Here are some examples for other dimensions. In the case of a two-dimensional A2 lattice space, the range of the lattice unit area is the range enclosed by a hypercuboid expressed by the following formula (5).

[0096]

number

[0097] In the case of an eight-dimensional E8 lattice space, the range of the lattice unit area is the range enclosed by a hypercuboid expressed by the following equation (6).

[0098]

number

[0099] In the case of a 24-dimensional Reach lattice space, the range of a lattice unit domain is the range enclosed by a hyper-rectangle expressed by the following equation (7).

[0100]

number

[0101] In this way, the range of the hyper-rectangle (the shape and size of the lattice unit area) is determined. Returning to the explanation of FIG.

[0102] The main entropy acquisition unit 110 acquires the entropy of the main data feature amount based on the result of estimation by the main data side probability estimation unit 109. Hereinafter, the entropy of the main data feature amount is referred to as the main entropy.

[0103] The main data side decoding unit 111 executes a main data feature decoding process. The main data feature decoding process is a process for processing information obtained based on the main data feature, and is a process for decoding the processing object. The processing object of the main data feature decoding process executed by the main data side decoding unit 111 is the noise-added main data feature. The contents of the decoding process of the main data side decoding unit 111 are updated by learning.

[0104] The reconstruction error calculation unit 112 calculates the difference between the decoding result of the main data side decoding unit 111 and the main data acquired by the main data acquisition unit 101. Hereinafter, the difference between the decoding result of the main data side decoding unit 111 and the main data acquired by the main data acquisition unit 101 is referred to as a reconstruction error. The difference between the decoding result of the main data side decoding unit 111 and the main data acquired by the main data acquisition unit 101 may be represented, for example, by the mean square error sum or by binary cross entropy.

[0105] The optimization unit 113 updates the learning network 100 based on the auxiliary entropy, the principal entropy, and the reconstruction error. The auxiliary entropy, the principal entropy, and the reconstruction error are all examples of the output of the learning network 100. Specifically, the optimization unit 113 updates the learning network 100 so as to reduce the optimization error, the principal entropy, and the auxiliary entropy. The objective function used by the optimization unit 113 is, for example, L=D+λ(R y +R z ).

[0106] The symbol L represents the objective function. The symbol D represents the reconstruction error. The symbol lambda is a predetermined constant. The symbol R y represents the auxiliary entropy. The symbol R z represents the principal entropy.

[0107] Since a small entropy means a short code length, the optimization unit 113 updates the learning network 100 to reduce the entropy. Also, since a smaller optimization error means higher accuracy of auto-encoding, the optimization unit 113 updates the learning network 100 to reduce the optimization error.

[0108] For example, the learning network 100 is updated by using a gradient method to solve the minimization problem of the objective function L. That is, the learning network 100 is updated by updating the values ​​of the parameters of the learning network 100 using, for example, the backpropagation method.

[0109] Specifically, updating the learning network 100 means updating the processing contents of the main data side encoding unit 102, the auxiliary data side encoding unit 103, the auxiliary data side probability estimating unit 106, the auxiliary data side decoding unit 107, and the main data side decoding unit 111.

[0110] <Relationship between autoencoding and learning network 100> The learning network 100 includes a main data-side noise adding unit 104. The main data-side noise adding unit 104 itself does not perform quantization processing. However, since the processing of the main data-side noise adding unit 104 is included, the performance of auto-encoding using vector quantization is improved by performing learning. The reason for this will be explained below.

[0111] For example, as described in Reference 1 below, when vector quantization is performed during learning, it is known that learning to improve the efficiency of auto-encoding processing using vector quantization cannot be performed because the gradient becomes 0. Therefore, in order to improve the efficiency of auto-encoding processing through learning, a process of adding noise to vectors is performed during learning instead of vector quantization itself.

[0112] By performing a process of adding noise instead of vector quantization itself, the performance of other processes that generate information used in vector quantization is improved, rather than the process of performing vector quantization itself. As a result, even if vector quantization is performed instead of a process of adding noise during auto-encoding, auto-encoding is performed more efficiently than before learning.

[0113] Reference 1: Balle, “Variational image compression with a scale hyperprior,” 2018

[0114] Regarding the learning unit 10, the main data side noise adding unit 104 adds noise to the noise-adding target vector. Therefore, by performing learning including the main data side noise adding unit 104, the gradient does not become 0. As a result, it is possible to update the contents of the main data side encoding unit 102, the auxiliary data side encoding unit 103, the auxiliary data side probability estimating unit 106, the auxiliary data side decoding unit 107, and the main data side decoding unit 111, which are also used in the self-encoding using vector quantization.

[0115] Therefore, what has been described so far as the network learning result specifically means the peripheral processing updated by the learning of the neural network (i.e., learning network 100) that adds noise instead of vector quantization. The peripheral processing is processing that generates information used in vector quantization. Specifically, the peripheral processing is processing that is executed by each of the main data side encoding unit 102, the auxiliary data side encoding unit 103, the auxiliary data side probability estimating unit 106, the auxiliary data side decoding unit 107, and the main data side decoding unit 111.

