system

Abstract

We provide the system. [Solution] A recording device for acquiring audio data, A recognition device for converting the aforementioned audio data into text data and classifying the speaker, An analytical device for selecting topics by analyzing past conversation data and information from external sources, A recognition device that acquires image data in real time and analyzes emotions and levels of interest, A presentation device for presenting selected topics, A system that includes this.

Description

Technical Field

[0001] The technology of the present disclosure relates to a system.

Background Art

[0002] Patent Document 1 discloses a persona chatbot control method performed by at least one processor, the method including steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to an explanation of a chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance as a response to the user utterance.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In modern business communication, it is difficult to find an appropriate topic during a conversation, and especially in an international setting, smooth conversation may be hindered due to language and cultural differences. Such problems often lead to interrupted conversations, uneasiness, and discomfort. Furthermore, since it is also cumbersome to adjust the topic depending on the conversation partner, there is a need for a mechanism that automatically makes an optimal topic proposal according to individual needs.

Means for Solving the Problems

[0005] This invention includes a recognition device that acquires speech, converts the speech data into text, and automatically classifies the speaker. Furthermore, it uses an analysis device that analyzes past conversation data and the latest external information to generate optimal topics tailored to each speaker. In addition, it is possible to smoothly advance the conversation by analyzing the other party's emotions and level of interest in real time using image recognition technology and updating the topic accordingly based on the results. This provides a system that automatically provides conversation content suitable for each individual person, enabling natural communication.

[0006] "Audio data" refers to digital sound information acquired through conversation or voice input.

[0007] A "recording device" is a hardware or software configuration device for collecting and storing audio data.

[0008] A "recognition device" is a device that analyzes audio data, transcribes it into text, and identifies the speaker.

[0009] "Text data" refers to written information generated from audio data, which visually records the content of a conversation.

[0010] "Frequency response" refers to the characteristic properties of individual sounds obtained by analyzing the frequency components of a sound waveform.

[0011] An "analysis device" is a device that takes in and analyzes historical data and information from external sources to derive new insights and proposals.

[0012] "Image data" refers to still images or video information acquired in digital format.

[0013] A "presentation device" is a device used to visually or audibly display selected information to the user.

[0014] "Updating" refers to the process of improving or changing currently used information and settings based on real-time data and analysis results. [Brief explanation of the drawing]

[0015] [Figure 1] This is a conceptual diagram showing an example of the configuration of a data processing system according to the first embodiment. [Figure 2] This is a conceptual diagram showing an example of the essential functions of a data processing device and a smart device according to the first embodiment. [Figure 3] This is a conceptual diagram showing an example of the configuration of a data processing system according to the second embodiment. [Figure 4] This is a conceptual diagram showing an example of the main functions of a data processing device and smart glasses according to the second embodiment. [Figure 5] This is a conceptual diagram showing an example of the configuration of a data processing system according to the third embodiment. [Figure 6] This is a conceptual diagram showing an example of the main functions of a data processing device and a headset-type terminal according to the third embodiment. [Figure 7] This is a conceptual diagram showing an example of the configuration of a data processing system according to the fourth embodiment. [Figure 8] This is a conceptual diagram showing an example of the main functions of a data processing device and a robot according to the fourth embodiment. [Figure 9] This shows an emotion map where multiple emotions are mapped. [Figure 10] This shows an emotion map where multiple emotions are mapped. [Figure 11] This is a sequence diagram showing the processing flow of the data processing system in Example 1. [Figure 12] This is a sequence diagram showing the processing flow of the data processing system in Application Example 1. [Figure 13] This is a sequence diagram showing the processing flow of the data processing system in Example 2, which incorporates an emotion engine. [Figure 14]It is a sequence diagram showing the processing flow of a data processing system in Application Example 2 when a sentiment engine is combined.

Embodiments for Carrying Out the Invention

[0016] Hereinafter, an example of an embodiment of a system according to the technology of the present disclosure will be described with reference to the accompanying drawings.

[0017] First, the terms used in the following description will be explained.

[0018] In the following embodiments, a numbered processor (hereinafter simply referred to as "processor") may be a single arithmetic unit or a combination of multiple arithmetic units. Also, the processor may be a single type of arithmetic unit or a combination of multiple types of arithmetic units. Examples of arithmetic units include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), an APU (Accelerated Processing Unit), and the like.

[0019] In the following embodiments, a numbered RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a work memory by the processor.

[0020] In the following embodiments, a numbered storage is one or more non-volatile storage devices that store various programs and various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), or magnetic tapes, etc.

[0021] In the following embodiments, the signed communication interface (I / F) is an interface that includes a communication processor and an antenna, etc. The communication interface manages communication between multiple computers. Examples of communication standards applicable to the communication interface include wireless communication standards such as 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), or Bluetooth (registered trademark).

[0022] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." That is, "A and / or B" means that it may be A alone, or B alone, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" applies when expressing three or more things linked by "and / or."

[0023] [First Embodiment]

[0024] Figure 1 shows an example of the configuration of the data processing system 10 according to the first embodiment.

[0025] As shown in Figure 1, the data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0026] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0027] The smart device 14 comprises a computer 36, a reception device 38, an output device 40, a camera 42, and a communication interface 44. The computer 36 comprises a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The reception device 38, output device 40, and camera 42 are also connected to the bus 52.

[0028] The reception device 38 is equipped with a touch panel 38A and a microphone 38B, etc., and receives user input. The touch panel 38A receives user input by detecting contact with an object (e.g., a pen or finger). The microphone 38B receives user input by detecting the user's voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0029] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form perceptible to the user 20 (e.g., audio and / or text). The display 40A displays visible information such as text and images according to instructions from the processor 46. The speaker 40B outputs audio according to instructions from the processor 46. The camera 42 is a small digital camera equipped with an optical system such as a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0030] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various types of information between processor 46 and processor 28 via network 54.

