CN111582434A

CN111582434A - Deep learning system

Info

Publication number: CN111582434A
Application number: CN201910404061.7A
Authority: CN
Inventors: 郑载训
Original assignee: Korea Treasure Platinum Co ltd
Current assignee: Korea Treasure Platinum Co ltd
Priority date: 2019-02-18
Filing date: 2019-05-15
Publication date: 2020-08-25
Anticipated expiration: 2039-05-15
Also published as: WO2020171321A1; CN111582434B; KR102391817B1; KR20200100388A

Abstract

A deep learning system is disclosed that is accessible over a communication network to an external deep learning system for learning a first artificial neural network model. The deep learning system includes: a data classification storage module configured to receive and store source learning data; a learning module configured to use the source learning data to cause a second artificial neural network model in the system to learn; an adjustment information transmission module configured to transmit adjustment information including learning-related information and parameter information of a second artificial neural network model to the external deep learning system to request adjustment of the first artificial neural network model, wherein the learning-related information and the parameter information are modified according to a result of the learning; and an adjustment module configured to update parameter information of the second artificial neural network model according to adjustment information transmitted from an external deep learning system.

Description

Deep learning system

Technical Field

The invention relates to an artificial intelligence learning system, in particular to a learning system based on deep learning.

Background

The deep learning algorithm is recently attracting attention as a machine learning system, which is also an artificial neural network and the number of hidden layers is much larger than that of the traditional neural network.

Generally, an artificial neural network is composed of three layers in total, i.e., an input layer for receiving input, a hidden layer for performing actual learning, and an output layer for returning a calculation result, as shown in fig. 1. In an artificial neural network having such a structure, each node means one neuron. Each node finds the weighted sum when data is input, applies the activation function, and transmits the calculation result to the next node. In this case, a weight is assigned to each node by learning. The results of the previous node (neuron) are weighted and input to the current node (neuron), and a bias is added to the weighted results.

In general, the deep learning process is performed by repeatedly performing a forward propagation process for calculating deep learning parameters and an objective function through a plurality of hidden layers in order from an input layer to an output layer and a backward propagation process for modifying weights through the hidden layers in order from the output layer to the input layer in consideration of errors. The weights modified during the deep learning process are repeatedly updated until the error is minimized, and all computers share the weights and deep learning parameters.

In other words, since the learning process using the learning data is performed in the deep learning algorithm to generate the artificial neural network model having the optimized parameters, as the number of pieces of the learning data increases, the artificial neural network model capable of more accurate prediction is formed. Thus, the deep learning process requires a large number of pieces of learning data.

However, in the case of a medical diagnostic imaging apparatus for determining whether a disease such as cervical cancer has occurred using a deep learning algorithm, there is a limitation in ensuring learning data due to a limitation in disclosure of personal information or medical information, and it is difficult to form an artificial neural network model that can accurately predict. Particularly in the case of rare diseases, the number of patients is small, it is difficult to share learning data between korea and foreign medical institutions, and it takes a long time to accumulate the learning data. Therefore, there is a need to find a new method for effectively solving these problems.

Further, in order to make a more accurate diagnosis in the medical diagnostic imaging apparatus, it is necessary to accurately classify learning data to be used for learning in addition to acquiring source learning data. That is, unless classification of learning data is accurately and clearly performed, accuracy in analyzing occurrence of a specific disease is inevitably lowered. Therefore, for the apparatus for determining whether a specific disease has occurred, it is necessary to share the learning data classification criteria and to share the learning data to be used for learning.

(prior art document)

(patent document)

Korean patent laid-open publication No. 10-2017-0083419

Disclosure of Invention

Technical problem

Therefore, the present invention has been devised in light of the above necessity, and a primary object of the present invention is to provide a deep learning system capable of changing an artificial neural network model into a model capable of making an accurate prediction like an artificial neural network model in which a large amount of source data is learned, in an environment in which only a small amount of source data can be learned.

