Image processing apparatus, control method of image processing apparatus, and storage medium

Abstract

The invention discloses an image processing apparatus, a control method of the image processing apparatus, and a storage medium. A mechanism is provided for assigning, to the obtained image, an identification image indicating that the obtained image is an image generated based on the learning model, if it is determined that the obtained image is an image generated based on the learning model. An image processing apparatus includes: an obtaining section for obtaining an image; a determination section configured to determine whether the image obtained by the obtaining section is an image generated based on a learning model; and an assigning section configured to, in a case where it is determined by the determining section that the image obtained by the obtaining section is an image generated based on the learning model, assign, to the image obtained by the obtaining section, a recognition image indicating that the image obtained by the obtaining section is an image generated based on the learning model.

Description

Technical Field

[0001] The present invention relates to an image processing apparatus, a control method for an image processing apparatus, and a storage medium. Background Art

[0002] In recent years, image generation techniques using artificial intelligence (AI) have become widespread. This has enabled individuals to easily generate realistic images. On the other hand, there have been cases of malicious use of AI-based image generation techniques (image generation techniques using AI), which has become a problem. For example, photos of fictional events, fictional incidents, or fictional accidents regarding real people, real facilities, etc. have been forged. To prevent such abuse of AI-based image generation techniques, countries are competing to formulate legislation, development guidelines, etc. regarding AI-generated content. As part of this legislation, legislation or similar legislation is being considered that requires AI-generated content to clearly state that it is generated by AI. In addition, in the future, it is likely that it will be necessary to clearly distinguish between AI-generated content and content other than AI-generated content. Japanese Unexamined Patent Application Publication No. 2006-211477 discloses an image forming apparatus that detects copyright-related read prohibition information (such as copyright marks and / or logos) from image data of banknotes, securities, etc. that are normally prohibited from being copied, and assigns the detected information as information indicating prohibition of copying to the image data.

[0003] However, in the image forming apparatus disclosed in Japanese Unexamined Patent Application Publication No. 2006-211477, even in a case where it is necessary to assign information regarding whether an image is generated by using AI, such assignment is not performed. Summary of the Invention

[0004] The present invention provides a mechanism in which, when it is determined that an obtained image is an image generated based on a learning model, a recognition image indicating that the obtained image is an image generated based on the learning model is assigned to the obtained image.

[0005] Accordingly, the present invention provides an image processing apparatus including: an obtaining unit configured to obtain an image; a determination unit configured to determine whether the image obtained by the obtaining unit is an image generated based on a learning model; and an assignment unit configured to, when the determination unit determines that the image obtained by the obtaining unit is an image generated based on the learning model, assign a recognition image to the image obtained by the obtaining unit, the recognition image indicating that the image obtained by the obtaining unit is an image generated based on the learning module.

[0006] Other features of the present invention will become apparent from the following description of exemplary embodiments with reference to the accompanying drawings. Brief Description of the Drawings

[0007] Figure 1 is a block diagram showing the hardware configuration of the image processing system according to the first embodiment.

[0008] Figure 2 is a block diagram showing the hardware configuration of the AI image generation server.

[0009] Figure 3 is a block diagram showing the hardware configuration of the general-purpose terminal.

[0010] Figure 4A is a flowchart showing the processing executed by the general-purpose terminal, and Figure 4B is a flowchart showing the processing executed by the AI image generation server.

[0011] Figure 5 is a flowchart showing the processing executed by the AI image generation server according to the second embodiment.

[0012] Figure 6A and Figure 6B is a diagram showing an example of the image of the image data.

[0013] Figure 7A , Figure 7B , Figure 7C and Figure 7D is a diagram showing an example of a variation of the image of the image data.

[0014] Figure 8 is a block diagram showing the hardware configuration of the general-purpose terminal according to the third embodiment.

[0015] Figure 9 is a flowchart showing the processing (AI image recognition processing) executed by the general-purpose terminal according to the third embodiment.

[0016] Figure 10 is a flowchart showing the processing (AI image notification processing) executed by the general-purpose terminal according to the third embodiment. Detailed Description of the Invention

[0017] The present invention will be described in detail below with reference to the drawings showing embodiments of the present invention.