[0031] Figure 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0032] As shown in Figure 2, in the data processing device 12, a specific processing is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" related to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 according to the specific processing program 56 executed on the RAM 30.

[0033] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0034] In the smart device 14, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The reception output program 60 is used in conjunction with a specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0035] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0036] This invention provides a system that automates the selection and provision of topics necessary for conversation, enabling users to communicate smoothly. This system consists of multiple modules that perform voice data acquisition, recognition, data analysis, topic generation, and topic updates based on real-time responses.

[0037] Regarding voice data acquisition and recognition

[0038] The device records audio at the start of a meeting or conversation and sends the data to a server. The server receives this audio data and converts it into text data using speech recognition technology. Simultaneously, it performs frequency analysis to identify individual speakers and records the statements of each identified speaker.

[0039] Data analysis and topic generation

[0040] The server collects past conversation logs and identifies each speaker's interests and concerns. Furthermore, it gathers the latest information through the internet and news APIs, selecting information highly relevant to each speaker. Based on this data, a generative AI model generates topics that are likely to interest each individual speaker.

[0041] Reaction analysis and topic presentation

[0042] The device uses its camera during the conversation to capture the other person's facial expressions and gestures. The server uses image recognition technology to evaluate the other person's emotions and level of interest in real time. Based on the speaker's responses, it assesses the appropriateness of the topic and adjusts the next topic offered as needed.

[0043] Specific example

[0044] This system is used by users participating in international conferences to support conversations with people they are meeting for the first time. The device records the conversation, and the server handles topic selection, freeing the user from the stress of choosing topics. Simultaneously, the system analyzes the other person's smiles and interesting reactions in real time, determining the next topic accordingly, ensuring a natural flow of conversation. Because the topics provided by the system adapt flexibly to the flow of the conversation, users can enjoy comfortable communication.

[0045] The following describes the processing flow.

[0046] Step 1:

[0047] The device activates the microphone at the start of a conversation and continuously records audio data. The recorded audio is sent to the server as packet data in real time.

[0048] Step 2:

[0049] The server inputs the received audio data into the speech recognition engine and converts it into text data. Simultaneously, it performs frequency analysis to extract voiceprint information and identify the speaker.

[0050] Step 3:

[0051] The server retrieves the speaker's past logs from the database. These logs record topics the speaker has shown interest in and keywords they have used to express interest, and the server uses this information to update the speaker's profile.

[0052] Step 4:

[0053] The server uses an external news API to collect the latest news data. The collected news data is analyzed and matched with speaker profile information to select highly relevant topics.

[0054] Step 5:

[0055] The server uses a generative AI model to generate topics that are likely to interest the speaker. These generated topics are then incorporated into the user's conversation in real time.

[0056] Step 6:

[0057] The device uses its camera to capture the facial expressions and gestures of the person it's talking to. The captured video data is then sent to a server.

[0058] Step 7:

[0059] The server analyzes the received video data using an image recognition engine and scores the emotions and level of interest of the person being spoken to in real time.

[0060] Step 8:

[0061] The server evaluates the effectiveness of the presented topic based on the scoring results. Based on the evaluation, it decides whether to continue with the topic or change it to a newly selected topic.

[0062] Step 9:

[0063] The user continues the conversation using the topic provided. The device prompts the user with the next topic as the conversation progresses, ensuring a smooth flow of conversation at all times.

[0064] (Example 1)

[0065] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0066] To facilitate smooth communication in conversations, it is necessary to automatically generate appropriate topics and dynamically adjust the flow of the conversation by evaluating the other person's responses in real time. Such a system would also be effective in overcoming language barriers when meeting someone for the first time or in cross-cultural exchange, but current technology does not yet offer sufficient automation or adaptability.

[0067] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0068] In this invention, the server includes acquisition means for acquiring audio signals, recognition means for converting the audio signals into strings and identifying the speaker, generation means for analyzing past dialogue history and information from external data sources to generate topics, evaluation means for acquiring video signals in real time and evaluating emotions and levels of interest, and notification means for notifying the generated topics. This enables the selection of appropriate topics that take into account the characteristics and interests of the other party, and the dynamic adjustment of conversation content based on reactions.

[0069] An "audio signal" is a representation of sound as an electrical signal, and it is the basis for acquiring and processing sound waves in digital or analog format.

[0070] "Acquisition means" refers to a device or method for collecting input data such as audio and video, and includes sensors, microphones, cameras, etc.

[0071] "Recognition means" refers to methods or devices that analyze audio signals or image data to identify specific patterns or information, and utilizes speech recognition or image recognition technologies.

[0072] "Generative means" refers to methods or devices for constructing and producing new information or content based on collected data, such as those used to automatically generate text or topics.

[0073] "Evaluation methods" refer to methods and devices for analyzing and judging qualitative information such as emotions and levels of interest from acquired data, and are particularly used for analyzing situations in real time.

[0074] "Notification means" refers to methods or devices for conveying information or results to the user, and includes providing information using screen displays, audio output, vibration, etc.

[0075] This invention is an automated system for assisting conversations, which uses audio and video signals to generate appropriate topics and dynamically adjusts the topics based on real-time responses.

[0076] Acquisition and conversion of audio signals

[0077] The device is equipped with a highly sensitive microphone, which captures the audio signal of the conversation in real time. This audio signal is transmitted to the server in digital format.

[0078] The server uses speech recognition software (e.g., a common speech recognition API) to convert the received audio signal into text. This organizes the conversation content as text data.

[0079] Speaker identification and data analysis

[0080] The server uses deep learning technology to identify speakers from the frequency characteristics of audio signals. It then analyzes stored past dialogue history to identify each speaker's interests and concerns.

[0081] When collecting the latest information from external data sources, the server uses a common data collection API.

[0082] Topic generation and adjustment

[0083] The server uses a generative AI model (e.g., a general generative AI platform) to generate topics highly relevant to each speaker. The generated topics are formatted with prompts and input into the model in the form of "Generate interesting news related to environmental issues."