Further, it is another object of the present invention to provide a deep learning system capable of reducing a learning time and overcoming an environment of learning data shortage by acquiring a result of learning by other artificial neural network models and reflecting the result in its own artificial neural network model.

Further, it is still another object of the present invention to provide a deep learning system that can obtain the same learning result as that obtained by sharing source learning data without sharing the source learning data between artificial neural network models operated by different operators.

Further, it is still another object of the present invention to provide a deep learning system capable of accurately determining whether a specific disease has occurred by sharing learning data classification criteria while acquiring learning results obtained by other artificial neural network models and reflecting them in its own artificial neural network model.

Technical scheme

In order to achieve the above object, according to an embodiment of the present invention, there is disclosed a deep learning system accessible through a communication network to an external deep learning system for learning a first artificial neural network model, the deep learning system including: a data classification storage module configured to receive and store source learning data; a learning module configured to use the source learning data to cause a second artificial neural network model in the system to learn; an adjustment information transmission module configured to transmit adjustment information including learning-related information and parameter information of a second artificial neural network model to the external deep learning system to request adjustment of the first artificial neural network model, wherein the learning-related information and the parameter information are modified according to a result of the learning; and an adjustment module configured to update parameter information of the second artificial neural network model according to adjustment information transmitted from an external deep learning system.

The deep learning system may also include a diagnosis module configured to diagnose whether a particular disease has occurred based on the first artificial neural network model optimized by repeated operations of the learning module and the adjustment module.

The data classification storage module may include: a data receiving unit configured to receive image data of an inspection target as source learning data from an external apparatus and store the received image data; and a data classification unit configured to classify and store the stored image data of the inspection target according to a plurality of multi-level classification criteria, and the data classification unit classifies and stores the image data of the inspection target using at least two or more of a first-level classification criterion which is a classification criterion of a color, a second-level classification criterion which is a classification criterion of a size of the inspection target, and a third-level classification criterion which is a classification criterion of a combination of a color and a form.

The adjustment information transmission module may preferentially transmit, as the adjustment information, a hyper-parameter required to perform synchronization on an initial learning state of the first artificial neural network model.

In some cases, the adjustment information transmitted by the adjustment information transmission module or the external deep learning system may constitute an artificial neural network model and include a weight value and a bias value of each layer modified according to a learning result.

The learning related information may include the number of learned pieces of source learning data in the source learning data to be learned.

The data classification unit also generates additional source learning data to adjust a balance of the number of source learning data classified and stored based on the classification criteria.

Advantageous effects

According to the above-described technical solution, since the parameters of the artificial neural network model obtained by learning by the first deep learning system are transmitted to the second deep learning system and used to update the parameters of the artificial neural network model of the second deep learning system, the same effect as that obtained by learning the same learning data can be obtained without sharing the source learning data. Accordingly, the present invention can overcome problems caused by exposure of personal information or medical information in advance, and can also shorten the time required to optimize an artificial neural network model due to limitations in sharing personal information or medical information.

Furthermore, the present invention is advantageous in that, since the artificial neural network model is partially self-optimized using parameters obtained from learning results generated from an external system, it is practically impossible to recover or infer source learning data, thereby minimizing damage caused by information exposure.

In particular, the present invention can achieve the same effect as that obtained by learning the same learning data without sharing the source learning data, and therefore, even if the artificial neural network model is in an environment where only a small amount of source data can be learned, the artificial neural network model can be changed to an artificial neural network model that learns a large amount of source learning data.

Also, according to still another embodiment, the present invention is advantageous in that since learning data classification criteria and parameters modified according to learning results are shared among different deep learning systems while accurately classifying learning data for learning in advance and learning on the basis of the classification criteria, diagnostic criteria can be shared to accurately diagnose whether a specific disease has occurred.

Drawings

Fig. 1 is a diagram showing a configuration of a general artificial neural network.

