Information processing device

Abstract

The data identification results are difficult to utilize. [Solution] The information processing device of the present disclosure comprises an input unit that inputs input data, including information based on target data and information based on the identification result of the target data by an identification model, to a model that analyzes data, and an output unit that outputs interpretation data from the model in multiple data formats, representing an interpretation of the identification result according to the input data.

Description

Technical Field

[0001] This disclosure relates to an information processing apparatus.

Background Art

[0002] In recent years, with the technology of generative AI (Artificial Intelligence), it has become easier to generate data such as voice, images, and text. Along with this, so-called fake data called deep fakes, such as non-existent voice, images, and text, have also been generated, and the abuse of such fake data has become a problem. For this reason, as described in Patent Document 1, it has become important to perform authenticity determination as to whether data such as voice, images, and text is real or not.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, simply performing authenticity determination of data makes the basis for such determination unclear. Also, when identifying types not limited to the authenticity of data, the basis for identifying the types is unclear. As a result, there arises a problem that it is difficult to utilize the identification result of data.

[0005] Therefore, one of the objects of this disclosure is to solve the above-described problem that it is difficult to utilize the identification result of data.

Means for Solving the Problems

[0006] An information processing apparatus which is one form of this disclosure is An input unit that inputs input data, including information based on target data and information based on the identification result of the target data by the identification model, to a model that analyzes data. An output unit that outputs interpretation data representing an interpretation of the identification result corresponding to the input data from the model in multiple data formats, Equipped with, This is the structure it takes. Furthermore, the information processing method, which is one form of this disclosure, Information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. This is the structure it takes. Furthermore, one form of this disclosure is a program, In an information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. To execute the process This is the structure it takes. [Effects of the Invention]

[0007] This disclosure, structured as described above, facilitates the utilization of data identification results. [Brief explanation of the drawing]

[0008] [Figure 1] This is a block diagram showing an example of the configuration of the information processing device related to this disclosure. [Figure 2] This is a block diagram showing an example of the configuration of the information processing device related to this disclosure. [Figure 3]It is a flowchart showing an example of a processing operation by an information processing apparatus according to the present disclosure. [Figure 4] It is a diagram showing an example of the state of processing by an information processing apparatus according to the present disclosure. [Figure 5] It is a block diagram showing an example of the configuration of an information processing apparatus according to the present disclosure. [Figure 6] It is a block diagram showing an example of the configuration of an information processing apparatus according to the present disclosure. [Figure 7] It is a diagram showing an example of the state of processing by an information processing apparatus according to the present disclosure. <开示にかかる情報処理装置による処理の様子の一例を示す図である。 [Figure 8] It is a block diagram showing an example of the hardware configuration of an information processing apparatus according to the present disclosure. [Figure 9] It is a block diagram showing an example of the configuration of an information processing apparatus according to the present disclosure. [Figure 10] It is a flowchart showing an example of a processing operation by an information processing apparatus according to the present disclosure.

Embodiments for Carrying Out the Invention

[0009] <The First Embodiment> The first embodiment of the present disclosure will be described with reference to the drawings. Note that the drawings may be relevant to any of the embodiments. [[ID=II]]

[0010] As an example, the information processing apparatus of the present disclosure performs authenticity determination as to whether the target data, i.e., voice data, is the voice of an actual person or the voice of a non-existent person generated by a generative AI (Artificial Intelligence), that is, fake data (deepfake), and is used to output interpretation data of such authenticity determination results. In particular, the information processing apparatus outputs in a plurality of data formats (multiple modalities) such as text, voice, and image as data explaining that it has been determined to be true or false, which is interpretation data of the authenticity determination result.

[0011] As a more specific example, the information processing apparatus in the present embodiment is, for example, an information processing terminal such as a personal computer or a smartphone used by a user. In this case, the information processing apparatus performs authenticity determination on voice data that the user views on the information processing apparatus, and outputs interpretation data of the authenticity determination result. Further, when the information processing apparatus has a communication function such as a smartphone, it may perform authenticity determination on voice data of the person with whom the user is communicating, and output interpretation data of the authenticity determination result.