[0084] Acquisition and evaluation of video signals

[0085] The camera on the device captures the video signal of the person you are talking to in real time. This allows the other person's facial expressions and gestures to be transmitted to the server.

[0086] The server uses this video signal and employs a common image recognition API to evaluate emotions and levels of interest in real time, and to confirm the appropriateness of the generated topic.

[0087] Specific example

[0088] Imagine a scenario where a user is participating in a cross-cultural exchange event. Using this system, the user can analyze the other person's smiles and surprised reactions, and then let the system choose the next topic accordingly. The system's dynamic adjustment function ensures that the conversation flows naturally and without interruption. In this way, the invention supports people's communication and provides a richer exchange experience.

[0089] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0090] Step 1:

[0091] The terminal acquires audio signals using a high-sensitivity microphone. The input is the audio signal from a conversation, which the terminal converts into a digital format before sending it to the server. Specifically, the microphone picks up the sound, and the built-in processing unit encodes it into a digital signal.

[0092] Step 2:

[0093] The server analyzes the received digital audio signal and performs speech recognition. The input is the acquired digital audio signal, and the output is a string generated based on it. The server uses a speech recognition API to convert the audio signal into text data while identifying the speaker. Specifically, the speech recognition algorithm maps the sound wave pattern to a linguistic structure.

[0094] Step 3:

[0095] The server analyzes past dialogue history using string data. The input for this step is the converted string data and past dialogue history, and the output is relevant data indicating the speaker's interests. The server uses natural language processing techniques to extract keywords from the text and identify the speaker's interests. Specifically, the text analysis engine evaluates each word and extracts the most important words.

[0096] Step 4:

[0097] The server collects information from external data sources and has a generative AI model generate topics. The input is the speaker's interest data and external data, and the output is a topic prompt sentence. By providing the prompt sentence to the generative AI model, the topics suggested by the model are obtained. Specifically, external data is collected via a RESTful API, and prompts such as "Please tell me the latest news on environmental issues" are sent to the generative AI model.

[0098] Step 5:

[0099] The terminal acquires video signals through the camera and transmits them to the server. The input is a real-time video signal, and the output is facial expression and movement data that the server uses for evaluation. The camera films the person being spoken to and sends the video to the server frame by frame in digital format.

[0100] Step 6:

[0101] The server analyzes the video signal and evaluates emotions and levels of interest. The input is real-time video data, and the output is the evaluation result of emotions and levels of interest. The server uses an image recognition API to determine emotions from facial expressions and movements. A visual information analysis algorithm analyzes image frames and performs actions to capture changes in facial expressions.

[0102] Step 7:

[0103] The server determines the appropriateness of the topic based on the evaluation results and updates the topic as needed. The output is the updated topic prompt. The server refers to the evaluation results and provides new input to the generating AI model to generate the next topic. Specifically, it feeds back the emotional data obtained in the previous step and adjusts the conversation content.

[0104] Step 8:

[0105] The terminal notifies the user of the generated topic. The input is the updated topic, and the output is the notification to the user. The terminal uses a display device and speech synthesis technology to convey the suggested topic to the user.

[0106] (Application Example 1)

[0107] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0108] To achieve smooth communication, selecting appropriate topics and adjusting conversations through real-time sentiment analysis are crucial. However, this is difficult to do with conventional systems, and there is a particular lack of technology to flexibly adapt the flow of dialogue while considering the user's emotions. Furthermore, even in customer support responses, there is a lack of technology to immediately generate and provide the most appropriate answers to user questions.

[0109] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0110] In this invention, the server includes means for converting audio information into text information and classifying the audio source, means for analyzing past conversation information and information from external sources to select a topic, and means for acquiring image information in real time and analyzing emotions and levels of interest. This enables adaptive communication based on the user's current emotions and interests.

[0111] "Audio information" refers to audio data, which is data provided through the user's voice.

[0112] "Recording means" refers to a device or system for collecting and recording audio information.

[0113] "Text information" refers to information obtained by converting audio information into written characters.

[0114] "Recognition means" refers to a system or technology that converts audio information into text information and classifies the audio source.

[0115] "Analysis means" refers to systems and technologies for analyzing past conversation information and information from external sources.

[0116] "External information sources" refer to external data providers such as the internet and news APIs.

[0117] "Image information" refers to visual data acquired in real time, including facial expressions and gestures.

[0118] "Means of delivery" refers to systems and devices used to present selected topics to users.

[0119] "Control means" refers to devices or systems that analyze the user's emotions from their facial expressions and adjust the conversation accordingly.

[0120] A system implementing this invention includes a program for acquiring voice information, converting it to text information, and controlling conversation adjustments. Specifically, a server receives voice information and converts it to text information using speech recognition software. The speech recognition technology used here is Google® Cloud Speech-to-Text, which analyzes speech with high accuracy.

[0121] Next, the server uses Azure® Machine Learning to generate user-related topics based on past conversation information and data collected from external sources. The generating AI model analyzes the data based on the prompt text and selects the most appropriate conversation content.

[0122] Meanwhile, the device acquires image information in real time and analyzes the user's emotions and interests using Amazon Rekognition. This information is interpreted by the server and used to verify whether the generated topics match the user's emotions. This process enables a flexible and engaging communication experience for the user.

[0123] For example, in customer support for an e-commerce site, it becomes possible to instantly generate the best answer to a user's question, scan the user's facial expression in real time, and provide additional information to enhance their sense of security. The following is an example of a prompt message.

[0124] "Please describe typical problems and solutions related to [product name] recently purchased by customers. Include any relevant recent news. Also, please address emotional feedback based on customer facial expressions."

[0125] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0126] Step 1:

[0127] When a user inputs voice information into a device, the device records the audio. This audio information is sent to a server. The server receives the audio information and converts the input audio data into text information using the Google Cloud Speech-to-Text API. This process yields the text data read from the audio.