[0018] Hereinafter, each embodiment of the present invention will be described in detail with reference to the drawings. However, the configurations described in the following embodiments are merely examples, and the scope of the present invention is not limited by the configurations described in the following embodiments. For example, each part (each unit or each component) constituting the present invention can be replaced with a part (unit or component) having any configuration capable of performing a similar function. In addition, any component can be added. Furthermore, any two or more configurations (features) of the embodiments can be combined.

[0019] <First Embodiment>

[0020] The following will refer to Figures 1 to 4B describe the first embodiment. Figure 1 is a block diagram showing the hardware configuration of the image processing system according to the first embodiment. As Figure 1 shown, the image processing system 10 includes an AI image generation server 101 and a general-purpose terminal 102, which are connected to each other via a network 100 so as to be able to communicate with each other. The AI image generation server 101 is composed of an information processing device. The AI image generation server 101 can generate information based on a learning model (generation step). In the present embodiment, the information is image data of a still image or a moving image (video), but is not limited thereto. For example, it may be audio data (voice data), text data, etc., or may be data including at least one of these data. Hereinafter, generating image data based on a learning model (including a trained model) may be referred to as "AI image generation". The image data is sent to the general-purpose terminal 102. The general-purpose terminal 102 is a device capable of performing various processes on the image data sent from the AI image generation server 101. The general-purpose terminal 102 is not particularly limited. For example, it may be a desktop personal computer, a notebook personal computer, a tablet terminal, a smartphone, etc.

[0021] Figure 2 is a block diagram showing the hardware configuration of the AI image generation server. As Figure 2As shown, the AI image generation server 101 includes a central processing unit (CPU) 201, a random access memory (RAM) 202, a read-only memory (ROM) 203, a storage unit 204, a graphics processing unit (GPU) 207, an AI recognition information distribution unit (information distribution unit) 208, and a network interface (network I / F) 209. The CPU 201 is a computer that controls the operation of the AI image generation server 101 based on a program loaded into the RAM 202. The ROM 203 is a boot ROM that stores, for example, a boot program of the image processing system 10. In addition, the ROM 203 also stores programs for causing the CPU 201 to execute each unit of the AI image generation server 101 and each part of the AI image generation service 101 (control method of the information processing device), etc. The storage unit 204 is a non-volatile device composed of a hard disk drive (HDD), a solid state drive (SSD), etc. The storage unit 204 stores a training model 205, an AI image generation program 206, etc. used in AI image generation. In the present embodiment, the training model 205 and the AI image generation program 206 are used as a generation unit (generation part) 210 for performing AI image generation. It should be noted that for AI image generation, any training model 205 and existing AI image generation programs such as Stable Diffusion can be used, but AI image generation is not limited to using any trained model 205 and existing AI image generation programs such as Stable Diffusion. The training model 205 and the AI image generation program 206 are loaded into the RAM 202 and executed by the CPU 201. The technology related to AI image generation is a well-known technology, so it will be omitted here.

[0022] In response to a request for AI image generation (AI image generation request) from the general-purpose terminal 102, AI image generation is performed. The CPU 201 issues an instruction to the GPU 207 so that the GPU 207 responds to the AI image generation request. The GPU 207 performs AI image generation processing according to the instruction. As a result, image data is generated. The image data includes recognition information that can identify that the image data is generated by AI image generation (i.e., the image data is image data generated based on a learning model). The AI recognition information distribution unit 208 performs distribution of the recognition information of the image data (information distribution step). The image data assigned with the recognition information is sent from the network I / F 209 to the general-purpose terminal 102 via the network 100, or stored in the storage unit 204. It should be noted that the network I / F 209 is connected to the network 100 and is responsible for input and output of various types of information. The connection between the network I / F 209 and the network 100 can be wired or wireless.