[0012] Note that the information processing apparatus in the present embodiment is not limited to being an information processing terminal used by a user. For example, the information processing apparatus may be an information processing apparatus used by an organization such as a police or a company, an information processing apparatus used in a call center, etc., and by being used to output interpretation data of the authenticity determination result of voice data as described above, it can be utilized, for example, for crime detection such as impersonation calls.

[0013] Hereinafter, an example of the configuration and operation of the information processing apparatus in the present embodiment will be described. The information processing apparatus is configured by one or more information processing apparatuses including an arithmetic unit and a storage unit. And, as shown in FIG. 1, the information processing apparatus includes a voice fake detection unit 11, a prompt generation unit 12, a multimodal interpretive AI unit 13, and a feedback reception unit 14. Each function of the voice fake detection unit 11, the prompt generation unit 12, the multimodal interpretive AI unit 13, and the feedback reception unit 14 can be realized by the arithmetic unit executing a program for realizing each function stored in the storage unit.

[0014] First, the information processing apparatus receives input of voice data and voice image data obtained by imaging the voice data as target data to be identified (step S1 in FIG. 3). For example, as shown in FIG. 4, it receives input of voice data composed of a time-series voice waveform and voice image data obtained by imaging a predetermined feature amount of the time-series voice.

[0015] The voice fake detection unit 11 performs a truth value determination on whether the target voice data is real data (Real) which is the voice of a real person, or fake data (Fake) which is the voice of a non-existent person generated by a generation AI or the like (step S2 in Figure 3). At this time, the voice fake detection unit 11 has been trained by machine learning using a large number of real and fake data, as well as correct answer data that represents the truth value of the truth value determination, and is configured to output the truth value determination result of the input voice data. The voice fake detection unit 11 then outputs the truth value determination result to the prompt generation unit 12. The truth value determination result of the voice data may be expressed as either true or false, or as the probability (proportion) of true or false.

[0016] The prompt generation unit 12 (input unit) generates a prompt (instruction data) that instructs the multimodal interpretability AI unit 13 to output interpretation data according to the content of the truthfulness determination result of the audio data, and inputs it to the multimodal interpretability AI unit 13 (step S3 in Figure 3). At this time, the prompt generation unit 12 generates a prompt that includes the truthfulness determination result from the audio fake detection unit 11 and instructs the unit to output explanatory interpretation data using the audio data and audio image data. Accordingly, the prompt generation unit 12 inputs the audio data and audio image data to the multimodal interpretability AI unit 13. Note that either the audio data or the audio image data may be input.

[0017] As an example, the prompt generation unit 12 generates a prompt that includes the truth value determination result of the audio data, as shown in Figure 4, and inputs it to the multimodal interpretability AI unit 13 along with the audio data and audio image data. In this embodiment, the prompt generates a prompt that includes the truth value determination result of "Fake" and requests interpretive data that explains why it was determined to be "Fake". In this embodiment, the prompt also requests an explanation using the input audio data and audio image data. The prompt generation unit 12 may also generate a prompt that explicitly instructs the unit to output interpretive data in multiple data formats, i.e., multimodally, and input it to the multimodal interpretability AI unit 13.

[0018] The multimodal interpretability AI unit 13 (output unit) is a model that analyzes the input data, and in particular, it is a multimodal model that has been trained to output responses to input prompts in multiple data formats, i.e., multimodally. In this embodiment, the multimodal interpretability AI unit 13 is configured to output interpreted data in at least two data formats (multimodal) from among data formats such as text data, audio data, and image data. Here, the multimodal interpretability AI unit 13 has been trained using input data such as prompts and audio data as described above, and training data in multiple data formats corresponding to the interpreted data. For example, as shown in Figure 2, the multimodal interpretability AI unit 13, in response to input prompts, audio data, and audio image data encoded by various encoders, generates multimodal interpreted data with a multimodal encoder, decodes it with various decoders and outputs it, calculates the loss between this output and the pre-prepared multimodal interpreted data which is the training data, and learns multimodal encoding to minimize the loss. However, the multimodal interpretability AI unit 13 may be implemented by an existing generative AI.