[0116] 7 is a diagram showing an example of a flow of processing executed by the learning unit 10 in the embodiment. N ] is obtained (step S101). The main data x is obtained from x1 to x N It is an N-dimensional vector with N elements from x1 to x N Each element from x1 to x is a tensor. N Each element up to may be a scalar or a vector.

[0117] Next, the main data encoding unit 102 executes a main data feature acquisition process (step S102). That is, the main data encoding unit 102 encodes the main data x. By encoding the main data, the main data feature y=f enc (x) is obtained. Function f enc (x) is a function (hereinafter referred to as the “main data encoding function”) that expresses the encoding process of the main data x. Note that the main data feature y is a tensor composed of k K-dimensional vectors.

[0118] Next, the auxiliary data side encoding unit 103 executes an auxiliary feature acquisition process (step S103). enc (y) is obtained. Function g enc (y) is a function (hereinafter referred to as the “main data feature encoding function”) expressing the encoding process of the main data feature y. The auxiliary feature z is a tensor such as a vector.

[0119] Next, the main data-side noise adding unit 104 executes a noise-added main data feature acquisition process (step S104). By executing the noise-added main data feature acquisition process, noise is added to the main data feature. That is, by executing the noise-added main data feature acquisition process, the noise-added main data feature y^=[y1^, y2^, . . . , y k ^] is obtained, where y i ^=y i +u y y i represents the i-th vector element of the main data feature y. y represents noise.

[0120] Next, the auxiliary data-side noise adding unit 105 executes a noise-added auxiliary feature acquisition process (step S105). By executing the noise-added auxiliary feature acquisition process, noise is added to the auxiliary feature. That is, by executing the noise-added auxiliary feature acquisition process, the noise-added auxiliary feature z^=[z1^, z2^, . . . , z w ^] is obtained. Note that w is an integer equal to or greater than 1. However, z i ^=z i +u z It is. i represents the i-th element of the auxiliary feature z. z represents noise.

[0121] Next, the auxiliary data side probability estimation unit 106 executes the auxiliary data side probability estimation process (step S106). The auxiliary data side probability estimation unit 106 estimates the auxiliary data side probability by executing the auxiliary data side probability estimation process. Specifically, the auxiliary data side probability is expressed by the following formula (8).

[0122]

number

[0123] The symbol on the left side of equation (8) represents the probability of the auxiliary data. Symbol h is the cumulative distribution function of the parametrized auxiliary feature.

[0124] Next, the auxiliary data side decoding unit 107 executes auxiliary feature decoding processing on the noise-added auxiliary feature (step S107). By executing the auxiliary feature decoding processing, the noise-added auxiliary feature is decoded. That is, by executing the auxiliary feature decoding processing, the auxiliary data θ=g dec (z^) is obtained. Function g dec (ẑ) is a function (hereinafter referred to as the “auxiliary feature decoding function”) that expresses the process of decoding the noise-added auxiliary feature ẑ. Note that the auxiliary data θ is a tensor such as a vector.

[0125] Next, the auxiliary entropy acquisition unit 108 acquires the auxiliary entropy based on the auxiliary data side probability (step S108). Specifically, the auxiliary entropy acquisition unit 108 acquires the auxiliary entropy by executing the process represented by the following formula (9).

[0126]

number

[0127] The symbol on the left side of equation (9) represents the auxiliary entropy.

[0128] Next, the main data side probability estimation unit 109 executes a main data side probability estimation process (step S109). The main data side probability estimation unit 109 executes the main data side probability estimation process to estimate the main data side probability based on the noise-added main data feature amount and the auxiliary data. The main data side probability is specifically expressed by the above-mentioned formula (1).

[0129] Next, the principal entropy acquisition unit 110 acquires the principal entropy based on the main data side probability (step S110). Specifically, the principal entropy acquisition unit 110 acquires the principal entropy by executing the process represented by the following formula (10).

[0130]

number

[0131] The symbol on the left side of equation (10) represents the principal entropy.

[0132] Next, the main data side decoding unit 111 executes a main data feature decoding process on the noise-added main data feature (step S111). The noise-added main data feature is decoded by executing the main data feature decoding process. Hereinafter, information obtained by decoded the noise-added main data feature is referred to as decoded main data. Therefore, the main data side decoding unit 111 executes the main data feature decoding process to decode the decoded main data x^=f dec (y^) is obtained. Function f dec (y^) is a function (hereinafter referred to as the “main data feature decoding function”) expressing the process of decoding the noise-added main data feature y^. Note that the decoded main data x^ is a tensor such as a vector.

[0133] Next, the reconstruction error calculation unit 112 acquires the difference between the decoded main data and the main data acquired by the main data acquisition unit 101 (step S112). The difference between the decoded main data and the main data acquired by the main data acquisition unit 101 is the reconstruction error.

[0134] Next, the optimization unit 113 updates the learning network 100 based on the auxiliary data entropy, the main data entropy, and the reconstruction error (step S113). Next, the optimization unit 113 determines whether a predetermined end condition for learning (hereinafter referred to as the "learning end condition") is satisfied (step S114). The learning end condition is, for example, a condition that the learning network 100 has been updated a predetermined number of times.