[0023] The identification information to be assigned to the image data is not particularly limited. Examples thereof include information indicating that the image data is output data output from the training model 205, and information on the input data to be input to the training model 205 when the output data is output. Other examples of the identification information include the training model 205, the program using the training model 205, information on the probability of the possibility that the image data is image data generated based on the training model 205, information on the AI image generation request, and the like. In addition, at least one of these is specified as the identification information. By making this assignment, it is possible to identify that the image data is image data generated by AI image generation. In the case where the entire image of the image data is not an AI-generated image but only a part of the image of the image data is an AI-generated image, the identification information may include information for identifying the AI-generated image part. The information for identifying the AI-generated image part may be indicated by, for example, the coordinates of the upper left corner and the lower right corner of the image of the image data, or may be pixel information constituting the AI-generated image part. It should be noted that, in addition to the identification information, for example, information on whether the image data has been generated based on the training model 205 certified by the rights holder and information on whether the image data itself has been certified by the rights holder may also be assigned to the image data. In addition, for example, information included in the AI image generation request from the general-purpose terminal 102 and information on copyright in the case where the image data generated by AI image generation is copyright-protected may be assigned to the image data. Such assignment of other information is also performed by the AI identification information assignment unit 208. The image data includes image main body data, which is the main information visualized by the image, and metadata, which is incidental information about the image main body data. The AI identification information assignment unit 208 is capable of assigning the identification information to either the image main body data or the metadata. In the present embodiment, it is assumed that the AI identification information assignment unit 208 assigns the identification information to the metadata. It should be noted that by operating the AI identification information assignment unit 208, it is possible to switch the assignment of the identification information to the image main body data and the assignment of the identification information to the metadata, or one of the two may be determined in advance in the program.

[0024] Figure 3 is a block diagram showing the hardware configuration of the general-purpose terminal. As Figure 3As shown, the general-purpose terminal 102 includes a CPU 301, a RAM 302, an SSD 303, a user I / F 304, and a network I / F 305. The CPU 301 controls the operation of the general-purpose terminal 102 based on the programs loaded into the RAM 302. The SSD 303 stores various programs, etc. For example, these programs also include system programs, AI image generation applications, etc. The user I / F 304 includes, for example, a display, a touch panel, a keyboard, a mouse, etc., and performs input / output processing for the user. The network I / F 305 is connected to the network 100 and is responsible for inputting and outputting various types of information. The connection between the network I / F 305 and the network 100 can be wired or wireless. It should be noted that the general-purpose terminal 102 may have, for example, a telephone function, a camera function, etc.

[0025] Figure 4A is a flowchart showing the processing executed by the general-purpose terminal, while Figure 4B is a flowchart showing the processing executed by the AI image generation server. As Figure 4A shown, in step S401, the CPU 301 of the general-purpose terminal 102 determines that parameters related to the image to be generated by the AI image generation server 101 have been received from the user via the user I / F 304. This parameter is not particularly limited. For example, it can be a keyword, text, image, etc. related to the image used in known AI image generation techniques.

[0026] In step S402, the CPU 301 sends an image generation request based on the parameter in step S401 to the AI image generation server 101 via the network I / F 305.

[0027] As Figure 4B shown, in step S411, the CPU 201 of the AI image generation server 101 determines that the image generation request sent in step S402 has been received via the network I / F 209.

[0028] In step S412, as described above, the CPU 201 controls the GPU 207 to perform AI image generation using the training model 205, etc.

[0029] In step S413, the CPU 201 controls the AI recognition information allocation unit 208 to allocate recognition information to the metadata of the image data that has been generated in step S412. Hereinafter, the image data generated by AI image generation may be referred to as "AI image data". In addition, the AI image to which recognition information has been allocated may be referred to as "image data with allocated recognition information".

[0030] In step S414, the CPU 201 transmits, via the network I / F 209, the image data to which identification information has been assigned in step S413 to the general-purpose terminal 102. As a result, the general-purpose terminal 102 can receive the image data to which identification information has been assigned.

