WO2023033241A1

WO2023033241A1 - Device for generating art psychological analysis data, and method for augmenting and effectively managing art psychological analysis data using same

Info

Publication number: WO2023033241A1
Application number: PCT/KR2021/016748
Authority: WO
Inventors: 김지훈; 박기석
Original assignee: 주식회사 파블로아트컴퍼니
Priority date: 2021-09-03
Filing date: 2021-11-16
Publication date: 2023-03-09
Also published as: JP2023543529A

Abstract

A device for generating art psychological analysis data, according to one embodiment of the present invention, may comprise: a user interface for providing an interactive environment for a user who is a person to be tested; a first drawing analysis unit which analyzes drawing data learned by using a first artificial intelligence model and inputted by the user via the user interface, and which cuts region-of-interest data from the drawing data and compresses the region-of-interest data or decompresses the compressed region-of-interest data in order to restore the drawing data; a storage unit for storing labeling information for the size, location, and angle of the compressed region-of-interest data; and a second drawing analysis unit which has a second artificial intelligence model for partially augmenting and learning an object of an art psychological analysis element included in the drawing data, and which generates art psychological analysis data through a combination of the object.

Description

Apparatus for generating data for psychological analysis of art and method for augmenting and efficiently managing data for psychological analysis of art using the same

The present invention relates to an apparatus for generating data for psychological analysis of art and a method for augmenting and efficiently managing data for psychological analysis of art using the same, and more particularly, to a region of interest (ROI) that is significant for psychological analysis of art using an autoencoder. Focusing on art psychoanalysis, the compression rate is improved along with the refinement of picture data for art psychoanalysis, and the artificial intelligence model for art psychoanalysis is learned by partially augmenting art psychoanalysis elements from picture data using Partial GAN (Generative Adversarial Network). It's about technology that can do it.

As a conventional technique for efficiently storing data in a storage space, an autoencoder using a data compression technique and a deep learning technique has been disclosed. An autoencoder can be divided into an encoder part and a decoder part. The encoder embeds the original image into a specific feature domain and transforms the value into a feature level, and the decoder compresses the feature level value into the original image. It is a part to release (recover), and many studies have been conducted on these encoders and decoders as neural networks using deep learning.

Considering the characteristics of the image for art psychological test, when the original image is applied to the autoencoder, there is a problem in that the amount of meaningless information is considerably larger than the information required for the actual art psychological test. For example, when a house is drawn in the HTP test, there are unnecessarily many blanks in the original image in addition to the information required for the psychological test of art. Although it may be an important factor in the psychological test result, details such as the shape and form of the window are not elements that are greatly reflected in the result of the psychological test of art.

In addition, when the original image is applied to the autoencoder, unnecessary elements must also be decompressed, so the characteristic dimension to be converted inevitably becomes enormous in proportion to the amount of unnecessary information, and there is a problem in that compression efficiency is inevitably reduced.

On the other hand, when learning an artificial intelligence model with limited learning data, a data augmentation technique is commonly used as a technique to increase performance through regularization of the artificial intelligence model. As an augmentation technique, geometric distortion ( geometry distortion, photometric distortion, GAN augmentation, and combining different images.

Among the augmentation techniques, GAN augmentation uses deep learning to generate an image similar to an actual image for a specific class called GAN (Generative Adversarial Network) to create a new image that did not exist before without feeling different from the shape of the existing image. It is a method to improve the performance of an artificial intelligence model by increasing the training data.

When generating data using a conventional GAN, the most important purpose is to create images of the same class without a sense of heterogeneity. However, what is important in art psychological testing is that it is important to create images of the same class without heterogeneity, but since the presence/absence of psychological analysis elements in the entire image and how they are drawn are important, conventional GANs are There is a problem in that it is difficult to create data while giving differences . In addition, when an image is randomly generated, additional labeling of the generated image must be performed, resulting in increased cost.

The present invention is a data generating device for psychological analysis of art that can efficiently store data in a storage unit of limited capacity by cutting and compressing significant region of interest data according to psychological analysis of art with an autoencoder and then compressing to improve the compression efficiency of picture data. Its purpose is to provide augmentation and efficient management of data for psychological analysis of art using this.