[0135] If the learning end condition is satisfied (step S114: YES), the process ends. On the other hand, if the learning end condition is not satisfied (step S114: NO), the process returns to step S101. The peripheral process at the time when the learning end condition is satisfied is used for vector quantization as the learned peripheral process.

[0136] The processes in steps S101 to S114 may be executed in any order as long as they do not violate the law of causality.

[0137] As described above, the auxiliary data-side probability estimation unit 106 may obtain a predetermined probability distribution by reading it from a predetermined storage device, etc. In such a case, the content of the auxiliary data-side probability estimation process is not updated by learning. Therefore, the learned auxiliary data-side probability estimation process is the same as the auxiliary data-side probability estimation process before learning.

[0138] More specifically, the update of the contents of the auxiliary data-side probability estimation process is an update of the parameterized auxiliary feature cumulative distribution function h. Therefore, when the auxiliary data-side probability estimation unit 106 acquires a pre-given probability distribution by, for example, reading it from a predetermined storage device, the learned parametrized auxiliary feature cumulative distribution function h is the same as the parametrized auxiliary feature cumulative distribution function h before learning.

[0139] In this way, the learning unit 10 updates the encoding and decoding processes in the process of self-encoding using vector quantization through learning. The process of self-encoding using vector quantization uses main data features, which are features to be self-encoded, and auxiliary features, which are features of the main data features. Furthermore, the process of self-encoding using vector quantization performs entropy coding on the result of vector quantizing the main data features, and entropy coding on the result of scalar quantizing the auxiliary features.

[0140] The encoding process in such self-encoding process using vector quantization specifically includes a main data feature acquisition process executed by the main data side encoding unit 102 and an auxiliary feature acquisition process executed by the auxiliary data side encoding unit 103. Moreover, the decoding process in such self-encoding process using vector quantization specifically includes an auxiliary feature decoding process executed by the auxiliary data side decoding unit 107 and a main data feature decoding process executed by the main data side decoding unit 111.

[0141] Auto-encoding using vector quantization using trained peripheral processing will be described with reference to Figs. 8 and 9. Trained means at the time when a learning end condition is satisfied. More specifically, an auto-encoding device 2 that executes encoding and decoding processes will be described as an example of a device that executes auto-encoding using vector quantization using trained peripheral processing. The auto-encoding device 2 is a type of auto-encoder (autoencoder).

[0142] Fig. 8 is a first explanatory diagram illustrating an overview of the self-encoding device 2 in the embodiment. Fig. 9 is a second explanatory diagram illustrating an overview of the self-encoding device 2 in the embodiment. More specifically, Fig. 8 is an explanatory diagram illustrating the encoding process executed by the self-encoding device 2, and Fig. 9 is an explanatory diagram illustrating the decoding process executed by the self-encoding device 2.

[0143] The auto-encoding device 2 is a type of auto-encoder, and thus includes an encoder and a decoder. Specifically, the auto-encoding device 2 includes an encoder 200 and a decoder 212. The encoder 200 includes an auto-encoding target acquisition unit 201, a learned main data side encoding unit 202, a learned auxiliary data side encoding unit 203, a vector quantization unit 204, a scalar quantization unit 205, a learned auxiliary data side probability estimation unit 206, a learned auxiliary data side decoding unit 207, an auxiliary entropy encoding unit 208, a main data side probability estimation unit 209, a main entropy encoding unit 210, and a data multiplexing unit 211.

[0144] The decoder 212 includes a coded data acquisition unit 213 , a data separation unit 214 , an auxiliary entropy decoding unit 215 , a learned auxiliary data side decoding unit 216 , a main entropy decoding unit 217 , and a learned main data side decoding unit 218 .

[0145] The auto-encoding target acquisition unit 201 acquires data to be subjected to auto-encoding as main data. Hereinafter, the target of auto-encoding will be referred to as the auto-encoding target.

[0146] The learned main data side encoding unit 202 executes a learned main data feature acquisition process on the auto-encoding target. By executing the learned main data feature acquisition process, the learned main data side encoding unit 202 acquires the main data feature of the auto-encoding target.

[0147] The learned auxiliary data side encoding unit 203 executes a learned auxiliary feature acquisition process on the main data feature to be auto-encoded. By executing the learned auxiliary feature acquisition process, the learned auxiliary data side encoding unit 203 acquires the auxiliary feature to be auto-encoded.

[0148] The vector quantization unit 204 executes a vector quantization process on the main data feature to be auto-encoded. By executing the vector quantization process, the vector quantization unit 204 acquires a vector quantized main data feature to be auto-encoded (hereinafter referred to as a "vector quantized feature").

[0149] The scalar quantization unit 205 executes a scalar quantization process on the auxiliary features to be auto-encoded. By executing the scalar quantization process, the scalar quantization unit 205 acquires scalar-quantized auxiliary features to be auto-encoded (hereinafter referred to as "scalar quantized features").