[0031] As described above, in the AI image generation server 101, image data is obtained through AI image generation (see step S412). In addition, in a case where it is necessary to assign identification information indicating that image data has been obtained through AI image generation to the image data, the assignment can be performed (see step S413). The general-purpose terminal 102 receives, via the network I / F 305, the image data to which identification information has been assigned, and the network I / F 305 serves as an acquisition unit capable of acquiring the image data to which identification information has been assigned from the AI image generation server 101. Then, the image data to which identification information has been assigned is displayed as an image on the touch panel of the user I / F 304 of the general-purpose terminal 102. At this time, the user of the general-purpose terminal 102 can refer to the metadata of the image data and confirm the identification information assigned to the metadata on the touch panel. Through this confirmation, the user of the general-purpose terminal 102 can understand that the image displayed on the touch panel includes an AI image. As a result, the user of the general-purpose terminal 102 can, for example, temporarily suspect that the AI image is a fake image.

[0032] <Second Embodiment>

[0033] Hereinafter, reference will be made to Figures 5 to 7D describe the second embodiment. The differences from the above-described first embodiment will be mainly described, and the description of similar matters will be omitted. The second embodiment is similar to the above-described first embodiment except that the AI identification information assignment unit assigns identification information to the image main body data among the image main body data and metadata included in the image data. Figure 5 is a flowchart showing the processing executed by the AI image generation server according to the second embodiment. In Figure 5 the flowchart shown, the processing is executed in the order of step S411, step S412, step S501, and step S414. Steps S411, S412, and S414 are similar to steps S411, S412, and S414 in the Figure 4B flowchart shown. In step S501, the CPU 201 of the AI image generation server 101 controls the AI identification information assignment unit 208 to assign identification information to the image main body data of the image data that has been generated in step S412. As a result, the image data to which identification information has been assigned is obtained.

[0034] Figure 6A and Figure 6B are diagrams showing examples of images of image data. Figure 6AThe image 600 shown is an image of the image main body data included in the AI image data that has been obtained through AI image generation in step S412. Figure 6B The image 601 shown is an image of the image data with identification information assigned to the image main body data in step S501. The image 601 has an image frame 602 and the character "AI" 603 superimposed on the image 601, and the character "AI" 603 is identification information indicating that the image 601 is an AI-generated image. As a result, when the user visually recognizes the image 601 on the touch panel of the user I / F 304 of the general-purpose terminal 102, the user can understand that the image 601 is an AI image. It should be noted that although the character 603 is "AI", they are not limited thereto, and any character indicating that the image is an AI-generated image may be used.

[0035] Figure 7A , Figure 7B , Figure 7C and Figure 7D are diagrams showing examples of variations of the image of the image data. Figure 7A The image 700 shown is an image of the image main body data included in the AI image data that has been obtained through AI image generation in step S412. The image 700 includes a partial image 702 that is an AI-generated image. Figure 7B The image 701A shown is an image of the image data with identification information assigned to the partial image 702 in step S501. The image 701A includes a partial image 703. The partial image 703 has an image frame 703a and the character "AI" 703b superimposed on the partial image 703, and the character "AI" 703b is identification information indicating that the partial image 703 is an AI-generated image. Figure 7C The image 701B shown is an image of the image data with identification information assigned to the partial image 702 in step S501. The image 701B includes a partial image 704. The partial image 704 is an image with a very light color applied over the entire partial image 704, and this very light color is identification information indicating that the partial image 702 is an AI-generated image, or the partial image 704 is an image with a marker such as a star mark superimposed thereon, and this star mark is identification information indicating that the partial image 702 is an image generated by AI. Figure 7DThe image 701C shown is an image of image data with identification information assigned to the partial image 702 in step S501. The image 701C includes a partial image 705. The partial image 705 is an image filled with black, and the black is the identification information indicating that the partial image 702 is an AI-generated image. As a result, by confirming the partial image 703 of the image 701A, the partial image 704 of the image 701B, and the partial image 705 of the image 701C, the user can understand that the images 701A to 701C all include AI images. It should be noted that, for example, invisible information such as digital watermark can be superimposed on the whole image or a partial image as the identification information indicating that the whole image or the partial image is an AI-generated image.