In addition, the present invention partially augments the object of the art psychoanalysis element included in the picture data with Partial GAN (Generative Adversarial Network), and inputs labeling information for the size, position, and angle of the object, which is used for psychological analysis of art. The purpose is to provide a device for generating data for psychological analysis of art that can reduce the cost of labeling work by learning artificial intelligence models, and a method for augmenting and efficiently managing data for psychological analysis of art using the device.

In addition, an object of the present invention is to provide an apparatus for generating data for psychological analysis of art that can derive various results of psychological analysis of art from picture data with a combination of Partial GAN and a method for augmenting and efficiently managing data for psychological analysis of art using the device.

However, the technical problem to be achieved in the present invention is not limited to the above-mentioned technical problems, and other technical problems not mentioned will become clear to those skilled in the art from the description below. You will be able to understand.

In order to achieve the above object, an apparatus for generating data for psychological analysis of art according to an embodiment of the present invention includes a user interface for providing an interactive environment to a user who is a test subject; It is learned using the first artificial intelligence model, and the picture data input from the user through the user interface is analyzed to cut the ROI data among the picture data and compress the ROI data or compress the compressed ROI data. a first picture analysis unit for restoring the picture data by releasing; a storage unit to store labeling information about the size, position, and angle of an object included in the compressed ROI data; and a second artificial intelligence model that partially augments and learns objects of the art psychoanalysis element included in the picture data, and generates data for art psychology analysis through a combination of the objects. can include

Also, the first artificial intelligence model may be an autoencoder for compressing or decompressing the ROI data.

The first picture analysis unit may cut the ROI data from among the picture data, normalize the ROI data to a predetermined size, and encode the normalized ROI data to compress characteristic information.

Also, the characteristic information may be labeling information about the size, position, and angle of an object included in the ROI data.

The first picture analysis unit decodes the characteristic information, decompresses the normalized ROI data, and adjusts the size of an object included in the normalized ROI data through the size information so that the object is the After recovering the size to be the same as the object of the picture data, identifying the position of the object included in the normalized ROI data through the location information, adjusting the position of the object to the same position as the object of the picture data The picture data can be restored.

In addition, the second artificial intelligence model may be a partial partial generative adversarial network (GAN) for partially augmenting an object included in normalized ROI data decompressed through decoding of the characteristic information.

And the Partial GAN is provided in plurality in the second picture analysis unit to partially augment the object, learns through feedback on the partial augmentation of the object, and analyzes art psychology with the partial augmentation of the object. data can be generated.

In addition, a database in which art-based psychological analysis information is pre-stored; a psychological state analysis unit that is learned using a third artificial intelligence model and primarily derives psychological state information of the user by analyzing the art psychological analysis data and the art-based psychological analysis information; and a communication unit for communicating with a terminal of a second user who is a picture analysis expert.

And the terminal of the second user provides the second user with picture data that the user inputs into the user interface, and inputs additional information including a psychological analysis opinion obtained by the second user analyzing the picture data. A second user interface may be provided to enable this.

In addition, the psychological state analysis unit adds the additional information to the psychological state information of the user primarily derived through the analysis of the art psychological analysis data and the art-based psychological analysis information to obtain the user's psychological state information as a second can be derived

A method for augmenting and efficiently managing data for psychological analysis of art according to an embodiment of the present invention includes a first picture analysis unit using a first artificial intelligence model, a second picture analysis unit and a second picture analysis unit using a second artificial intelligence model. 3 learning the psychological state analysis unit using an artificial intelligence model; cutting the ROI data among the picture data input by the user through a user interface and compressing the ROI data by the first picture analysis unit; storing labeling information about the size, position, and angle of an object included in the compressed ROI data in a storage unit; restoring the picture data using the size information and location information after the first picture analyzer decompresses the compressed ROI data through the characteristic information; and generating data for psychological analysis of art through a combination of the objects, by the second picture analysis unit, which has learned by partially augmenting the object of the psychological analysis element included in the picture data.