[0150] The learned auxiliary data side probability estimation unit 206 executes a learned auxiliary data side probability estimation process. The learned auxiliary data side probability estimation unit 206 executes the learned auxiliary data side probability estimation process to estimate the auxiliary data side probability of the auto-encoding target based on the scalar quantized feature.

[0151] The learned auxiliary data side decoding unit 207 executes a learned auxiliary feature decoding process on the scalar quantized feature. That is, the learned auxiliary data side decoding unit 207 decodes the scalar quantized feature. Hereinafter, information obtained by decoded the scalar quantized feature is referred to as quantized auxiliary data. Therefore, the learned auxiliary data side decoding unit 207 is a process for obtaining the quantized auxiliary data by decoding the scalar quantized feature.

[0152] The auxiliary entropy coding unit 208 performs entropy coding of the scalar quantized feature based on the scalar quantized feature and the auxiliary data side probability to be self-encoded. The entropy coding is, for example, arithmetic coding.

[0153] The main data side probability estimating unit 209 estimates the main data side probability of the auto-encoding target based on the vector quantized feature amount and the quantized auxiliary data.

[0154] The main entropy coding unit 210 performs entropy coding of the vector quantized feature quantity based on the vector quantized feature quantity and the main data side probability of the self-coding target. The entropy coding is, for example, arithmetic coding.

[0155] The data multiplexing unit 211 outputs the entropy-coded vector quantized feature amount and the entropy-coded scalar quantized feature amount to the decoder 212. In this manner, the encoder 200 encodes the object to be auto-encoded.

[0156] The encoded data acquisition unit 213 acquires the entropy-encoded vector quantized feature amount and the entropy-encoded scalar quantized feature amount.

[0157] The data separation unit 214 acquires the entropy-coded vector quantization feature and the entropy-coded scalar quantization feature acquired by the encoded data acquisition unit 213. The data separation unit 214 outputs the entropy-coded scalar quantization feature to the auxiliary entropy decoding unit 215, and outputs the entropy-coded vector quantization feature to the main entropy decoding unit 217.

[0158] The trained auxiliary entropy decoding unit 215 performs entropy decoding on the entropy-encoded scalar quantized feature using the trained parametrized auxiliary feature cumulative distribution function.

[0159] The learned auxiliary data side decoding unit 216 executes learned auxiliary feature decoding processing on the result of entropy decoding by the learned auxiliary entropy decoding unit 215 .

[0160] The main entropy decoding unit 217 performs entropy coding of the entropy coded vector quantization feature based on the entropy coded vector quantization feature and the result of the learned auxiliary feature decoding process by the learned auxiliary data side decoding unit 216. More specifically, the main entropy decoding unit 217 performs entropy decoding of the entropy coded vector quantization feature using a decoding cumulative distribution function. The decoding cumulative distribution function is a parameterized main data feature cumulative distribution function whose parameter values ​​are values ​​indicated by the result of the learned auxiliary feature decoding process by the learned auxiliary data side decoding unit 216.

[0161] The learned main data side decoding unit 218 executes a learned main data feature decoding process on the result of decoding by the main entropy decoding unit 217.

[0162] In this manner, the decoder 212 decodes the auto-encoding target encoded by the encoder 200. Also, in this manner, the auto-encoding device 2 performs auto-encoding of the auto-encoding target.

[0163] 10 is a flowchart showing an example of the flow of processing executed by the encoder 200 in the embodiment. N ] is obtained (step S201). The autoencoding target X is X1 to X N It is an N-dimensional vector with N elements from X1 to X N Each element from X1 to X is a tensor. N Each element up to may be a scalar or a vector.

[0164] Next, the learned main data side encoding unit 202 executes a learned main data feature acquisition process (step S202). That is, the learned main data side encoding unit 202 encodes the auto-encoding target X. By encoding the auto-encoding target, the main data feature Y of the auto-encoding target is obtained as Y=F. enc (Y) is obtained. Function F enc (X) is the trained main data encoding function. Note that the main data feature Y is a tensor such as a vector.

[0165] Next, the learned auxiliary data side encoding unit 203 executes a learned auxiliary feature acquisition process (step S203). By executing the learned auxiliary feature acquisition process, the auxiliary feature Z to be auto-encoded is obtained by enc (Y) is obtained. Function G enc (Y) is the trained main data feature encoding function. The auxiliary feature Z is a tensor such as a vector.

[0166] Next, the vector quantization unit 204 performs vector quantization on the main data feature quantity Y to be auto-encoded (step S204). By performing vector quantization, the vector quantized feature quantity Y^=[Y1^, Y2^, . . . , Y k ^]=[Q(Y1), Q(Y2),..., Q(Y k )] is obtained. Y i represents the i-th element of the main data feature Y to be auto-encoded. i represents the i-th element of the vector quantized feature Ŷ. Q is a function expressed by the following equation (11).

[0167]

number

[0168] The symbol Λ denotes the set of all lattice points.