[0128] Step 2:

[0129] The server analyzes the acquired text information and uses Azure Machine Learning to compare it with past conversation data and external information sources from the internet. In this analysis process, a generative AI model generates appropriate topics via prompt sentences based on the input text information, and content relevant to the user is selected. As a result, topics based on the user's interests are output.

[0130] Step 3:

[0131] While the user is operating the device, the device uses its camera in real time to acquire image information of the user. The server analyzes this image information using Amazon Rekognition and recognizes the user's emotions and level of interest by interpreting their facial expressions. Based on the emotion data obtained from the input image information, the conversation content is adjusted.

[0132] Step 4:

[0133] The server references the acquired sentiment data and generated topics to determine the optimal flow of conversation. It then presents appropriate topics to the user using the provided means. Topics are suggested to match the prompt text and sentiment data, allowing the user to experience more comfortable communication. This step completes the system's output to the user.

[0134] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0135] This invention is a system that integrates speech recognition, emotion recognition, and data analysis to facilitate smooth communication in conversations. This system analyzes the user's emotional state in real time and provides appropriate topics for conversation, thereby achieving effective communication.

[0136] About the basic configuration

[0137] The system primarily consists of a recording device for acquiring speaker voice data, a recognition device for converting voice data into text and classifying speakers, an analysis device for analyzing past conversation data and external information, an emotion engine for recognizing the user's facial expressions and tone of voice, and a presentation device for presenting selected topics. These devices work together to provide the user with an optimal conversational experience.

[0138] Regarding the acquisition and recognition of audio data

[0139] The device acquires audio data in real time via its microphone and sends it to the server. The server uses speech recognition technology to convert the audio into text data and simultaneously classify the speaker. Frequency characteristics are used for speaker classification to accurately identify each individual speaker.

[0140] About the Emotion Engine

[0141] The device uses its camera and microphone to collect the user's facial expressions and tone of voice during conversations. An emotion engine on the server analyzes this data to recognize the user's emotions in real time, and uses the results to influence the selection of topics in the conversation.

[0142] Regarding the selection and presentation of topics

[0143] The server integrates data obtained from past conversation logs and external sources with the results of the emotion engine's analysis, and generates topics that are interesting and relevant to the user through a generative AI model. The selected topics are presented to the user via the device and used as material to continue the conversation.

[0144] Specific example

[0145] For example, when a user is talking to a colleague they've just met at work, if the system's emotion engine determines the user is nervous, it will prioritize suggesting topics that will help them relax. Furthermore, if it's clear the user is interested in a particular topic, the system will provide more in-depth information on that topic. In this way, the system addresses the user's emotional needs, helping to build natural and smooth communication.

[0146] The following describes the processing flow.

[0147] Step 1:

[0148] The device activates the microphone at specific locations and times to record surrounding sounds. The recorded audio data is converted to a digital format in real time and sent to the server.

[0149] Step 2:

[0150] The server inputs the received audio data into the speech recognition engine, which converts it into text data. During this process, frequency analysis is performed simultaneously, and the speaker is identified and classified by analyzing the voiceprint.

[0151] Step 3:

[0152] The server retrieves the speaker's past conversation logs from the database. The logs contain past topics, interests, and frequently used words, and the server updates the speaker profile based on this information.

[0153] Step 4:

[0154] The server connects to an external news API to collect the latest topics and related information. This data is then matched with speaker profiles to create a list of topics that are highly relevant to the speaker.

[0155] Step 5:

[0156] The device uses its camera and microphone to capture the user's facial expressions and tone of voice during a conversation. This emotion data is then sent to the server's emotion engine.

[0157] Step 6:

[0158] The server's emotion engine analyzes the user's emotional data and determines their emotional state. The emotional results are fed back into topic selection in real time.

[0159] Step 7:

[0160] The server uses a generated AI model to integrate past logs, external information, and sentiment data to generate topics that will interest the user. These topics are then prioritized, and a refined list is created.

[0161] Step 8:

[0162] The terminal presents the user with topics from the server. If the user shows interest in a topic, the system facilitates continued conversation by delving deeper into related information.

[0163] Step 9:

[0164] The user continues the conversation based on the suggested topics. The device continues to monitor the user's emotions and the content of the conversation, and optimizes the conversation environment by suggesting new topics as needed.

[0165] (Example 2)

[0166] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart device 14 as the "terminal".

[0167] Conventional conversation support systems have struggled to achieve natural and smooth communication because they cannot accurately recognize the user's emotional state from their voice and facial expressions, and they do not adequately provide topics of interest or relevance. Therefore, there is a need for a system that can analyze the user's emotions and interests in real time and provide topics based on that analysis.

[0168] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0169] In this invention, the server includes recording means for acquiring audio data, speech recognition means for converting audio data into text data and classifying speakers, data analysis means for analyzing past conversation logs and data from external sources to generate selected topics, sentiment analysis means for acquiring facial expression data and voice features in real time and recognizing emotional states, and generative model utilization means for presenting topics that are highly relevant and interesting to the user. This makes it possible to achieve natural and effective communication by analyzing the user's emotional state and providing appropriate topics accordingly.

[0170] "Voice data" refers to information obtained by capturing a user's speech as a digital signal.

[0171] "Recording means" refers to devices or mechanisms for acquiring and storing audio data.

[0172] "Speech recognition means" refers to technologies and methods for converting speech data into text data and identifying the speaker.

[0173] "Data analysis methods" refer to techniques and methods for generating topics by analyzing information obtained from past conversation logs and external sources.

[0174] "Emotional analysis methods" refer to technologies and methods for analyzing a user's facial expressions and tone of voice in real time to determine their emotional state.

[0175] "Generative model utilization methods" refer to techniques and methods for generating and presenting relevant topics and prompts using generative models.

[0176] "Frequency response" refers to the acoustic characteristics used to analyze the properties of speech and identify the speaker.