In addition, according to an embodiment of the present invention, in the method for augmenting and efficiently managing data for psychological analysis of art, the psychological state analysis unit analyzes the data for psychological analysis of art and the art-based psychological analysis information previously stored in a database to determine the user's Primary derivation of psychological state information; Receiving picture data of the user through a second user interface and a communication unit provided in a terminal of a second user who is an expert in psychological analysis; After the second user analyzes the picture data, inputting additional information including a psychological analysis opinion on the picture data into the second user interface; secondarily deriving mental state information of the user by adding the additional information received through the communication unit to the mental state information of the user primarily derived by the mental state analysis unit; and providing the first or second derived psychological state information of the user to the user through the user interface.

A computer program stored in a computer-readable recording medium may be provided for executing a method of augmenting and efficiently managing data for psychological analysis of art according to embodiments of the present invention.

The present invention does not store picture data in a storage unit, but rather improves the compression efficiency of picture data by compressing the data of the region of interest to be used for psychological analysis of art among picture data, and efficiently stores the data in a storage unit with limited capacity. can

In addition, the present invention partially augments an object of an art psychoanalysis element from picture data with a partial GAN (Generative Adversarial Network), and inputs labeling information about the size, position and angle of the object, thereby reducing the cost of labeling work. can

In addition, the present invention can derive various art psychological analysis results from picture data by GAN combination.

However, the effects obtainable in the present invention are not limited to the effects mentioned above, and other effects not mentioned will be clearly understood by those skilled in the art from the description below. You will be able to.

1 is a block diagram schematically showing an apparatus for generating data for psychological analysis of art according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram illustrating an example of a process of cutting the ROI data and compressing the ROI data of the first picture analysis unit shown in FIG. 1 .

FIG. 3 is a conceptual diagram illustrating an example of a picture data recovery process of the first picture analysis unit shown in FIG. 1 .

FIG. 4 is a diagram showing an example of a GAN learning process of the second picture analysis unit shown in FIG. 1 .

FIG. 5 is a diagram illustrating an example of result data for psychological analysis of art generated by a GAN combination by the second picture analysis unit shown in FIG. 1 .

6 is a schematic block diagram of an apparatus for generating data for psychological analysis of art according to another embodiment of the present invention.

7 is a flowchart illustrating a method for augmenting and efficiently managing data for psychological analysis of art according to an embodiment of the present invention.

8 is a flowchart illustrating a method for augmenting and efficiently managing data for psychological analysis of art according to another embodiment of the present invention.

Hereinafter, with reference to the accompanying drawings, embodiments of the present invention will be described in detail so that those skilled in the art can easily carry out the present invention. However, since the description of the present invention is only an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiments described in the text. That is, since the embodiment can be changed in various ways and can have various forms, it should be understood that the scope of the present invention includes equivalents capable of realizing the technical idea. In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all of them or only such effects, the scope of the present invention should not be construed as being limited thereto.

The meaning of terms described in the present invention should be understood as follows.

Terms such as "first" and "second" are used to distinguish one component from another, and the scope of rights should not be limited by these terms. For example, a first element may be termed a second element, and similarly, a second element may be termed a first element. It should be understood that when an element is referred to as “connected” to another element, it may be directly connected to the other element, but other elements may exist in the middle. On the other hand, when an element is referred to as being “directly connected” to another element, it should be understood that no intervening elements exist. Meanwhile, other expressions describing the relationship between components, such as “between” and “immediately between” or “adjacent to” and “directly adjacent to” should be interpreted similarly.

Singular expressions should be understood to include plural expressions unless the context clearly dictates otherwise, and terms such as “comprise” or “having” refer to a described feature, number, step, operation, component, part, or It should be understood that it is intended to indicate that a combination exists, and does not preclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs, unless defined otherwise. Terms defined in commonly used dictionaries should be interpreted as consistent with meanings in the context of related art, and cannot be interpreted as having ideal or excessively formal meanings unless explicitly defined in the present invention.