[0036] <Third Embodiment>

[0037] Hereinafter, reference will be made to Figures 8 to 10 describe the third embodiment. The differences from the above first embodiment will be mainly described, and the description of similar matters will be omitted. The third embodiment is similar to the above first embodiment except that the generation unit for generating AI images and the assignment unit for assigning identification information are built in separate devices.

[0038] Figure 8 is a block diagram showing the hardware configuration of a general-purpose terminal according to the third embodiment. As in the above respective embodiments, Figure 8 the general-purpose terminal 800 shown is a device communicably connected to an AI image generation server 101 as an external device. In the third embodiment, in addition to the CPU 301, the RAM 302, the SSD 303, the user I / F 304, and the network I / F 305, the general-purpose terminal 800 further includes an AI image recognition unit (determination unit) 801 and an AI identification information assignment unit (information assignment unit) 802. The network I / F 305 can obtain existing image data (acquisition step) generated by the AI image generation service 101 from the AI image generation server 101. As the existing image data, there are three types of image data: AI image data obtained by AI image generation and before identification information is assigned, image data with identification information assigned in a state where the identification information has been assigned to the AI image data, and non-AI image data not based on AI image generation (non-AI image data not generated by AI image generation).

[0039] The AI image recognition unit 801 determines whether the image data obtained by the network I / F 305 is image data generated by AI image generation (determination step). This determination is made based on the presence or absence of recognition information in the image data. In addition, the CPU 301 can be used as a quantization unit that quantifies, using probabilities, scores, etc., the probability that the image data obtained by the network I / F 305 is data obtained by AI image generation, that is, the probability indicating certainty. In this case, the AI image recognition unit 801 converts the probability into a percentage, and if the numerical value of the probability is greater than or equal to a threshold value (N%), the AI image recognition unit 701 determines that the image data obtained by the network I / F 305 is image data generated by AI image generation. In addition, if the numerical value of the probability is less than the threshold value (N%), the AI image recognition unit 801 determines that the image data obtained by the network I / F 305 is not image data generated by AI image generation. The threshold value can be pre-stored in the SSD (storage unit) 303 or can be appropriately set via the user I / F (operation unit) 304. In addition, it is preferable that the threshold value can be appropriately changed via the user I / F 304. The AI recognition information allocation unit 802 can perform processing similar to that of the AI recognition information allocation unit 208 of the AI image generation server 101. Specifically, as a result of the determination by the AI image recognition unit 801, if it is determined that the image data is image data generated by AI image generation, the AI recognition information allocation unit 802 can allocate recognition information to the image data (information allocation step). In addition, the AI recognition information allocation unit 802 can allocate a probability.

[0040] Figure 9 is a flowchart showing the processing (AI image recognition processing) executed by the general-purpose terminal according to the third embodiment. As Figure 9 shown, in step S900, the CPU 301 determines whether the recognition start button (not shown) for determining whether the image data obtained by the network I / F 305 is image data generated by AI image generation has been operated, that is, whether the recognition start button has been pressed. The recognition start button is provided, for example, on the user interface 304 (user I / F 304). As a result of the determination in step S900, if it is determined that the recognition start button has been operated, the AI image recognition processing proceeds to step S901. On the other hand, as a result of the determination in step S900, if it is determined that the recognition start button has not been operated, the AI image recognition processing remains at step S900 and waits.

[0041] In step S901, the CPU 301 controls the AI image recognition unit 801 to determine whether recognition information has been assigned to the image data obtained by the network I / F 305. As a result of the determination in step S901, in the case where it is determined that recognition information has been assigned to the image data obtained by the network I / F 305, the AI image recognition process ends. In this case, the image data obtained by the network I / F 305 is stored in the SDD 303 as it is. On the other hand, as a result of the determination in step S901, in the case where it is determined that recognition information has not been assigned to the image data obtained by the network I / F 305, the AI image recognition process proceeds to step S903.