[0177] In an embodiment of this invention, the system has a configuration that integrates multiple functions. First, the terminal is equipped with a microphone to acquire the user's voice data in real time. The terminal then transmits the acquired voice data to a server. The server uses speech recognition technology to convert the voice data into text data and identify the speaker. General speech recognition software and libraries are used here.

[0178] Furthermore, the device collects user facial expression data through its camera. This image data and audio data are combined and analyzed in real time by an emotion analysis engine on a server to evaluate the user's emotional state. In this process, the user's tone of voice and facial movements are important indicators.

[0179] The server performs data analysis based on sentiment analysis results, past conversation history, and external information sources. This analysis utilizes machine learning models to generate topics relevant to the user using an AI model. These generated topics are then presented to the user as prompts via the terminal.

[0180] Specifically, for example, when a user is discussing a new project with a colleague, the server detects the user's level of expectation and generates a prompt such as, "Have you ever brainstormed new ideas?" This prompt helps to stimulate the conversation and facilitate deeper discussion.

[0181] Such systems can accurately capture a user's emotional state and interests, and provide conversational support accordingly, thereby enabling natural and effective communication.

[0182] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0183] Step 1:

[0184] When a user speaks into the microphone, the device acquires audio data in real time. During this process, the audio signal is converted into digital data and sent to the server in a fixed buffer size. Analog audio signals are used as input, and digital audio data is obtained as output.

[0185] Step 2:

[0186] The server receives digital audio data and converts it into text data using speech recognition technology. Simultaneously, it analyzes the frequency characteristics to identify the speaker. The input is digital audio data, and the output consists of text data and speaker identification information. Speech recognition software is used in this process.

[0187] Step 3:

[0188] The device uses its camera to collect user facial expression data. This data, along with audio features, is sent to a server for sentiment analysis. An emotion engine processes this data to estimate the user's emotional state. The input consists of image data and audio feature data, and the output is the user's emotion category.

[0189] Step 4:

[0190] The server performs data analysis based on sentiment analysis results, past conversation logs, and data from external sources. This analysis utilizes machine learning algorithms and a generative AI model to determine topics highly relevant to the user. The input consists of sentiment analysis results and various data sources, and the output is a generated prompt message.

[0191] Step 5:

[0192] The terminal presents the user with a prompt message obtained from the server. This is done either as on-screen text or as voice guidance using speech synthesis. The input is the generated prompt message, and the output is the user experience provided. Through this process, the user can obtain the information and topics necessary for the next conversation.

[0193] (Application Example 2)

[0194] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart device 14 will be referred to as the "terminal."

[0195] In modern face-to-face retail, smooth communication between service staff and customers is crucial. However, instantly grasping a customer's emotions and interests and providing appropriate topics and services is not easy. Therefore, there is a need for effective customer service methods that improve customer satisfaction and increase purchasing intent.

[0196] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0197] In this invention, the server includes recording means for acquiring voice data, recognition means for converting voice data into text data and classifying speakers, analysis means for selecting topics by analyzing past conversation data and information from external sources, recognition means for acquiring image data in real time and analyzing emotions and levels of interest, presentation means for presenting selected topics, and auxiliary means aimed at analyzing the customer's emotional state in real time during customer service and presenting appropriate customer service methods or topics. This makes it possible for customer service employees to provide appropriate communication according to the customer's emotional state, thereby improving customer satisfaction and purchasing intent.

[0198] "Recording means" refers to devices for acquiring audio data, primarily equipment that uses microphones or recording devices to collect speech.

[0199] A "recognition means" is a device that converts acquired audio data into text data and has the function of identifying the speaker. It also includes the ability to analyze emotions and levels of interest from image data.

[0200] An "analysis tool" is a device that selects topics based on past conversation data and information obtained from external sources, and is primarily a device that performs data analysis and association.

[0201] A "presentation means" is a device that provides selected topics to the user and outputs information via a visual or auditory interface.

[0202] "Auxiliary tools" are devices that analyze a customer's emotional state in real time during customer service and suggest appropriate customer service methods or topics. They primarily operate in conjunction with emotion recognition technology.

[0203] Embodiments of the present invention are primarily realized as customer service support systems for physical retail stores. This system aims to enable employees to understand the customer's emotional state in real time during customer service and to communicate appropriately accordingly. Each component, including the server, terminals, and users, operates in proper coordination.

[0204] First, the terminal is equipped with a microphone to collect voice data, recording conversations during customer service in real time. The recorded voice data is sent to a server, where it is converted into text data using speech recognition technology. The server then performs speaker recognition, enabling individual customer identification.

[0205] Meanwhile, the terminal uses its camera to acquire customer facial expression data. This data is then used by an emotion recognition engine on the server to analyze the customer's emotional state and interests. The emotion engine uses the customer's voice tone and facial expressions to perform a real-time assessment of their emotions.

[0206] Next, the server analyzes past conversation history and external information, and uses a generative AI model to generate appropriate topics and customer service methods based on that analysis. These generated topics are then presented on the interface screen via the terminal, providing information to the employee.

[0207] For example, if a customer in a store shows interest in a new product but is hesitant to purchase it, the system will provide information such as, "This new product was very popular last week and is a limited production item." This allows employees to provide accurate guidance to customers and encourage them to make a purchase.

[0208] An example of a prompt would be, "Suggest topics based on the sentiment analysis results in your conversation with the customer. If the customer seems a little anxious, suggest topics that might help them feel more at ease." This allows the generative AI model to learn specific ways of responding to customers and improve the quality of customer service.

[0209] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0210] Step 1:

[0211] The terminal uses a microphone to acquire conversational audio data in real time within the physical store. The acquired audio data is transmitted from the terminal to the server. The input here is the audio signal, and the output is the audio data transmitted to the server.