The data generation device 1 for psychological analysis of art of the present invention improves the compression efficiency of picture data to be used for psychological analysis of art through encoding and decoding of an autoencoder, and partially It is a system that derives various art psychological analysis results through picture data augmentation, and the components of the system to implement this are as follows.

Referring to FIG. 1 , an apparatus for generating data for psychological analysis of art (1) according to an embodiment of the present invention includes a user interface (10), a first picture analysis unit (20), a storage unit (30), and a second picture analysis unit. A unit 40 is provided, and a psychological state analysis unit 50 and a database 60 may be further provided to provide psychological state information to the user 2 who is a test subject who wants to receive the art psychological analysis service.

The user interface 10 provides an environment in which the user 2 can interact with the data generating device 1 for psychological analysis of art, and is a hardware device for receiving commands from the user 2 and converting them into electronic data. and software programs. For example, an input device such as a keyboard, mouse, or touch pen, an output device such as a display, and a drawing application for processing and displaying data such as outlines and colors input through the input device on the output device in real time may be included. there is.

In addition, the user interface 10 receives picture data by the user 2, and the picture data of the present invention may be a picture that includes an object to be an art psychoanalysis element drawn by the user and items constituting the object. . In the present invention, the object and item of picture data and the art psychological test method for the art psychological analysis service are not limited, but the art psychological test method presents drawing a house, tree, and person as a task, and then It may be an HTP test that analyzes the user's psychological state from information such as color and location. Accordingly, the object of the picture data drawn by the user may be at least one of a house, a tree, and a person, and items constituting a house, which is an object of the picture data, include a house frame (roof, wall, chimney, etc.), windows and doors, The items that make up a tree can include stems, roots, leaves, and fruits.

The first picture analysis unit 20 analyzes the picture data input from the user through the user interface 10, crops Region of Interest (ROI) data among the picture data, and then compresses the ROI data. (encoding) or recovering (decoding) picture data by decompressing compressed ROI data, and may be learned using the first artificial intelligence model 21 to perform encoding and decoding processes.

In this case, the first artificial intelligence model 21 may be an autoencoder for compressing or decompressing ROI data through encoding and decoding processes.

An example of a process in which the first picture analysis unit 20 learned with the first artificial intelligence model 21 compresses the ROI data from the picture data is as follows.

FIG. 2 is a conceptual diagram illustrating an example of a process of cutting out the ROI data of the first picture analysis unit shown in FIG. 1 and compressing the ROI data, and FIG. 3 is a picture data recovery process of the first picture analysis unit shown in FIG. 1 It is a conceptual diagram showing an example.

Referring to FIG. 2 , the first picture analysis unit 20 cuts the data of the region of interest including the object to be used as an evaluation factor in the psychological analysis of art among the picture data input to the user interface 10 ('ROI in FIG. 2). Cropping'), normalizing the ROI data to a preset size ('Nomalization' in FIG. 2), and encoding the normalized ROI data ('Encoding' in FIG. 2) to compress characteristic information. At this time, the characteristic information is labeling information about the size, position, and angle of the object included in the region of interest data so that the first artificial intelligence model 21 can learn, and the first artificial intelligence model ( 21) means attribute (tag) information that can be recognized.

Labeling information about the size, position, and angle of an object compressed through encoding by the first picture analysis unit 20 may be stored in the storage unit 30 .

Meanwhile, picture data input to the user interface 10 may include input time information of pixels. This is a problem in the conventional non-face-to-face psychological analysis of art, in which the inspector cannot observe in real time the moment the user 2, the test subject, draws a picture, making it difficult to grasp the priority and importance of objects in the picture data. For example, when drawing a tree, psychological analysis results can be different depending on whether the trunk or roots are drawn first. difficult.