[0042] In step S903, the CPU 301 controls the AI image recognition unit 801 to determine whether the image data obtained by the network I / F 305 is image data generated by AI image generation. As a result of the determination in step S903, in the case where it is determined that the image data obtained by the network I / F 305 is image data generated by AI image generation, the AI image recognition process proceeds to step S904. On the other hand, as a result of the determination in step S903, in the case where it is determined that the image data obtained by the network I / F 305 is not image data generated by AI image generation, the AI image recognition process ends. In this case, the image data obtained by the network I / F 305 is stored in the SDD 303 as it is.

[0043] In step S904, the CPU 301 controls the AI recognition information assignment unit 802 to assign recognition information to the image data obtained by the network I / F 305. The recognition information can be assigned to the image main body data, the metadata, or both the image main body data and the metadata. After step S904 has been executed, the AI image recognition process ends. The image data to which the recognition information has been assigned is stored in the SDD 303. It should be noted that Figure 9 in the flowchart shown, the processing order of step S901 and step S903 can be swapped, that is, reversed.

[0044] In addition, as described above, there are three types of image data as the existing image data obtained from the AI image generation server 101. The first type of image data is the AI image data obtained by AI image generation and before the identification information is assigned. The second type of image data is the image data to which the identification information is assigned in a state where the identification information has been assigned to the AI image data. The third type of image data is non-AI image data that is not based on AI image generation (non-AI image data that is not generated by AI image generation). In addition, in the determination in step S901, the first type of image data is determined as "No", the second type of image data is determined as "Yes", and the third type of image data is determined as "No". In addition, in the determination in step S903, the first type of image data is determined as "Yes", while the third type of image data is determined as "No".

[0045] Figure 10 FIG. is a flowchart showing a process (AI image notification process) executed by a general-purpose terminal according to the third embodiment. Here, it is assumed that the user interface 304 includes a speaker. As Figure 10 shown, in step S1000, the CPU 301 determines whether a button (not shown) for opening the image data stored in the SDD 303 during the process of the flowchart shown in Figure 9 has been operated. For example, the button is arranged on the user interface 304. As a result of the determination in step S1000, when it is determined that the button has been operated, the AI image notification process proceeds to step S1001. On the other hand, as a result of the determination in step S1000, when it is determined that the button has not been operated, the AI image notification process remains at step S1000 and waits.

[0046] In step S1001, the CPU 301 controls the AI image recognition unit 801 to determine whether the image data to be opened in step S1000 is image data generated by AI image generation. This determination is made based on the presence or absence of identification information. As a result of the determination in step S1001, when it is determined that the image data to be opened in step S1000 is image data generated by AI image generation, the AI image notification process proceeds to step S1002. On the other hand, as a result of the determination in step S1001, when it is determined that the image data to be opened in step S1000 is not image data generated by AI image generation, the AI image notification process ends.

[0047] In step S1002, the CPU 301 controls the speaker of the user interface 304 to voice - notify that the image data to be opened in step S1000 is image data generated by AI image generation. This notification allows the user to know that the image data is image data generated by AI image generation before opening the image data. It should be noted that the notification target in step S1002 is the image data stored in the SDD 303 after step S901 is executed and the image data stored in the SDD 303 after step S904 is executed. In addition, the notification through the user interface 304 is not limited to voice notification, but can be, for example, notification through an image, notification through lighting, notification through vibration, etc., or a combination of these. In addition, after step S1002 is executed, the CPU 301 can determine whether to forcibly open the image data to be opened in step S1000. This determination can be made based on, for example, whether there is an operation on a button provided on the user interface 304 for forcibly opening the image data. In addition, when the button has been operated, the image data will be forcibly opened.