[0169] Next, the scalar quantization unit 205 performs scalar quantization on the auxiliary feature Z to be auto-encoded (step S205). By executing the scalar quantization process, the scalar quantization unit 205 obtains the scalar quantized feature Z^=round(Z), where round represents a rounding process.

[0170] Next, the learned auxiliary data side probability estimation unit 206 executes the learned auxiliary data side probability estimation process (step S206). The learned auxiliary data side probability estimation unit 206 executes the learned auxiliary data side probability estimation process to estimate the auxiliary data side probability of the auto-encoding target based on the scalar quantized feature value Z^. Specifically, the auxiliary data side probability of the auto-encoding target is expressed by the following formula (12).

[0171]

number

[0172] The symbol on the left side of equation (12) represents the probability of the auxiliary data to be auto-encoded. The symbol H is the cumulative distribution function of the learned parametrized auxiliary features.

[0173] The learned auxiliary data side decoding unit 207 executes a learned auxiliary feature decoding process on the scalar quantized feature Z^ (step S207). That is, the learned auxiliary data side decoding unit 207 decodes the scalar quantized feature to obtain the quantized auxiliary data Θ=G dec (Z^) is obtained. Function G dec (Z^) is the learned auxiliary feature decoding function. The quantized auxiliary data Θ is a tensor such as a vector.

[0174] Next, the auxiliary entropy coding unit 208 performs entropy coding of the scalar quantized feature Z^ based on the scalar quantized feature Z^ and the auxiliary data side probability to be auto-encoded (step S208).

[0175] Next, the main data side probability estimating unit 209 estimates the main data side probability of the auto-encoding target based on the vector quantized feature value Y^ and the quantized auxiliary data Θ (step S209).

[0176] Next, the main entropy coding unit 210 performs entropy coding of the vector quantized feature value Y^ based on the vector quantized feature value Y^ and the main data side probability to be auto-coded (step S210).

[0177] Next, the data multiplexing unit 211 outputs the entropy-encoded vector quantized feature amount and the entropy-encoded scalar quantized feature amount to the decoder 212 (step S211).

[0178] The series of processes from step S201 to step S211 is an example of the encoding process by the encoder 200. Note that the processes from step S201 to step S211 may be executed in any order as long as they do not violate the law of causality.

[0179] 11 is a flowchart showing an example of the flow of processing executed by the decoder 212 in the embodiment. The encoded data acquisition unit 213 acquires the result of encoding by the encoder 200 (step S301). Specifically, the result of encoding by the encoder 200 is the entropy-encoded vector quantization feature and the entropy-encoded scalar quantization feature output in step S211.

[0180] Next, the data separation unit 214 separates the entropy-coded scalar quantization feature acquired in step S301 from the entropy-coded vector quantization feature acquired in step S302 (step S302). Specifically, separating means outputting the entropy-coded scalar quantization feature acquired in step S301 to the auxiliary entropy decoding unit 215, and outputting the entropy-coded vector quantization feature acquired in step S302 to the main entropy decoding unit 217.

[0181] Next, the trained auxiliary entropy decoding unit 215 performs entropy decoding on the entropy-encoded scalar quantized feature using the trained parametrized auxiliary feature cumulative distribution function (step S303).

[0182] Next, the learned auxiliary data side decoding unit 216 executes a learned auxiliary feature decoding process on the result of entropy decoding by the learned auxiliary entropy decoding unit 215 (step S304).

[0183] Next, the principal entropy decoding unit 217 performs entropy decoding of the entropy-coded vector quantization feature quantity using the decoded cumulative distribution function (step S305).

[0184] Next, the learned main data side decoding unit 218 executes a learned main data feature decoding process on the result of decoding by the main entropy decoding unit 217 (step S306).

[0185] A series of processes from step S301 to step S306 is an example of the decoding process by the decoder 212. Note that each process from step S301 to step S306 is executed after the encoding process such as step S211 is executed by the encoder 200, and may be executed in any order as long as it does not violate the law of causality.

[0186] <Hardware description> 12 is a diagram showing an example of a hardware configuration of a learning device 1 in an embodiment. The learning device 1 includes a control unit 11 including a processor 91 such as a CPU (Central Processing Unit) and a memory 92 connected by a bus, and executes a program. The learning device 1 functions as a device including the control unit 11, an input unit 12, a communication unit 13, a storage unit 14, and an output unit 15 by executing the program.

[0187] More specifically, the processor 91 reads out a program stored in the storage unit 14, and stores the read out program in the memory 92. The processor 91 executes the program stored in the memory 92, whereby the learning device 1 functions as a device including a control unit 11, an input unit 12, a communication unit 13, a storage unit 14, and an output unit 15.

[0188] The control unit 11 controls the operation of various functional units included in the learning device 1. The control unit 11 controls, for example, the operation of the output unit 15. The control unit 11 records, for example, various pieces of information generated by learning in the storage unit .

[0189] Input unit 12 includes input devices such as a mouse, a keyboard, a touch panel, etc. Input unit 12 may be configured as an interface that connects these input devices to study device 1. Input unit 12 accepts input of various types of information to study device 1.