Accordingly, the data generation device 1 for psychological analysis of art enables the user's psychological state to be analyzed in consideration of the order in which objects included in the picture data are drawn, and for this purpose, the first picture analysis unit 20 provides a user interface ( 10) determines the drawing order of objects and items included in the picture data received, and stores the order information in which objects and items are drawn in the storage unit 30, and the second picture analysis unit 40 determines the user's art When creating psychological analysis data, by including the information on the order in which objects and items are drawn in the user's art psychological analysis data, considering the order in which objects and items are drawn in the psychological state analysis unit 50, the user's Let your psychological state be analyzed. At this time, in the database 60, the order in which objects and items to be transmitted are drawn when the psychological state of the user is analyzed so that the mental state analyzer 50 analyzes the user's psychological state in consideration of the order in which the objects and items are drawn. It is preferable that psychological analysis information according to is pre-stored.

Hereinafter, some examples of psychoanalysis information according to the order in which the test subject, the user 2, draws a picture are introduced.

Example 1: When a test subject is instructed to draw a family member, if a specific family member (especially a parent) is drawn last, there is a high possibility of a lack of attachment to that member.

Example 2: If the test subject is instructed to draw family members including herself, there is a high possibility that the test subject lacks self-esteem or narcissism if she draws herself last.

Example 3: If a person draws a specific object, erases it, and repeats drawing, there is a high possibility of separation anxiety symptoms such as lack of confidence, anxiety, and obsessive compulsiveness.

According to the above embodiment, it is possible to overcome the limitations of treatment based on non-face-to-face psychological analysis of art, in which it is difficult to observe the process of drawing a picture in real time. That is, it is possible to provide more objective psychological analysis results because it is possible to know exactly which object the examinee considers more important when drawing a picture.

The storage unit 30 stores the labeling information on the size, position, and angle of the object compressed through the first picture analysis unit 20, compared to storing picture data input to the user interface 10 as it is. The compression efficiency of data is relatively improved, and through this, data can be efficiently stored in a limited capacity.

This storage unit 30 is a volatile or non-volatile storage medium installed inside the data generating device 1 for art psychoanalysis, or storage that exists outside the data generating device 1 for art psychoanalysis accessible through a network. can be a medium.

The first picture analysis unit 20 learns to recover picture data through labeling information about the size, position, and angle of an object stored in the storage unit 30 in the decoding process of the first artificial intelligence model 21. , An example of a process in which the first picture analysis unit 20 recovers picture data through decoding is as follows.

Referring to FIG. 3, the first picture analysis unit 20 decompresses the normalized ROI data by decoding the characteristic information stored in the storage unit 30 ('Decoding' in FIG. 3), and When the compression is released, the size of the object included in the normalized ROI data is adjusted ('Resize' in FIG. 3) through the size information stored in the storage unit 30 so that the object is restored to the same size as the object of the picture data. and, when the size adjustment of the object is completed, the location of the object included in the normalized ROI data is identified through the location information stored in the storage unit 30, and then the location of the object is adjusted to the same location as that of the object in the picture data. ('Localization' in FIG. 3) to restore picture data.

Referring back to FIG. 1, the second picture analysis unit 40 partially augments the object of the psychological art analysis element included in the normalized ROI data decompressed in the decoding process of the first picture analysis unit 30. In order to do this, it learns with the second artificial intelligence model 41, and through the learning of the second artificial intelligence model 41, data for art psychological analysis is created with a combination of objects.

At this time, the second artificial intelligence model 41 may be a partial GAN for partially augmenting an object included in the normalized ROI data.

In addition, a plurality of Partial GANs are provided in the second picture analysis unit 40 to partially augment an object included in the normalized ROI data, and learn through feedback on partial augmentation of the object, and Partial augmentation can generate data for psychological analysis of art. In other words, the combination of Partial GAN can derive various data for psychological analysis of art.

An example of the learning process of the Partial GAN is as follows. In the following example, the Partial GAN performs a plurality of separate creations of a house frame, a window, and a door, which are items constituting a house, which is one of the objects for HTP inspection. It can be done with GAN.

4 is a diagram showing an example of the GAN learning process of the second picture analysis unit shown in FIG. 1, and FIG. 5 shows an example of result data for art psychological analysis generated by the GAN combination of the second picture analysis unit shown in FIG. it is a drawing

Referring to FIG. 4, the second picture analysis unit 40 is composed of a Partial GAN, a second artificial intelligence model, which is a GAN for generating a house frame, a GAN for generating a window, and a GAN for generating a door, which are elements of psychological analysis of art. , it is possible to perform learning to create house frames, windows, and doors through input of normalized ROI data.