[0019] The multimodal interpretability AI unit 13 then outputs interpretive data, which is a multimodal explanation, in response to inputs such as prompts and audio data (step S4 in Figure 3). For example, as shown in Figure 4, it outputs interpretive data consisting of multiple data formats such as text data, audio data, and audio image data. In this case, the audio data of the interpretive data is data extracted from a part of the input audio data, and the audio image data is data in which the audio features that formed the basis for the "Fake" determination are displayed as points or rectangles on the input audio image data. In this way, the multimodal interpretability AI unit 13 may output interpretive data using the input audio data and audio image data.

[0020] The feedback receiving unit 14 (learning unit) receives feedback from users who have obtained the interpretation data output as described above, in the form of requests regarding the interpretation data. The feedback may include requests for further explanation from the user, and may be tailored to the user's level of knowledge (expertise level). For example, as shown in Figure 4, in response to text data such as "The harmonic components of the speech have distinctive features" included in the output interpretation data, feedback may include requests for further explanation such as "What is the meaning of the harmonic component features in the field of speech?"

[0021] When the feedback receiving unit 14 receives feedback from a user, it inputs the feedback itself, along with input data such as prompts that were entered when the interpretation data was output, to the multimodal interpretability AI unit 13. The multimodal interpretability AI unit 13 then learns to output interpretation data that takes the feedback into account. Specifically, the feedback receiving unit 14 prepares interpretation data that is desirable to be output as an answer to the request in the feedback, and uses this interpretation data as training data for the multimodal interpretability AI unit 13 to learn from it. For example, in response to feedback such as "What is the meaning of harmonic component features in the speech field?" in the example above, the feedback receiving unit 14 obtains explanatory data about "harmonic component features in the speech field" from a web server on the internet or a predetermined database, and uses this explanatory data as training data for the multimodal interpretability AI unit 13 to learn from it. As a result, when the same prompt as described above is input later, the multimodal interpretability AI unit 13 will output interpretation data that includes an explanation of "harmonic component features in the speech field."

[0022] As described above, the information processing device in this embodiment outputs multimodal interpretation data, allowing the user to obtain an explanation for why the audio data was judged as "Fake" in multiple data formats such as text, audio, and images, making it easier to understand the basis of the judgment. Therefore, even users with limited knowledge of the field of audio processing or the field of machine learning models used for the judgment can understand the basis of the judgment from the interpretation data in various data formats. As a result, the results of the truthfulness judgment of the audio data can be effectively utilized.

[0023] Furthermore, the voice fake detection unit 11 and the multimodal interpretability AI unit 13 described above do not necessarily have to be equipped on the information processing device. For example, the voice fake detection unit 11 and the multimodal interpretability AI unit 13 may be equipped on other information processing devices connected to the information processing device, and the information processing device may request the other information processing device to determine the authenticity of the voice data and obtain the result of the authenticity determination, or request the output of interpretive data via a prompt and obtain multimodal interpretive data.

[0024] <Second Embodiment> Next, a second embodiment of the present disclosure will be described with reference to the drawings. The drawings may be relevant to either embodiment.

[0025] The information processing device in this embodiment has the same configuration as Embodiment 1 described above. In addition, the information processing device has the following configuration. The following will mainly describe the configuration that differs from the above.

[0026] As shown in Figure 5, the information processing device in this embodiment includes an interpretability generation unit 15 in addition to the configuration described in Embodiment 1. The functions of the interpretability generation unit 15 can be realized by the arithmetic unit executing a program for realizing each function stored in the memory device.

[0027] The interpretability generation unit 15 generates interpretability data that represents an interpretation of the authenticity determination result of the audio data by the audio fake detection unit 11. The interpretability data generated by the interpretability generation unit 15 is, for example, data that explains the basis for the determination result. As an example, as shown in the "Interpretability Image" in Figure 7, when the authenticity determination result of the audio fake detection unit 11 is "Fake," the interpretability data is data in which the audio feature quantities that formed the basis for the "Fake" determination are displayed as dots on the audio image data. However, the interpretability generation unit 15 may generate interpretability data of any content.

[0028] In this embodiment, the prompt generation unit 12 (input unit) inputs the above-mentioned interpretability data to the multimodal interpretability AI unit 13, generates a prompt that instructs the output of interpretable data based on such interpretability data, and inputs it to the multimodal interpretability AI unit 13. For example, as shown in Figure 7, the prompt generation unit 12 generates a prompt that includes the truth value determination result of the audio data and uses the interpretability data, which is audio data, audio image data, and interpretable image, to request an explanation, and inputs it to the multimodal interpretability AI unit 13 along with the audio data, audio image data, and interpretable image.