[0212] Step 2:

[0213] The server converts the received audio data into text data using speech recognition software. During this process, it identifies the speaker and classifies the content of each identified conversation. The input here is audio data, and the output is text data and speaker identification information.

[0214] Step 3:

[0215] The terminal acquires images of the customer's facial expressions via its camera and sends them to the server. The input is image data, and the output is the transfer of real-time image data to the server. Image acquisition is performed to obtain a clear image focusing on a specific area of ​​the face.

[0216] Step 4:

[0217] The server integrates acquired image data and voice tone data and analyzes the customer's emotions and level of interest using an emotion recognition engine. The inputs here are image data and voice tone data, and the output is the customer's emotional state and level of interest. The data is analyzed by calculating the rate of change in facial expressions and the pitch variation of voice tone.

[0218] Step 5:

[0219] The server analyzes past conversation history and data from external sources, combines this with sentiment recognition results, and generates relevant topics via a generative AI model. The input for this step is past conversation data, sentiment analysis results, and external information, while the output is suggested topics. The server uses the generative AI model to generate topics according to the prompt.

[0220] Step 6:

[0221] The generated topics are presented to the user (employee) via a visual or audio interface through the terminal. The input here is topic suggestions, and the output is information delivered to the user as display or audio guidance. The displayed information is organized to facilitate specific customer service actions.

[0222] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0223] Data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of data generation model 58 is ChatGPT (registered trademark) (Internet search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0224] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart device 14.

[0225] [Second Embodiment]

[0226] Figure 3 shows an example of the configuration of the data processing system 210 according to the second embodiment.

[0227] As shown in Figure 3, the data processing system 210 includes a data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0228] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0229] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication interface 44. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, and camera 42 are also connected to the bus 52.

[0230] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0231] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0232] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0233] Figure 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Figure 4, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0234] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0235] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0236] In the smart glasses 214, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0237] Next, the identification processing performed by the identification processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0238] This invention provides a system that automates the selection and provision of topics necessary for conversation, enabling users to communicate smoothly. This system consists of multiple modules that perform voice data acquisition, recognition, data analysis, topic generation, and topic updates based on real-time responses.

[0239] Regarding voice data acquisition and recognition

[0240] The device records audio at the start of a meeting or conversation and sends the data to a server. The server receives this audio data and converts it into text data using speech recognition technology. Simultaneously, it performs frequency analysis to identify individual speakers and records the statements of each identified speaker.

[0241] Data analysis and topic generation

[0242] The server collects past conversation logs and identifies each speaker's interests and concerns. Furthermore, it gathers the latest information through the internet and news APIs, selecting information highly relevant to each speaker. Based on this data, a generative AI model generates topics that are likely to interest each individual speaker.

[0243] Reaction analysis and topic presentation

[0244] The device uses its camera during the conversation to capture the other person's facial expressions and gestures. The server uses image recognition technology to evaluate the other person's emotions and level of interest in real time. Based on the speaker's responses, it assesses the appropriateness of the topic and adjusts the next topic offered as needed.

[0245] Specific example

[0246] This system is used by users participating in international conferences to support conversations with people they are meeting for the first time. The device records the conversation, and the server handles topic selection, freeing the user from the stress of choosing topics. Simultaneously, the system analyzes the other person's smiles and interesting reactions in real time, determining the next topic accordingly, ensuring a natural flow of conversation. Because the topics provided by the system adapt flexibly to the flow of the conversation, users can enjoy comfortable communication.

[0247] The following describes the processing flow.

[0248] Step 1:

[0249] The device activates the microphone at the start of a conversation and continuously records audio data. The recorded audio is sent to the server as packet data in real time.

[0250] Step 2:

[0251] The server inputs the received audio data into the speech recognition engine and converts it into text data. Simultaneously, it performs frequency analysis to extract voiceprint information and identify the speaker.

[0252] Step 3:

[0253] The server retrieves the speaker's past logs from the database. These logs record topics the speaker has shown interest in and keywords they have used to express interest, and the server uses this information to update the speaker's profile.

[0254] Step 4:

[0255] The server uses an external news API to collect the latest news data. The collected news data is analyzed and matched with speaker profile information to select highly relevant topics.

[0256] Step 5:

[0257] The server uses a generative AI model to generate topics that are likely to interest the speaker. These generated topics are then incorporated into the user's conversation in real time.

[0258] Step 6:

[0259] The device uses its camera to capture the facial expressions and gestures of the person it's talking to. The captured video data is then sent to a server.

[0260] Step 7:

[0261] The server analyzes the received video data using an image recognition engine and scores the emotions and level of interest of the person being spoken to in real time.

[0262] Step 8:

[0263] The server evaluates the effectiveness of the presented topic based on the scoring results. Based on the evaluation, it decides whether to continue with the topic or change it to a newly selected topic.

[0264] Step 9:

[0265] The user continues the conversation using the topic provided. The device prompts the user with the next topic as the conversation progresses, ensuring a smooth flow of conversation at all times.

[0266] (Example 1)

[0267] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0268] To facilitate smooth communication in conversations, it is necessary to automatically generate appropriate topics and dynamically adjust the flow of the conversation by evaluating the other person's responses in real time. Such a system would also be effective in overcoming language barriers when meeting someone for the first time or in cross-cultural exchange, but current technology does not yet offer sufficient automation or adaptability.

[0269] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0270] In this invention, the server includes acquisition means for acquiring audio signals, recognition means for converting the audio signals into strings and identifying the speaker, generation means for analyzing past dialogue history and information from external data sources to generate topics, evaluation means for acquiring video signals in real time and evaluating emotions and levels of interest, and notification means for notifying the generated topics. This enables the selection of appropriate topics that take into account the characteristics and interests of the other party, and the dynamic adjustment of conversation content based on reactions.

[0271] An "audio signal" is a representation of sound as an electrical signal, and it is the basis for acquiring and processing sound waves in digital or analog format.