Referring to FIG. 5, the second picture analysis unit 40, through a plurality of learned Partial GAN combinations, shows a house in which a house frame and a window are combined, a house in which a door and a house frame are combined, and a house in which a window, a door, and a house frame are combined. Houses of various cases can be created, and by learning through them, various data for art psychological analysis related to the house can be created.

Referring back to FIG. 1 , the psychological state analysis unit 50 analyzes the art psychological analysis data generated by the second picture analysis unit 40 and the art-based psychological analysis information previously stored in the database 60 to analyze the user's psychological state. State information is first derived, and for this purpose, it is learned using the third artificial intelligence model 51.

At this time, the third artificial intelligence model 51 is composed of multiple layers and means a learning model configured to implement functions similar to those of a human neural network, and is based on RNN (Recurrent Neural Network), CNN (Convolutional Neural Network), and attention. An artificial neural network model such as a model of is used, and may be learned to derive psychological state information of a user by searching from art-based psychological analysis information previously stored in the database 60 .

In addition, the psychological state information of the user primarily derived by the psychological state analysis unit 50 may be output from a display provided in the data generating device 1 for psychological analysis of art through the user interface 10 and provided to the user. there is.

The database 60 stores art psychoanalysis information in advance. The art-based psychoanalysis information includes which objects and items are included or omitted when a user draws a picture, whether objects and items are drawn large or small, objects and items, and It includes psychological analysis result data on how the items differ from general social norms, accumulated face-to-face or non-face-to-face psychological counseling data, and thesis data related to art-based psychological analysis. It can be updated based on the results of its own psychological analysis obtained in 1).

Hereinafter, the data generating device 1 for art psychological analysis according to another embodiment of the present invention will be described in detail, focusing on differences from the data generating device 1 for psychological analysis of art according to an embodiment of the present invention. would.

Referring to FIG. 6 , an apparatus for generating data for psychological analysis of art (1) according to another embodiment of the present invention includes a second user's terminal ( 4) may be further provided with a communication unit 70 for communication.

The terminal 4 of the second user provides the second user 3 with picture data input by the user 2 to the user interface 10, or the second user 3 is a data generating device for psychological analysis of art. A second user interface may be provided to enable input of additional information including psychological analysis opinions derived by analyzing the picture data received from (1).

The psychological state analysis unit 50 receives the additional information through the communication unit 70 after the additional information is input to the second user interface, the data for psychological analysis of art and the database 60 of the second picture analysis unit 40 ), the user's psychological state information can be derived secondarily by adding the additional information to the user's psychological state information derived primarily through the analysis of art-based psychoanalysis information pre-stored in ).

That is, the secondly derived user's mental state information is added with additional information including the psychological analysis opinion of the second user 3, and the information changed from the user's psychological state information derived primarily according to the addition of the additional information. It may be user's mental state information.

On the other hand, it is preferable that the data generation device 1 for art psychological analysis outputs the user's psychological state information first or secondarily derived by the psychological state analysis unit 50 through the user interface 10 . However, when the secondly derived user's mental state information is output, the user's firstly derived psychological state information without additional information can be controlled to be output as well. can provide information.

Hereinafter, the process of the psychological art analysis method using the data generating device 1 for psychological analysis of art according to an embodiment of the present invention will be described in detail.

7, the first picture analysis unit 20 using the first artificial intelligence model 21, the second picture analysis unit 40 and the third artificial intelligence using the second artificial intelligence model 41 The psychological state analysis unit 50 may be trained using the model 51 (S710).

Thereafter, the first picture analysis unit 20 cuts the object of the art psychological analysis element and the data of the region of interest including the items constituting the object among the picture data input by the user 2 through the user interface 10. The ROI data is compressed (S720), and labeling information about the size, position, and angle of the compressed ROI data may be stored in the storage unit 30 (S730).