Fig. 2 is a diagram illustrating connections between deep learning systems according to an embodiment of the present invention.

Fig. 3 is a diagram showing the configuration of a deep learning system according to an embodiment of the present invention.

Fig. 4 is a diagram for further describing a deep learning process according to an embodiment of the present invention.

Fig. 5 is a diagram illustrating a multi-level classification flow of image data of an inspection target according to another embodiment of the present invention.

Fig. 6 is a diagram for describing in detail a multi-level classification criterion of image data of an inspection target according to still another embodiment of the present invention.

Detailed Description

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, if it is determined that related well-known functions or elements may unnecessarily obscure the gist of the present invention, a detailed description thereof will be omitted.

For reference, the term "module" used herein is defined as a functional and structural combination of hardware for implementing the technical spirit of the present invention and software required for driving the hardware. That is, a "module" refers to a logical unit of predetermined code and hardware resources for executing the code, and it does not necessarily mean physically connected code or one type of hardware, as is apparent to those skilled in the art.

Hereinafter, a deep learning system according to an embodiment of the present invention will be described with reference to the accompanying drawings.

First, fig. 2 illustrates the connections between the deep learning systems 100, 200, and 300 according to an embodiment of the present invention.

As shown in fig. 2, the deep learning systems 100, 200 and 300 according to the embodiment of the present invention transmit and receive required information to and from each other through a communication network. It is assumed that each of the deep learning systems 100, 200, and 300 may include the same internal artificial neural network model.

Each of the deep learning systems 100, 200, and 300 transmits and receives information to and from a deep learning system designated by an administrator, and updates its own parameters using parameters obtained by learning by other deep learning systems to optimize an internal artificial neural network model.

The configuration of the deep learning systems 100, 200, and 300 shown in fig. 2 will be described below with reference to fig. 3.

FIG. 3 shows a block diagram of a deep learning system 100 according to an embodiment of the invention. As an example, the deep learning system 100 including the Convolutional Neural Network (CNN) -based artificial neural network model 150 includes: a data classification storage module 110 configured to receive and store source learning data in a memory; a learning module 120 configured to train an artificial neural network model 150 in the system using the source learning data; an adjustment information transmission module 130 configured to transmit adjustment information including learning-related information and parameter information of the artificial neural network model 150 modified with the learning result to the external deep learning system 200 or 300 to request adjustment of the artificial neural network model in the system; and an adjustment module 140 configured to update parameter information of the artificial neural network model 150 according to adjustment information transmitted from the external deep learning system 200 or 300.

According to an embodiment, in addition to the above elements, the deep learning system 100 according to an embodiment of the present invention may further include a diagnosis module 160 configured to diagnose whether a specific disease has occurred based on the artificial neural network model 150 optimized by the repeated operations of the learning module 120 and the adjustment module 140.

The source learning data is data transmitted through a separate storage device or a separate data providing server in which the source learning data is stored, and through an image capturing device connectable to the deep learning system 100. When the deep learning system 100 according to an embodiment of the present invention is used as part of a medical diagnostic imaging apparatus, the source learning data may be image data of an object to be examined (hereinafter, referred to as an examination target) captured for diagnosing a specific disease.

Meanwhile, the adjustment information transmission module 130 preferentially transmits, as the adjustment information, the hyper-parameters required to perform synchronization on the initial learning state of the artificial neural network model included in the external deep learning system 200 or 300, and then transmits, as the adjustment information, the weight values W1, W2, W3. of each layer, which are modified according to the learning result and the bias value. The adjustment information transmission module 130 is a control module for controlling the overall operation of the deep learning system 100, and can perform control of authentication and interoperation with another system.

The above learning related information may be the number of pieces of learned source learning data in the source learning data to be learned, the ratio of the learned data to the data to be learned, or the number of pieces of learned source learning data and the number of times learning is performed. It is important to have the artificial neural network model of the partner-side deep learning system 200 or 300 learn in the same manner as learning in one deep learning system 100.