[0272] "Acquisition means" refers to a device or method for collecting input data such as audio and video, and includes sensors, microphones, cameras, etc.

[0273] "Recognition means" refers to methods or devices that analyze audio signals or image data to identify specific patterns or information, and utilizes speech recognition or image recognition technologies.

[0274] "Generative means" refers to methods or devices for constructing and producing new information or content based on collected data, such as those used to automatically generate text or topics.

[0275] "Evaluation methods" refer to methods and devices for analyzing and judging qualitative information such as emotions and levels of interest from acquired data, and are particularly used for analyzing situations in real time.

[0276] "Notification means" refers to methods or devices for conveying information or results to the user, and includes providing information using screen displays, audio output, vibration, etc.

[0277] This invention is an automated system for assisting conversations, which uses audio and video signals to generate appropriate topics and dynamically adjusts the topics based on real-time responses.

[0278] Acquisition and conversion of audio signals

[0279] The device is equipped with a highly sensitive microphone, which captures the audio signal of the conversation in real time. This audio signal is transmitted to the server in digital format.

[0280] The server uses speech recognition software (e.g., a general speech recognition API) to convert the received speech signal into a string of characters. As a result, the conversation content is organized as text data.

[0281] Speaker identification and data analysis

[0282] The server uses deep learning technology to identify the speaker from the frequency characteristics of the speech signal. Subsequently, it analyzes the stored past conversation history to identify the interests and concerns of each speaker.

[0283] When collecting the latest information from an external data source, the server uses a general information collection API.

[0284] Topic generation and adjustment

[0285] The server uses a generative AI model (e.g., a general generative AI platform) to generate topics highly relevant to each speaker. The generated topics are formalized by a prompt sentence and input into the model in a form such as "Please generate interesting news related to environmental issues."

[0286] Acquisition and evaluation of video signals

[0287] The camera installed on the terminal acquires the video signal of the person during the conversation in real time. As a result, the expression and gesture of the person are transmitted to the server.

[0288] The server uses this video signal to evaluate the emotion and degree of interest in real time using a general image recognition API, and to confirm the appropriateness of the generated topic.

[0289] Specific example

[0290] Imagine a scenario where a user is participating in a cross-cultural exchange event. Using this system, the user can analyze the other person's smiles and surprised reactions, and then let the system choose the next topic accordingly. The system's dynamic adjustment function ensures that the conversation flows naturally and without interruption. In this way, the invention supports people's communication and provides a richer exchange experience.

[0291] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0292] Step 1:

[0293] The terminal acquires audio signals using a high-sensitivity microphone. The input is the audio signal from a conversation, which the terminal converts into a digital format before sending it to the server. Specifically, the microphone picks up the sound, and the built-in processing unit encodes it into a digital signal.

[0294] Step 2:

[0295] The server analyzes the received digital audio signal and performs speech recognition. The input is the acquired digital audio signal, and the output is a string generated based on it. The server uses a speech recognition API to convert the audio signal into text data while identifying the speaker. Specifically, the speech recognition algorithm maps the sound wave pattern to a linguistic structure.

[0296] Step 3:

[0297] The server analyzes past dialogue history using string data. The input for this step is the converted string data and past dialogue history, and the output is relevant data indicating the speaker's interests. The server uses natural language processing techniques to extract keywords from the text and identify the speaker's interests. Specifically, the text analysis engine evaluates each word and extracts the most important words.

[0298] Step 4:

[0299] The server collects information from external data sources and has a generative AI model generate topics. The input is the speaker's interest data and external data, and the output is a topic prompt sentence. By providing the prompt sentence to the generative AI model, the topics suggested by the model are obtained. Specifically, external data is collected via a RESTful API, and prompts such as "Please tell me the latest news on environmental issues" are sent to the generative AI model.

[0300] Step 5:

[0301] The terminal acquires video signals through the camera and transmits them to the server. The input is a real-time video signal, and the output is facial expression and movement data that the server uses for evaluation. The camera films the person being spoken to and sends the video to the server frame by frame in digital format.

[0302] Step 6:

[0303] The server analyzes the video signal and evaluates emotions and levels of interest. The input is real-time video data, and the output is the evaluation result of emotions and levels of interest. The server uses an image recognition API to determine emotions from facial expressions and movements. A visual information analysis algorithm analyzes image frames and performs actions to capture changes in facial expressions.

[0304] Step 7:

[0305] The server determines the appropriateness of the topic based on the evaluation results and updates the topic as needed. The output is the updated topic prompt. The server refers to the evaluation results and provides new input to the generating AI model to generate the next topic. Specifically, it feeds back the emotional data obtained in the previous step and adjusts the conversation content.

[0306] Step 8:

[0307] The terminal notifies the user of the generated topic. The input is the updated topic, and the output is the notification to the user. The terminal utilizes a display device and voice synthesis technology to perform the operation of conveying the proposed topic to the user.

[0308] (Application Example 1)

[0309] Next, Application Example 1 will be described. In the following description, the data processing device 12 is referred to as the "server", and the smart glasses 214 are referred to as the "terminal".

[0310] To achieve smooth communication, it is important to select appropriate topics and adjust conversations through real-time sentiment analysis. However, it is difficult to perform this with conventional systems, and there is a lack of technology to flexibly adapt the flow of conversations considering the user's sentiment. Furthermore, in customer support responses, there is also a lack of technology to immediately generate and provide the optimal answer to the user's questions.

[0311] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0312] In this invention, the server includes means for converting voice information into text information and classifying voice sources, means for analyzing past conversation information and information from external information sources to select topics, and means for acquiring image information in real time and analyzing sentiment and degree of interest. Thereby, adaptable communication based on the user's current sentiment and interest becomes possible.

[0313] "Voice information" refers to voice data, which is data provided by the user's voice.

[0314] "Recording means" refers to a device or system for collecting and recording voice information.

[0315] "Text information" is information obtained by converting voice information into characters.