[0190] The communication unit 13 includes a communication interface for connecting the learning device 1 to an external device. The communication unit 13 communicates with the external device via wired or wireless communication. The external device is, for example, a device that transmits main data used for learning. The communication unit 13 acquires the main data used for learning by communicating with the device that transmits the main data. The external device is, for example, the auto-encoding device 2. The communication unit 13 transmits the network learning result to the auto-encoding device 2 by communicating with the auto-encoding device 2. Note that the main data does not necessarily need to be input via the communication unit 13, and may be input to the input unit 12.

[0191] The storage unit 14 is configured using a computer-readable storage medium device such as a magnetic hard disk device or a semiconductor storage device. The storage unit 14 stores various information related to the learning device 1. The storage unit 14 stores information inputted via, for example, the input unit 12 or the communication unit 13. The storage unit 14 stores various information generated by, for example, the execution of learning.

[0192] The storage unit 14 stores in advance, for example, a probability distribution used to acquire the occurrence probability of each element of a tensor indicating an auxiliary feature. The storage unit 14 stores in advance, for example, a parametrized auxiliary feature cumulative distribution function. The storage unit 14 stores in advance, for example, a parametrized main data feature cumulative distribution function. The storage unit 14 stores in advance, for example, representative vector information. The storage unit 14 stores, for example, a result of hypercuboid division.

[0193] The storage unit 14 stores, in advance, for example, initial values ​​of the parameters of the learning network 100. The initial values ​​are, for example, random values. The storage unit 14 stores, for example, network learning results.

[0194] The output unit 15 outputs various information. The output unit 15 includes a display device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, or an organic EL (Electro-Luminescence) display. The output unit 15 may be configured as an interface that connects these display devices to the learning device 1. The output unit 15 outputs, for example, information input to the input unit 12. The output unit 15 may display, for example, the results of learning.

[0195] 13 is a diagram showing an example of the configuration of the control unit 11 included in the learning device 1 in the embodiment. The control unit 11 includes a learning unit 10, a memory control unit 120, a communication control unit 130, and an output control unit 140. The memory control unit 120 records various information in the memory unit 14. The communication control unit 130 controls the operation of the communication unit 13. The output control unit 140 controls the operation of the output unit 15.

[0196] 14 is a diagram showing an example of a hardware configuration of the autoencoding device 2 in the embodiment. The autoencoding device 2 includes a control unit 21 including a processor 93 such as a CPU (Central Processing Unit) and a memory 94 connected by a bus, and executes a program. The autoencoding device 2 functions as a device including the control unit 21, an input unit 22, a communication unit 23, a storage unit 24, and an output unit 25 by executing the program.

[0197] More specifically, the processor 93 reads out the program stored in the storage unit 24 and stores the read out program in the memory 94. The processor 93 executes the program stored in the memory 94, whereby the autoencoding device 2 functions as a device including a control unit 21, an input unit 22, a communication unit 23, a storage unit 24, and an output unit 25.

[0198] The control unit 21 controls the operation of various functional units included in the self-encoding device 2. The control unit 21 controls, for example, the operation of the output unit 25. The control unit 21 records, in the storage unit 24, various pieces of information generated by, for example, encoding by the encoder 200 and decoding by the decoder 212.

[0199] The input unit 22 includes input devices such as a mouse, a keyboard, and a touch panel. The input unit 22 may be configured as an interface that connects these input devices to the auto-encoding device 2. The input unit 22 accepts input of various information to the auto-encoding device 2.

[0200] The communication unit 23 includes a communication interface for connecting the auto-encoding device 2 to an external device. The communication unit 23 communicates with the external device via wired or wireless communication. The external device is, for example, a device that transmits the auto-encoding target. The communication unit 23 acquires the auto-encoding target by communicating with the device that transmits the auto-encoding target. The external device is, for example, the learning device 1. The communication unit 23 receives the network learning result by communicating with the learning device 1. Note that the auto-encoding target does not necessarily need to be input via the communication unit 23, and may be input to the input unit 22.

[0201] The storage unit 24 is configured using a computer-readable storage medium device such as a magnetic hard disk device or a semiconductor storage device. The storage unit 24 stores various information related to the self-encoding device 2. The storage unit 24 stores information inputted via, for example, the input unit 22 or the communication unit 23. The storage unit 24 stores various information generated by, for example, the execution of encoding by the encoder 200 and decoding by the decoder 212.

[0202] The storage unit 24 stores, for example, a network learning result. The storage unit 24 stores, for example, representative vector information in advance. The storage unit 24 stores, for example, a result of hyper-rectangle division.

[0203] The output unit 25 outputs various information. The output unit 25 includes a display device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, or an organic EL (Electro-Luminescence) display. The output unit 25 may be configured as an interface that connects these display devices to the learning device 1. The output unit 25 outputs, for example, information input to the input unit 22. The output unit 25 may output, for example, the result of auto-encoding of the auto-encoding target.