The deep learning systems 100, 200 and 300 according to another embodiment of the present invention can be re-implemented by modifying the data classification storage module 110 among the above-described elements. That is, the data classification storage module 110 may include: a data receiving unit configured to receive image data of an inspection target as source learning data from an external apparatus and store the received image data in a memory; and a data classification unit configured to classify and store the stored image data of the inspection target according to a plurality of multi-level classification criteria.

In this case, the data classifying unit may classify and store the image data of the inspection target using at least two or more of a first-level classification criterion which is a classification criterion of color, a second-level classification criterion which is a classification criterion of the size of the inspection target, and a third-level classification criterion which is a classification criterion of a combination of color and shape.

The first-stage classification criterion of the data classification unit includes, as a classification criterion value, a color value for identifying one or more of an acetic acid reaction image, a Lugol solution reaction image, a green filter image, and a normal image.

When the data classification storage module 110 is a system capable of classifying according to data classification criteria, the adjustment information transmission module 130 preferably includes at least two or more of the first-level classification criteria, the second-level classification criteria, and the third-level classification criteria in the adjustment information and transmits the adjustment information to the external deep learning system 200 or 300 to synchronize the interoperable systems.

While the operation of the deep learning system 100, 200 or 300 including the above elements will be further described, a process in which the first deep learning system 100 and the second deep learning system 200 interoperate with each other to optimize the artificial neural network model will be described. For convenience of description, the artificial neural network model included in the first deep learning system 100 will be referred to as a second artificial neural network model, and the artificial neural network model included in the second deep learning system 200 will be referred to as a first artificial neural network model.

Fig. 4 is a diagram for additionally describing a deep learning process according to an embodiment of the present invention.

In fig. 4, it is assumed that the second deep learning system 200 has a smaller number of pieces of source learning data than the first deep learning system 100 and has a different operating system. Further, it is assumed that each of the first deep learning system 100 and the second deep learning system 200 has a storage unit in which source learning data to be learned is stored.

Under these assumptions, the system administrator accesses the first deep learning system 100 and sets up a deep learning system to be interoperated (S10). The setting may be performed by inputting the connection information and the authentication information of the second deep learning system 200.

With this arrangement, the adjustment information transmission module 130 operable as a control module of the first deep learning system 100 can access the second deep learning system 200 and perform the mutual system authentication process (S20).

When mutual system authentication is normally completed, the first deep learning system 100 and the second deep learning system 200 can transmit and receive required information. In this case, the adjustment information transmission module 130 sets the hyper-parameter as the adjustment information to perform synchronization on the initial learning state of the second artificial neural network model of the second deep learning system 200 and preferentially transmits the adjustment information to the second deep learning system 200 (S30). When receiving the hyper-parameters as the adjustment information, the adjustment module of the second deep learning system 200 updates the parameters associated with the first artificial neural network model to perform synchronization with the second artificial neural network model.

When the initial learning states of the two deep learning systems are synchronized with each other through the transmission and reception of the hyper-parameters, the learning module 120 of the first deep learning system 100 trains the second artificial neural network model 150 in the system using the source learning data stored in the memory (S40). The weight values W1, W2, W3. of the layers constituting the second artificial neural network model 150 are modified by learning.

When the learning module 120 learns a predetermined number of pieces of source learning data, the adjustment information transmission module 130 transmits adjustment information including learning-related information (the number of learned pieces of data in the source learning data) and parameter information (weight values and bias values) of the second artificial neural network model modified according to the learning result to the second deep learning system 200 to request adjustment of the first artificial neural network model (S50).

The adjustment model included in the second deep learning system 200 updates parameter information (i.e., weight values) of the first artificial neural network model in the second deep learning system 200 according to the adjustment information transmitted from the external first deep learning system 100 (S60). The second deep learning system 200 can also obtain the effect of learning the source learning data of the first deep learning system 100 by updating the weight values of the first artificial neural network model included in the external second deep learning system 200 using the weight values modified by the first deep learning system 100 during one learning process.