[0048] Other embodiments

[0049] Embodiments of the present invention can also be implemented by a computer of a system or apparatus that reads and executes computer - executable instructions (e.g., one or more programs) recorded on a storage medium (which can also be more completely referred to as a "non - transitory computer - readable storage medium") to perform one or more of the functions of the above - described embodiments, and / or includes one or more circuits (e.g., an application - specific integrated circuit (ASIC)) for performing one or more of the functions of the above - described embodiments. Moreover, embodiments of the present invention can be implemented by a method of, for example, reading and executing the computer - executable instructions from the storage medium by the computer of the system or apparatus to perform one or more of the functions of the above - described embodiments, and / or controlling the one or more circuits to perform one or more of the functions of the above - described embodiments. The computer can include one or more processors (e.g., a central processing unit (CPU), a micro - processing unit (MPU)), and can include a network of separate computers or separate processors to read and execute the computer - executable instructions. The computer - executable instructions can be provided to the computer, for example, from a network or the storage medium. The storage medium can include, for example, a hard disk, a random access memory (RAM), a read - only memory (ROM), the memory of a distributed computing system, an optical disc (such as a compact disc (CD), a digital versatile disc (DVD), or a Blu - ray disc (BD) TM )、a flash device, and a memory card, etc.

[0050] Embodiments of the present invention can also be implemented by the following method, that is, a software (including a computer program product of computer programs / instructions) that executes the functions of the above embodiments is provided to a system or device through a network or various storage media, and a computer (central processing unit (CPU), microprocessing unit (MPU)) of the system or device reads and executes the computer programs / instructions.

[0051] Although the present invention has been described with reference to exemplary embodiments, it should be understood that the present invention is not limited to the disclosed exemplary embodiments. The scope of the appended claims should be given the broadest interpretation so as to cover all such variations and equivalent structures and functions.

[0052] This application claims priority to Japanese Patent Application No. 2024-007426, filed on January 22, 2024, the entire content of which is incorporated herein by reference.

Claims

1. An image processing apparatus, comprising: an acquisition unit configured to acquire an image; a determination unit configured to determine whether the image acquired by the acquisition unit is an image generated based on a learning model; and an assignment unit configured to, when the determination unit determines that the image acquired by the acquisition unit is an image generated based on the learning model, assign a recognition image to the image acquired by the acquisition unit, the recognition image indicating that the image acquired by the acquisition unit is an image generated based on the learning model.

2. The image processing apparatus according to claim 1, further comprising: a generation unit configured to generate an image based on the learning model, and wherein the acquisition unit acquires the image generated by the generation unit.

3. The image processing apparatus according to claim 1, wherein the image processing apparatus is a device communicably connected to an external device, the external device includes a generation unit configured to generate an image based on the learning model, and the acquisition unit is capable of acquiring the image from the external device.

4. The image processing apparatus according to claim 1, further comprising: a quantization unit configured to quantify the probability of the possibility that the image is an image generated based on the learning model, and wherein when the value quantified by the quantization unit is greater than or equal to a threshold, the determination unit determines that the image is an image generated based on the learning model, and when the value is less than the threshold, the determination unit determines that the image is not an image generated based on the learning model.

5. The image processing apparatus according to claim 4, further comprising: a storage unit configured to store the threshold; and an operation unit configured to perform an operation of changing the threshold.

6. The image processing apparatus according to any one of claims 1 to 5, wherein the image includes metadata.

7. The image processing apparatus according to any one of claims 1 to 5, wherein the assignment unit assigns at least one of the following information indicating information as the recognition image: the image is output data output from the learning model, input data to be input into the learning module when outputting the output data, the learning model, a program using the learning model, and the probability of the possibility that the image is an image generated based on the learning model.

8. A control method for controlling an image processing apparatus, the control method comprising: an acquisition step of acquiring an image; a determination step of determining whether the image acquired in the acquisition step is an image generated based on a learning model; and an assignment step of, when it is determined in the determination step that the image acquired in the acquisition step is an image generated based on the learning model, assigning a recognition image to the image acquired in the acquisition step, the recognition image indicating that the image acquired in the acquisition step is an image generated based on the learning model.

9. A non-transitory computer-readable storage medium storing a program for causing a computer to execute a control method for controlling an image processing apparatus, The control method includes: An obtaining step of obtaining an image; A determining step of determining whether the image obtained in the obtaining step is an image generated based on a learning model; And An assigning step of, when it is determined in the determining step that the image obtained in the obtaining step is an image generated based on the learning model, assigning a recognition image to the image obtained in the obtaining step, where the recognition image indicates that the image obtained in the obtaining step is an image generated based on the learning model.

Images