[0029] Furthermore, as shown in Figure 6, the multimodal interpretability AI unit 13 (output unit) in this embodiment takes interpretable data such as interpretable images as input, in addition to prompts, feedback, audio data and audio image data, and is machine-trained using training data in multiple data formats corresponding to the interpretable data that can be output. As a result, the multimodal interpretability AI unit 13 outputs interpretable data, which is a multimodal explanation, in response to inputs such as prompts, audio data, and interpretable data such as interpretable images. As an example, as shown in Figure 7, it outputs interpretable data consisting of multiple data formats such as text data, audio data, and audio image data. In this case, the multimodal interpretability AI unit 13 may output interpretable data in at least one data format using the interpretable data which is the input interpretable image. In the example in Figure 7, interpretable data is output consisting of an interpretable image in which the features of the audio are displayed as points, and an image in which the regions in which the features appear are indicated by rectangular frames.

[0030] Furthermore, the feedback receiving unit 14 (learning unit) in this embodiment is configured in the same manner as described above. Therefore, when the feedback receiving unit 14 receives feedback from a user who has obtained the outputted interpretation data as shown in Figure 7, it inputs the feedback, along with input data such as prompts that were input when the interpretation data was output, to the multimodal interpretability AI unit 13. The multimodal interpretability AI unit 13 then learns to output interpretation data that takes the feedback into account for the input data. As a result, when the same prompt as described above is input later, the multimodal interpretability AI unit 13 will output interpretation data corresponding to the content of the feedback as described above.

[0031] As described above, according to this embodiment, an explanation of the authenticity determination result of audio data can be obtained in multiple data formats such as text, audio, and images, using pre-generated interpretability data. As a result, it becomes easier to understand the basis of the determination result, and the results of the authenticity determination of audio data can be effectively utilized.

[0032] <Third Embodiment> Next, a third embodiment of this disclosure will be described.

[0033] The information processing device described above performs authenticity determination on audio data and outputs interpretation data for the determination result. However, the data that the information processing device in this embodiment determines the authenticity of is not limited to audio data, but can be any data. For example, the data that the information processing device determines the authenticity of may be image data such as still images or moving images, or text data. Furthermore, if the information processing device is an information processing terminal used by a user or an information processing device used by an organization, it will perform authenticity determination on video data or text data that the user views or browses, or that the organization analyzes, and output interpretation data of the authenticity determination result in a multimodal manner. In this case, the interpretation data may be data that utilizes the image data or text data that is the subject of the authenticity determination.

[0034] Furthermore, the information processing device in this embodiment is not limited to determining the authenticity of target data such as audio data or image data, but may also identify and determine a pre-set type of the target data such as audio data or image data. The information processing device may then output multimodal interpretation data for the determination result of identifying the type of the target data such as audio data or image data. For example, the information processing device may identify the age group of the person who uttered the audio data (e.g., teenagers, 20s, 30s, ...) as a type, and output multimodal interpretation data for the identified age group. Alternatively, for example, if the target data is image data such as X-rays, electrocardiograms, or endoscopes in the medical field, the information processing device may identify the type of disease or lesion from such image data and output multimodal interpretation data for the identification result. In this way, it is possible to support physicians in making decisions based on image interpretation.

[0035] <Fourth Embodiment> Next, a fourth embodiment of this disclosure will be described with reference to the drawings. This embodiment shows an outline of the information processing device, etc., described in the embodiments described above. Note that the drawings may be relevant to any of the embodiments.