After updating the parameters of the first artificial neural network model using the adjustment information transmitted from the outside, the learning module of the second deep learning system 200 also learns its own source learning data (S70), and performs the learning according to the learning-related information transmitted from the first deep learning system 100.

When learning is completed based on the learning-related information, parameters of the first artificial neural network model are also modified. By transmitting the adjustment information including the modified weight values and the bias values to the first deep learning system 100 by the adjustment information transmission module 130 included in the second deep learning system 200 (S80), the adjustment module 140 included in the first deep learning system 100 updates the parameter information of the second artificial neural network model 150 based on the received adjustment information (S90). Therefore, the first deep learning system 100 can also obtain the effect of learning the source data of the second deep learning system 200.

Meanwhile, the first learning system 100 and the second deep learning system 200 repeat the above adjustment process until a specific performance is obtained, that is, until the artificial neural network model is optimized (S100), thereby optimizing the first artificial neural network model or the second artificial neural network model.

As described above, by transmitting the parameters obtained by learning its own source learning data to the partner-side deep learning system to update the parameters of the artificial neural network model, the same effect as that obtained by learning the same learning data can be obtained without sharing the source learning data. Therefore, advantageously, it is possible to solve problems caused by exposure of personal information or medical information in advance, and it is also possible to shorten the time required to optimize the artificial neural network model due to the limitation on sharing personal information or medical information.

Furthermore, the present invention is advantageous in that since the artificial neural network model is partially optimized using parameters obtained from the learning results of the external system, it is practically impossible to recover or infer the source learning data, thereby minimizing the damage caused by the information exposure.

In the above embodiments, the process has been described in which the two deep learning systems 100 and 200 interoperate with each other to optimize the artificial neural network model without sharing the source learning data.

A deep learning system according to still another embodiment configured to perform learning to optimize an artificial neural network model after accurately classifying learning data for learning to improve accuracy of analyzing occurrence of a specific disease, while two deep learning systems 100 and 200 interoperate with each other to optimize the artificial neural network model, will be described below. The deep learning system may be implemented by dividing the data classification storage module 110 shown in fig. 3 into a data receiving unit and a data classification unit. In the following embodiments, data classification criteria required for accurately diagnosing whether or not cervical cancer has occurred will be described as examples.

First, the data receiving unit is used to receive image data of an inspection target corresponding to source learning data from an external apparatus such as an image capturing apparatus and store the received image data, and the data classifying unit is used to classify and store the image data of the inspection target according to a plurality of multi-level classification criteria. In this case, the data classifying unit may classify the image data of the inspection target using at least two or more of a first-level classification criterion which is a classification criterion of color, a second-level classification criterion which is a classification criterion of the size of the inspection target, a third-level classification criterion which is a classification criterion of a combination of color and shape, and a fourth-level classification criterion which is a classification criterion of exposure and focus.

In some cases, the data classifying unit may classify the image data of the examination target first according to a first-level classification criterion, which is a classification criterion of color, secondly, classify the image data of the examination target classified by the first level according to a second-level classification criterion, which is a classification criterion of the size of the examination target, for example, the size of cervix uteri, and thirdly, classify the image data of the examination target according to a third-level classification criterion, which is a classification criterion of a combination of color and shape in the image data of the examination target classified by the second level.

The first-order classification criteria may include color values for identifying one or more of an acetic acid response image, a Lugol solution response image, a green filter image, and a normal image as classification criteria values.

The third level of classification criteria may include a combination of colors and shapes for identifying one or more of blood, mucus, rings, colposcopes, treatment scars, and surgical tools in the image data of the examination target.