Thereafter, the first picture analysis unit 20 receives characteristic information from the storage unit 30, decompresses the compressed ROI data to generate normalized ROI data, and uses size information and location information to generate Picture data can be recovered (S740).

After that, the second picture analysis unit 40 partially augmented and learned the object of the art psychoanalysis element included in the picture data can generate data for art psychoanalysis with a combination of the objects (S750).

After that, the psychological state analysis unit 50 analyzes the art psychological analysis data generated by the second picture analysis unit 40 and the art-based psychological analysis information previously stored in the database 60 to obtain the user's psychological state information as 1 The difference can be derived (S760).

The process (S710 to S760) of the art psychoanalysis method using the art psychoanalysis data generating device 1 according to an embodiment of the present invention is the art psychoanalysis data generating device described with reference to FIGS. 1 to 5 ( 1), and since details are the same as those described above, duplicate descriptions will be omitted for convenience.

Furthermore, the process of the psychological art analysis method using the data generating device 1 for psychological analysis of art according to an embodiment of the present invention may further include a second step of deriving psychological state information of the user. Same as

Hereinafter, a data generating apparatus for art psychological analysis according to another embodiment of the present invention (1) focusing on differences from the art psychological analysis method using the data generating apparatus (1) for art psychological analysis according to an embodiment of the present invention. ) will be explained in detail about the process of the art psychoanalysis method.

In addition, a description of the process (S810 to S860) of the art psychoanalysis method, which is the same as the process of the art psychoanalysis method of one embodiment, is omitted for convenience, except that only the code is different from that of the embodiment.

Referring to FIG. 8 , after the psychological state information of the user is first derived, the second user 3 who is a psychological analysis expert communicates with the second user interface provided in the terminal 4 of the second user and the communication unit 70. The user's picture data may be provided from the data generating device 1 for psychological analysis of art (S870).

Thereafter, the second user may analyze the user's picture data and input additional information including a psychological analysis opinion on the picture data to the second user interface (S880).

After that, the mental state analysis unit 50 adds additional information received from the terminal 4 of the second user through the communication unit 70 to the first derived mental state information of the user, thereby converting the user's mental state information to 2 The difference can be derived (S890).

Thereafter, the apparatus 1 for generating data for psychological analysis of art may provide the first or second derived psychological state information of the user to the user through the user interface 10 (S900). At this time, the data generating device 1 for psychological analysis of art may output the user's psychological state information derived primarily and the user's psychological state information derived secondarily through control of the display.

The process (S860 to S900) of the art psychological analysis method using the art psychological analysis data generating device 1 according to another embodiment of the present invention is performed in the art psychological analysis data generating device 1 described with reference to FIG. It can be performed by, and since the details are the same as those described above, redundant descriptions will be omitted for convenience.

On the other hand, the art psychoanalysis method of the data generating device 1 for art psychoanalysis according to the embodiments of the present invention is implemented as an application or implemented in the form of program instructions that can be executed through various computer components and is computer readable. It can be recorded on a recording medium. The computer readable recording medium may include program instructions, data files, data structures, etc. alone or in combination.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like.

Detailed descriptions of the preferred embodiments of the present invention disclosed as described above are provided to enable those skilled in the art to implement and practice the present invention. Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art will understand that the present invention can be variously modified and changed without departing from the scope of the present invention. For example, those skilled in the art can use each configuration described in the above-described embodiments in a manner of combining with each other. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The present invention may be embodied in other specific forms without departing from the technical spirit and essential characteristics of the present invention. Accordingly, the above detailed description should not be construed as limiting in all respects and should be considered as illustrative. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of the present invention are included in the scope of the present invention. The invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein. In addition, claims that do not have an explicit citation relationship in the claims may be combined to form an embodiment or may be included as new claims by amendment after filing.