[0036] First, the hardware configuration of the information processing device 100 in this disclosure will be described. The information processing device 100 is composed of a general information processing device, and as an example, it is equipped with the following hardware configuration as shown in Figure 8. ·CPU(Central Processing Unit)101(Arithmetic unit) ROM (Read Only Memory) 102 (Storage Device) • RAM (Random Access Memory) 103 (Storage Device) • Program group 104 loaded into RAM 103 • Storage device 105 for storing the program group 104 • Drive device 106 for reading and writing to external storage medium 110 of the information processing device. • Communication interface 107 connecting to a communication network 111 outside the information processing device. • Input / output interface 108 for data input and output. • Bus 109 connecting each component

[0037] Figure 8 shows an example of the hardware configuration of the information processing device 100, and the hardware configuration of the information processing device is not limited to the case described above. For example, the information processing device may consist of only a part of the configuration described above, such as not having a drive device 106. In addition, the information processing device may use a GPU (Graphic Processing Unit), DSP (Digital Signal Processor), MPU (Micro Processing Unit), FPU (Floating point number Processing Unit), PPU (Physics Processing Unit), TPU (Tensor Processing Unit), quantum processor, microcontroller, or a combination thereof instead of the CPU described above.

[0038] The information processing device 100 can be equipped with the input unit 121 and output unit 122 shown in Figure 9 by having the CPU 101 acquire the program group 104 and execute it. The program group 104 is, for example, stored in advance in a storage device 105 or ROM 102, and the CPU 101 loads it into RAM 103 and executes it as needed. The program group 104 may also be supplied to the CPU 101 via a communication network 111, or it may be stored in advance in a storage medium 110, and the drive device 106 reads the program and supplies it to the CPU 101. However, the input unit 121 and output unit 122 described above may be constructed with dedicated electronic circuits to realize such means.

[0039] The input unit 121 inputs input data, which includes information based on the target data and information based on the identification result of the identification model for the target data, to the model that analyzes the data (step S101 in Figure 10). The output unit 122 outputs interpretation data from the model in multiple data formats, which represents the interpretation of the identification result according to the input data (step S102 in Figure 10).

[0040] According to the above configuration, the information processing device 100 inputs input data, including, for example, audio data or audio image data which are the target data, and a truth / false determination result which is an example of the identification result of the audio data by the identification model, to a model that analyzes the data. The information processing device then outputs interpretation data from the model, which represents an interpretation of the truth / false determination result which is an example of the identification result according to the input data, in multiple data formats such as text, audio, and images. This makes it easy to understand the basis of the identification result from the interpretation data in various data formats, and makes effective use of the results of the truth / false determination of the audio data.

[0041] Furthermore, at least one of the functions of the input unit 121 and output unit 122 described above may be performed on an information processing device installed and connected to any location on the network, that is, it may be performed using so-called cloud computing.

[0042] Furthermore, the aforementioned programs can be stored and supplied to a computer using various types of non-transitory computer-readable media. Non-transitory computer-readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (e.g., magneto-optical disks), CD-ROMs (Read Only Memory), CD-Rs, CD-R / Ws, and semiconductor memory (e.g., mask ROMs, PROMs (Programmable ROMs), EPROMs (Erasable PROMs), flash ROMs, and RAMs (Random Access Memory)). Programs may also be supplied to a computer using various types of transient computer-readable media. Examples of transient computer-readable media include electrical signals, optical signals, and electromagnetic waves. Transitory computer-readable media can be supplied to a computer via wired communication channels such as electric wires and optical fibers, or via wireless communication channels.

[0043] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure are possible, as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each of the embodiments described above can be combined with other embodiments as appropriate.