[0204] 15 is a diagram showing an example of the configuration of the control unit 21 included in the self-encoding device 2 in the embodiment. The control unit 21 includes an self-encoding execution unit 20, a storage control unit 220, a communication control unit 230, and an output control unit 240. The self-encoding execution unit 20 performs self-encoding on the self-encoding target. The self-encoding execution unit 20 includes an encoder 200 and a decoder 212. The self-encoding execution unit 20 performs encoding by the encoder 200 and decoding by the decoder 212, thereby performing self-encoding of the self-encoding target.

[0205] The memory control unit 220 records various information in the memory unit 24. The communication control unit 230 controls the operation of the communication unit 23. The output control unit 240 controls the operation of the output unit 25.

[0206] The learning device 1 configured in this way learns peripheral processing for vector quantization using representative vector information, which is information indicating the positions of representative vectors arranged in a lattice pattern in a vector space such as LatticeVQ. The learning device 1 then estimates the occurrence probability of the representative vector using the result of the hypercuboid division. As described above, when using the representative vector, the occurrence probability of the representative vector may be estimated using Voronoi division, but integration is not easy.

[0207] Therefore, the learning device 1 that estimates the occurrence probability of a representative vector using a hypercuboid obtained by hypercuboid division can reduce the burden required to obtain the process of auto-encoding using vector quantization. This reduces the burden from the learning stage required to realize auto-encoding using vector quantization. Therefore, the learning device 1 can reduce the burden required for auto-encoding using vector quantization.

[0208] Furthermore, the learning device 1 configured in this manner uses the representative vector to learn peripheral processing for vector quantization, and therefore the burden required for learning the representative vector can be reduced. Furthermore, the learning device 1 configured in this manner uses the representative vector to learn peripheral processing for vector quantization, and therefore does not use memory that is required when learning the representative vector. Therefore, the learning device 1 can reduce the frequency with which a memory shortage problem occurs, and can process main data of larger dimensions.

[0209] As mentioned above, in learning vector quantization, it is necessary to add noise instead of quantization. However, when using representative vector information such as LatticeVQ, the range of the Voronoi region becomes a quantization error, so it is not easy to generate noise whose probability distribution is Gaussian.

[0210] Meanwhile, the learning device 1 configured in this way uses, in learning, only samples within the Voronoi region as noise among samples randomly generated within a (K-1)-dimensional sphere circumscribing the Voronoi region in a K-dimensional vector space. This makes it possible to generate noise that follows a Gaussian distribution. Therefore, the learning device 1 can reduce the burden required for obtaining the process of auto-encoding using vector quantization. Therefore, the learning device 1 can reduce the burden required for auto-encoding using vector quantization.

[0211] The auto-encoding device 2 configured in this manner performs auto-encoding using vector quantization using the results of learning by the learning device 1. This makes it possible to reduce the burden required for auto-encoding using vector quantization.

[0212] (Modification) In adding noise to the main data feature quantity, samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space may be used as noise points. In adding noise to the main data feature quantity, samples uniformly generated within a hypercuboid obtained by hypercuboid division may be used as noise points.

[0213] Therefore, the process of adding noise to the main data feature (i.e., vector noise adding process) may be any one of the first noise adding process, the second noise adding process, and the third noise adding process. The first noise adding process is a process of adding, as noise, samples within a Voronoi region among samples randomly generated within a (K-1)-dimensional sphere circumscribing a Voronoi region in a K-dimensional vector space. In other words, the first noise adding process is the process described with reference to Figs. 3 and 4.

[0214] The second noise adding process adds, as noise, samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space. The third noise adding process adds, as noise, samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid.

[0215] The learning device 1 may be implemented using a plurality of information processing devices communicably connected via a network. In this case, each functional unit of the learning device 1 may be distributed and implemented in the plurality of information processing devices.

[0216] The self-encoding device 2 may be implemented using a plurality of information processing devices communicably connected via a network. In this case, each functional unit of the self-encoding device 2 may be distributed and implemented in the plurality of information processing devices.

[0217] All or part of the functions of the learning device 1 and the autoencoding device 2 may be realized using hardware such as an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). The program may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into computer systems. The program may be transmitted via a telecommunications line.

[0218] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and designs that do not deviate from the gist of the present invention are also included. [Explanation of symbols]

[0219] 1...Learning device, 10...Learning unit, 100...Learning network, 101...Main data acquisition unit, 102...Main data side encoding unit, 103...Auxiliary data side encoding unit, 104...Main data side noise adding unit, 105...Auxiliary data side noise adding unit, 106...Auxiliary data side probability estimating unit, 107...Auxiliary data side decoding unit, 108...Auxiliary entropy acquisition unit, 109...Main data side probability estimating unit, 110...Main entropy acquisition unit, 111...Main data side decoding unit, 112...Reconstruction error calculation unit, 113...Optimization unit, 2...Auto-encoding device, 200...Encoder, 201...Auto-encoding target acquisition unit, 202...Learned main data side encoding unit, 203...Learned auxiliary data side encoding unit, 204...Vector quantization unit, 205...Scalar quantization unit, 206...learned auxiliary data side probability estimation unit, 207...learned auxiliary data side decoding unit, 208...auxiliary entropy coding unit, 209...main data side probability estimation unit, 210...main entropy coding unit, 211...data multiplexing unit, 213...encoded data acquisition unit, 214...data separation unit, 215...auxiliary entropy decoding unit, 216...learned auxiliary data side decoding unit, 217...main entropy decoding unit, 218...learned main data side decoding unit, 11...control unit, 12...input unit, 13...communication unit, 14...storage unit, 15...output unit, 120...storage control unit, 130...communication control unit, 140...output control unit, 21...control unit, 22...input unit, 23...communication unit, 24...storage unit, 25...output unit, 20...self-encoding execution unit, 220: memory control unit, 230: communication control unit, 240: output control unit, 91: processor, 92: memory, 93: processor, 94: memory< / latticevq>