Claims

a user interface for providing an interactive environment to a user who is a test subject;

It is learned using the first artificial intelligence model, and the picture data input from the user through the user interface is analyzed to cut the ROI data among the picture data and compress the ROI data or compress the compressed ROI data. a first picture analysis unit for restoring the picture data by releasing;

a storage unit to store labeling information about the size, position, and angle of an object included in the compressed ROI data; and

A second artificial intelligence model that partially augments and learns objects of art psychoanalysis elements included in the picture data, and a second picture analysis unit that generates data for art psychology analysis through a combination of the objects; A data generating device for psychological analysis of art, characterized in that.
According to claim 1,

The first artificial intelligence model,

Data generating device for psychological analysis of art, characterized in that the autoencoder (autoencoder) for compressing or decompressing the region of interest data.
According to claim 2,

The first picture analysis unit,

After cutting the region of interest data among the picture data, normalizing the region of interest data to a preset size, and encoding the normalized region of interest data to compress characteristic information.
According to claim 3,

The characteristic information is

Data generating device for psychological analysis of art, characterized in that the labeling information for the size, position, angle of the object included in the region of interest data.
According to claim 4,

The first picture analysis unit,

decoding the characteristic information to decompress the normalized ROI data;

Adjusting the size of an object included in the normalized ROI data through the size information to restore the object to have the same size as the object of the picture data;

After identifying the position of the object included in the normalized region-of-interest data through the location information, the position of the object is adjusted to the same position as the object of the picture data to restore the picture data. Data generating device for analysis.
According to claim 5,

The second artificial intelligence model,

A data generating device for psychological analysis of art, characterized in that it is a Partial Generative Adversarial Network (GAN) for partially augmenting an object included in the normalized ROI data decompressed through decoding of the characteristic information.
According to claim 6,

The Partial GAN,

In order to partially augment the object, a plurality of second picture analysis units are provided, learning through feedback on the partial augmentation of the object, and generating data for psychological analysis of art with partial augmentation of the object. A data generating device for psychological analysis of art.
According to claim 1,

A database pre-stored with art-based psychological analysis information;

a psychological state analysis unit that is learned using a third artificial intelligence model and primarily derives psychological state information of the user by analyzing the art psychological analysis data and the art-based psychological analysis information; and

A data generating device for psychological analysis of art, characterized in that it further comprises; a communication unit for communicating with a terminal of a second user who is an expert in picture analysis.
According to claim 8,

The terminal of the second user,

Picture data input by the user to the user interface is provided to the second user, and a second user interface is provided so that the second user can input additional information including a psychological analysis opinion obtained by analyzing the picture data. Data generating device for psychological analysis of art, characterized in that.
According to claim 9,

The psychological state analysis unit,

The psychological state information of the user is secondarily derived by adding the additional information to the psychological state information of the user primarily derived through the analysis of the art psychological analysis data and the art-based psychological analysis information. Data generating device for analysis.
Learning a first picture analysis unit using a first artificial intelligence model, a second picture analysis unit using a second artificial intelligence model, and a psychological state analysis unit using a third artificial intelligence model;

cutting the ROI data among the picture data input by the user through a user interface and compressing the ROI data by the first picture analysis unit;

storing labeling information about the size, position, and angle of an object included in the compressed ROI data in a storage unit;

restoring the picture data using the size information and location information after the first picture analyzer decompresses the compressed ROI data through the characteristic information; and

and generating data for psychological analysis of art through a combination of the objects by the second picture analysis unit, which has partially augmented and learned an object of the psychological analysis element of art included in the picture data. Augmentation and efficient management of data for analysis.
According to claim 11,

first deriving psychological state information of the user by the psychological state analysis unit analyzing the art psychological analysis data and art-based psychological analysis information previously stored in a database;

Receiving picture data of the user through a second user interface and a communication unit provided in a terminal of a second user who is an expert in psychological analysis;

After the second user analyzes the picture data, inputting additional information including a psychological analysis opinion on the picture data into the second user interface;

secondarily deriving mental state information of the user by adding the additional information received through the communication unit to the mental state information of the user primarily derived by the mental state analysis unit; and

Providing first or secondly derived psychological state information of the user to the user through the user interface; Augmentation and efficient management of data for psychological analysis of art, characterized in that it further comprises.
A computer program stored in a computer-readable recording medium for executing the augmentation and efficient management method of data for psychological analysis of art according to at least one of claims 11 and 12.