For example, blood appears predominantly as red flowing down from the center of the cervix, mucus appears predominantly as yellowish flowing down from the center of the cervix, and the ring is located predominantly at the center of the cervix and appears generally and visibly as a line of boomerang shape. Colposcopes and other surgical tools appear in colors other than pink to indicate the cervix (silver, blue, etc.). Therefore, as described above, by using a combination of the color and the form of each foreign substance, foreign substances affecting the cervix can be classified.

Meanwhile, the data classifying unit may separately classify images that are not classified according to the above-described three classification criteria (i.e., images in which no lesion is identified) according to the exposure and focus classification criteria. For example, the classification (fourth classification) may be performed using a characteristic that the histogram is extremely biased to one side when the exposure is insufficient or excessive, or using a characteristic that an edge is not detected or the color contrast is not sufficiently high when the focus is out of range. During the first to fourth classification processes described above, deep learning techniques may be used to perform the classification.

In addition, the data classification unit may also generate additional source learning data (image data of the inspection target) to adjust the balance of the number of image data of the inspection target, that is, the source learning data classified and stored based on the classification criterion. This is to prevent excessive learning from being performed only on a specific type of cervical cancer image data or normal learning from being performed on a specific type (or kind) of image.

The process of classifying the image data of the cervix as the image data of the examination target to optimize the artificial neural network model will be described in detail with reference to fig. 5 and 6.

Fig. 5 shows a multi-level classification flow of image data of an inspection target according to another embodiment of the present invention, and fig. 6 is a diagram for describing in detail a multi-level classification criterion of image data of an inspection object according to still another embodiment of the present invention.

Referring to fig. 5, first, the data receiving unit included in the data classification storage module 110 receives image data of the cervix uteri from an external device such as an image capturing device and stores the received image data (S200).

The data classifying unit classifies and stores one or more unclassified image data pieces of the cervix according to a plurality of multi-level classification criteria (S210).

For example, the data classifying unit preferentially performs the first classification on the unclassified image data of the cervix uteri according to the first-level classification criterion (which is a classification criterion of color).

For the first classification, the data classifying unit may classify the acetic acid reaction image, the lugol solution reaction image, the green filter image, and the normal image by using a color value for identifying the image as a classification standard value.

In detail, since a white spot appears in the cervix of the acetic acid reaction image, the white spot can be distinguished from the cervix and vagina expressed in pink. Since brown or dark orange appears in the Lugol's solution response image and green completely appears in the green filter image, it is possible to classify unclassified image data of the cervix uteri by using a color value indicating an image feature as a classification standard value.

When the first classification is completed, the data classifying unit performs a second classification according to a second-level classification criterion that is a classification criterion for the size of the cervix uteri in the image data subjected to the first classification.

The cervix is a 500 won coin sized circle, usually located in the center of the image. Accordingly, the data classifying unit may classify the image data into an image of only the enlarged cervix, an image of the entire cervix, an image of the cervix occupying about 80%, an image of the cervix occupying about 50%, and an image including the cervix, the colposcope, and other parts, and the like, for the size of the cervix in the image, for the second time.

Subsequently, the data classifying unit performs a third classification of the foreign substance affecting the cervix based on a third classification criterion which is a classification criterion of a combination of a color and a shape in the secondarily classified cervix image data.

As mentioned above, blood appears mainly as red flowing down from the center of the cervix, mucus appears mainly as yellowish flowing down from the center of the cervix, and the ring is mainly located at the center of the cervix and appears usually and apparently as a line in the shape of a boomerang. Colposcopes and other surgical tools appear in colors other than pink to indicate the cervix (silver, blue, etc.). Therefore, as described above, by using a combination of the color and the form of each foreign substance, foreign substances affecting the cervix can be classified.

In some cases, the data classification unit may perform a fourth classification on the third classified images based on the exposure and the focus.

As described above, the image data classified according to the multi-level classification criteria is stored in the memory on the basis of the classification criteria.