Claims

1. a learning unit that updates, by learning, encoding and decoding processes in the auto-encoding process using vector quantization, the auto-encoding process using main data features that are features to be auto-encoded and auxiliary features that are features of the main data features, the auto-encoding process performing vector quantization of the main data features and scalar quantization of the auxiliary features; Equipped with the learning unit executes a process of adding noise to the main data feature quantity in the learning; The noise adding process includes: A process of adding, as noise, samples within a Voronoi region among samples randomly generated within a (K-1)-dimensional sphere circumscribing the Voronoi region in a K-dimensional (K is an integer equal to or greater than 2) vector space; A process of adding samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space as noise; A process of adding samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, that includes one lattice point of the vector space, and that has a shape of a hypercuboid, as noise; Either one of the following: Learning device.

2. The vector quantization is LatticeVQ. The learning device according to claim 1 .

3. an auto-encoding target acquisition unit that acquires an auto-encoding target; The present invention relates to an auto-encoding process using vector quantization, the auto-encoding process using main data features which are features to be auto-encoded, and auxiliary features which are features of the main data features, the auto-encoding process performing vector quantization of the main data features and scalar quantization of the auxiliary features, and a learning unit which updates encoding and decoding processes in the auto-encoding process by learning, the learning unit executing a process of adding noise to the main data features in the learning, the process of adding noise being performed by randomly generating samples within a (K-1)-dimensional sphere circumscribing a Voronoi region in a K-dimensional (K is an integer of 2 or more) vector space, the process of adding noise being performed by applying a noise-applying process to the Voronoi region among samples randomly generated within a (K-1)-dimensional sphere circumscribing a Voronoi region in the Voronoi region. a process of adding, as noise, samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space; and a process of adding, as noise, samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid; and an auto-encoding execution unit that performs auto-encoding by vector quantization of the target acquired by the auto-encoding target acquisition unit using a trained encoding process and a trained decoding process obtained using a learning device, the trained encoding process and the trained decoding process being one of the following processes: a process of adding, as noise, samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid; A self-encoding device comprising:

4. a learning step of updating, by learning, encoding and decoding processes in an auto-encoding process using vector quantization, the auto-encoding process using main data features which are features to be auto-encoded and auxiliary features which are features of the main data features, the auto-encoding process performing vector quantization of the main data features and scalar quantization of the auxiliary features; having The learning step includes a process of adding noise to the main data feature quantity in the learning step, The noise adding process includes: A process of adding, as noise, samples within a Voronoi region among samples randomly generated within a (K-1)-dimensional sphere circumscribing the Voronoi region in a K-dimensional (K is an integer equal to or greater than 2) vector space; A process of adding samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space as noise; A process of adding samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, that includes one lattice point of the vector space, and that has a shape of a hypercuboid, as noise; Either one of the following: How to learn.

5. An auto-encoding target acquisition step of acquiring an auto-encoding target; The present invention relates to an auto-encoding process using vector quantization, the auto-encoding process using main data features which are features to be auto-encoded, and auxiliary features which are features of the main data features, the auto-encoding process performing vector quantization of the main data features and scalar quantization of the auxiliary features. The present invention further relates to an auto-encoding process using vector quantization, the auto-encoding process performing vector quantization of the main data features and scalar quantization of the auxiliary features. The present invention relates to an auto-encoding process using vector quantization, the auto-encoding process using vector quantization of the main data features and scalar quantization of the auxiliary features. The present invention relates to an auto-encoding process using vector quantization, a process of adding, as noise, samples uniformly generated within a (K-1)-dimensional sphere whose volume is approximated to the volume of a Voronoi region in a K-dimensional vector space; and a process of adding, as noise, samples uniformly generated within a region that divides a vector space in which representative vectors are arranged in a lattice pattern, includes one lattice point of the vector space, and has a shape of a hypercuboid. An auto-encoding execution step of performing auto-encoding by vector quantization of the target acquired in the auto-encoding target acquisition step using a trained encoding process and a trained decoding process obtained using a learning device, the trained encoding process and the trained decoding process being one of the following processes; The self-encoding method has the following structure:

6. A program for causing a computer to function as the learning device according to claim 1 or 2.

7. A program for causing a computer to function as the self-encoding device according to claim 3.