[0044] <Note> Some or all of the above embodiments may also be described as follows. The general configuration of the information processing apparatus, information processing method, and program in this disclosure is described below. However, this disclosure is not limited to the configurations described below. Furthermore, some or all of the configurations and functions described in Appendices 2 to 8, which are dependent on Appendice 1 below, may also be dependent on other Appendices 9, 10, and 11 in the same way as Appendices 2 to 8. Moreover, not limited to Appendices 1, 9, 10, and 11, some or all of the configurations and functions described as appendices may also be dependent on similar hardware, software, various recording means for recording software, or systems, without departing from the embodiments described above. (Note 1) An input unit that inputs input data, including information based on target data and information based on the identification result of the target data by the identification model, to a model that analyzes data. An output unit that outputs interpretation data representing an interpretation of the identification result corresponding to the input data from the model in multiple data formats, Equipped with an information processing device. (Note 2) The information processing device described in Appendix 1, The output unit outputs the interpretation data using information based on the target data from the model in at least one data format. Information processing device. (Note 3) The information processing device described in Appendix 2, The input unit inputs data representing the interpretation of the identification result of the target data by the identification model to the model as input data. The output unit outputs the interpretation data using information based on data representing the interpretation from the model in at least one data format. Information processing device. (Note 4) The information processing device described in Appendix 1, The input unit inputs instruction data to the model, along with the input data, which instructs the model to output the interpretation data in multiple data formats based on the input data. Information processing device. (Note 5) The information processing device described in Appendix 4, The input unit generates instruction data that instructs the output of interpretation data according to the identification content of the target data, and inputs the instruction data together with the input data to the model. Information processing device. (Note 6) The information processing device described in Appendix 1, The system includes a learning unit that inputs the aforementioned input data and user requests regarding the outputted interpretation data into the model, and performs machine learning on the model to output the interpretation data corresponding to the requests. Information processing device. (Note 7) The information processing device described in Appendix 6, The learning unit inputs training data that corresponds to the requested answer into the model, and trains the model to output interpretation data based on the training data. Information processing device. (Note 8) The information processing device described in Appendix 2, The input unit inputs the input data to the model, which includes at least one of audio data and image data obtained by visualizing the audio data, and a truth value determination result as the identification result of the audio data by the identification model. The output unit outputs the interpretation data from the model as interpretation data for the truth value determination result of the audio data, in a data format that uses at least the audio data and at least one of the image data, and text data. Information processing device. (Note 9) Information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. Information processing methods. (Note 10) In an information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. A program that executes a process. (Note 11) An input unit that inputs input data, including information based on target data and information based on the identification result of the target data by the identification model, to a model that analyzes data. An output unit that obtains interpretation data representing an interpretation of the identification result corresponding to the input data output from the model in multiple data formats, and outputs the interpretation data. Equipped with an information processing device. [Explanation of symbols]

[0045] 11. Voice fake detection unit 12 Prompt generation unit 13. Multimodal Interpretability AI Unit 14. Feedback Reception Department 15 Interpretability generator 100 Information Processing Devices 101 CPU 102 ROM 103 RAM 104 Program Groups 105 Storage device 106 Drive unit 107 Communication Interface 108 Input / Output Interfaces 109 Bus 110 Storage medium 111 Communication Network 121 Input section 122 Output section

Claims

1. An input unit that inputs input data, including information based on target data and information based on the identification result of the target data by the identification model, to a model that analyzes data. An output unit that outputs interpretation data representing an interpretation of the identification result corresponding to the input data from the model in multiple data formats, Equipped with an information processing device.

2. An information processing apparatus according to claim 1, The output unit outputs the interpretation data using information based on the target data from the model in at least one data format. Information processing device.

3. An information processing apparatus according to claim 2, The input unit inputs data representing the interpretation of the identification result of the target data by the identification model to the model as input data. The output unit outputs the interpretation data using information based on data representing the interpretation from the model in at least one data format. Information processing device.

4. An information processing apparatus according to claim 1, The input unit inputs instruction data to the model, along with the input data, which instructs the model to output the interpretation data in multiple data formats based on the input data. Information processing device.

5. An information processing apparatus according to claim 4, The input unit generates instruction data that instructs the output of interpretation data according to the identification content of the target data, and inputs the instruction data together with the input data to the model. Information processing device.

6. An information processing apparatus according to claim 1, The system includes a learning unit that inputs the aforementioned input data and user requests regarding the outputted interpretation data into the model, and performs machine learning on the model to output the interpretation data corresponding to the requests. Information processing device.

7. An information processing apparatus according to claim 6, The learning unit inputs training data that corresponds to the requested answer into the model, and trains the model to output interpretation data based on the training data. Information processing device.

8. An information processing apparatus according to claim 2, The input unit inputs the input data to the model, which includes at least one of audio data and image data obtained by imaging the audio data, and a truth value determination result as the identification result of the audio data by the identification model. The output unit outputs the interpretation data from the model as interpretation data for the truth value determination result of the audio data, in a data format that uses at least the audio data and at least one of the image data, and text data. Information processing device.

9. Information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. Information processing methods.

10. In an information processing device, Input data, which includes information based on the target data and information based on the identification result of the identification model of the target data, is input to the model that analyzes the data. The model outputs interpretation data in multiple data formats that represents an interpretation of the identification result corresponding to the input data. A program that executes a process.

Images