When the classification of the unclassified image data is completed, the first deep learning system 100 and the second deep learning system 200 optimize their own artificial neural network models by the process shown in fig. 4. In this case, in order to apply the multi-level classification criteria of the image data employed by the single deep learning system in the same manner, the adjustment information transmission module 130 preferably includes at least two or more of the first level classification criteria, the second level classification criteria, and the third level classification criteria in the adjustment information and transmits the adjustment information to the partner-side deep learning system.

In S220 not described in fig. 5, learning is performed on the image data classified and stored according to the classification criteria, and learning is performed in S40 of fig. 4.

As described above, by the different deep learning systems that share the learning data classification criteria and the parameters modified according to the learning result while accurately classifying the learning data to be used for learning in advance and performing learning based on the classification criteria, even an artificial neural network model under an environment where only a small amount of source data can be learned can be changed to an artificial neural network model that learns a large amount of source learning data, and an artificial neural network model established by learning data made according to a variety of classification criteria can diagnose whether a specific disease has occurred more accurately than the conventional systems.

Although the present invention has been described with reference to the embodiments shown in the drawings, it will be understood by those skilled in the art that the embodiments are merely illustrative of the present invention and various modifications and equivalents may be made without departing from the spirit and scope of the present invention. Therefore, the technical scope of the present invention should be defined by the following claims.

Claims

1. A deep learning system accessible over a communication network to an external deep learning system for learning a first artificial neural network model, the deep learning system comprising:

a data classification storage module configured to receive and store source learning data;

a learning module configured to use the source learning data to cause a second artificial neural network model in the deep learning system to learn;

an adjustment information transmission module configured to transmit adjustment information including learning-related information and parameter information of the second artificial neural network model to the external deep learning system to request adjustment of the first artificial neural network model, wherein the learning-related information and the parameter information are modified according to a result of the learning; and

an adjustment module configured to update the parameter information of the second artificial neural network model according to adjustment information transmitted from the external deep learning system.

2. The deep learning system of claim 1, further comprising a diagnosis module configured to diagnose whether a particular disease has occurred based on the first artificial neural network model optimized by repeated operations of the learning module and the adjustment module.

3. The deep learning system of claim 2,

the data classification storage module comprises:

a data receiving unit configured to receive image data of an inspection target as source learning data from an external apparatus and store the received image data; and

a data classification unit configured to classify and store the stored image data of the inspection target according to a plurality of multi-level classification criteria,

wherein the data classifying unit classifies and stores the image data of the inspection target using at least two or more of a first-level classification criterion as a classification criterion for colors, a second-level classification criterion as a classification criterion for a size of the inspection target, and a third-level classification criterion as a classification criterion for a combination of colors and morphologies.

4. The deep learning system of claim 1,

the data classification storage module comprises:

5. The deep learning system of any one of claims 1 to 4, wherein the adjustment information transmission module preferentially transmits, as the adjustment information, hyper-parameters required to perform synchronization on an initial learning state of the first artificial neural network model.

6. The deep learning system according to any one of claims 1 to 4, wherein the parameter information included in the adjustment information transmitted by the adjustment information transmission module or the external deep learning system constitutes an artificial neural network model and includes a weight value and a bias value of each layer modified according to a learning result.

7. The deep learning system according to any one of claims 1 to 4, wherein the learning-related information includes any one of a ratio of learning data to source learning data to be learned and a number of pieces of learned source learning data.

8. The deep learning system of claim 3 or 4, wherein the data classification unit further generates additional source learning data to adjust a balance of the number of the source learning data classified and stored based on the classification criteria.

9. The deep learning system of claim 3 or 4, wherein the first-level classification criterion of the data classification unit includes a color value for identifying one or more of an acetic acid reaction image, a Lugol solution reaction image, a green filter image, and a normal image as a classification criterion value.

10. The deep learning system of claim 3 or 4, wherein the adjustment information transmission module comprises at least two or more of the first level classification criteria, the second level classification criteria, and the third level classification criteria in the adjustment information and transmits the adjustment information to the external deep learning system.