[0316] "Recognition means" refers to a system or technology that converts audio information into text information and classifies the audio source.

[0317] "Analysis means" refers to systems and technologies for analyzing past conversation information and information from external sources.

[0318] "External information sources" refer to external data providers such as the internet and news APIs.

[0319] "Image information" refers to visual data acquired in real time, including facial expressions and gestures.

[0320] "Means of delivery" refers to systems and devices used to present selected topics to users.

[0321] "Control means" refers to devices or systems that analyze the user's emotions from their facial expressions and adjust the conversation accordingly.

[0322] A system implementing this invention includes a program for acquiring voice information, converting it to text information, and controlling conversation adjustments. Specifically, a server receives voice information and converts it to text information using speech recognition software. The speech recognition technology used here is Google Cloud Speech-to-Text, which analyzes voice with high accuracy.

[0323] Next, the server uses Azure Machine Learning to generate user-related topics based on past conversation information and data collected from external sources. The generating AI model analyzes the data based on the prompt text and selects the most appropriate conversation content.

[0324] Meanwhile, the device acquires image information in real time and analyzes the user's emotions and interests using Amazon Rekognition. This information is interpreted by the server and used to verify whether the generated topics match the user's emotions. This process enables a flexible and engaging communication experience for the user.

[0325] For example, in customer support for an e-commerce site, it becomes possible to instantly generate the best answer to a user's question, scan the user's facial expression in real time, and provide additional information to enhance their sense of security. The following is an example of a prompt message.

[0326] "Please describe typical problems and solutions related to [product name] recently purchased by customers. Include any relevant recent news. Also, please address emotional feedback based on customer facial expressions."

[0327] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0328] Step 1:

[0329] When a user inputs voice information into a device, the device records the audio. This audio information is sent to a server. The server receives the audio information and converts the input audio data into text information using the Google Cloud Speech-to-Text API. This process yields the text data read from the audio.

[0330] Step 2:

[0331] The server analyzes the acquired text information and uses Azure Machine Learning to compare it with past conversation data and external information sources from the internet. In this analysis process, a generative AI model generates appropriate topics via prompt sentences based on the input text information, and content relevant to the user is selected. As a result, topics based on the user's interests are output.

[0332] Step 3:

[0333] While the user is operating the device, the device uses its camera in real time to acquire image information of the user. The server analyzes this image information using Amazon Rekognition and recognizes the user's emotions and level of interest by interpreting their facial expressions. Based on the emotion data obtained from the input image information, the conversation content is adjusted.

[0334] Step 4:

[0335] The server references the acquired sentiment data and generated topics to determine the optimal flow of conversation. It then presents appropriate topics to the user using the provided means. Topics are suggested to match the prompt text and sentiment data, allowing the user to experience more comfortable communication. This step completes the system's output to the user.

[0336] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0337] This invention is a system that integrates speech recognition, emotion recognition, and data analysis to facilitate smooth communication in conversations. This system analyzes the user's emotional state in real time and provides appropriate topics for conversation, thereby achieving effective communication.

[0338] About the basic configuration

[0339] The system primarily consists of a recording device for acquiring speaker voice data, a recognition device for converting voice data into text and classifying speakers, an analysis device for analyzing past conversation data and external information, an emotion engine for recognizing the user's facial expressions and tone of voice, and a presentation device for presenting selected topics. These devices work together to provide the user with an optimal conversational experience.

[0340] Regarding the acquisition and recognition of audio data

[0341] The device acquires audio data in real time via its microphone and sends it to the server. The server uses speech recognition technology to convert the audio into text data and simultaneously classify the speaker. Frequency characteristics are used for speaker classification to accurately identify each individual speaker.

[0342] About the Emotion Engine

[0343] The device uses its camera and microphone to collect the user's facial expressions and tone of voice during conversations. An emotion engine on the server analyzes this data to recognize the user's emotions in real time, and uses the results to influence the selection of topics in the conversation.

[0344] Regarding the selection and presentation of topics

[0345] The server integrates data obtained from past conversation logs and external sources with the results of the emotion engine's analysis, and generates topics that are interesting and relevant to the user through a generative AI model. The selected topics are presented to the user via the device and used as material to continue the conversation.

[0346] Specific example

[0347] For example, when a user is talking to a colleague they've just met at work, if the system's emotion engine determines the user is nervous, it will prioritize suggesting topics that will help them relax. Furthermore, if it's clear the user is interested in a particular topic, the system will provide more in-depth information on that topic. In this way, the system addresses the user's emotional needs, helping to build natural and smooth communication.

[0348] The following describes the processing flow.

[0349] Step 1:

[0350] The device activates the microphone at specific locations and times to record surrounding sounds. The recorded audio data is converted to a digital format in real time and sent to the server.

[0351] Step 2:

[0352] The server inputs the received audio data into the speech recognition engine, which converts it into text data. During this process, frequency analysis is performed simultaneously, and the speaker is identified and classified by analyzing the voiceprint.

[0353] Step 3:

[0354] The server retrieves the speaker's past conversation logs from the database. The logs contain past topics, interests, and frequently used words, and the server updates the speaker profile based on this information.

[0355] Step 4:

[0356] The server connects to an external news API to collect the latest topics and related information. This data is then matched with speaker profiles to create a list of topics that are highly relevant to the speaker.

[0357] Step 5:

[0358] The device uses its camera and microphone to capture the user's facial expressions and tone of voice during a conversation. This emotion data is then sent to the server's emotion engine.

[0359] Step 6:

[0360] The server's emotion engine analyzes the user's emotional data and determines their emotional state. The emotional results are fed back into topic selection in real time.

[0361] Step 7:

[0362] The server uses a generated AI model to integrate past logs, external information, and sentiment data to generate topics that will interest the user. These topics are then prioritized, and a refined list is created.

[0363] Step 8:

[0364] The terminal presents the user with topics from the server. If the user shows interest in a topic, the system facilitates continued conversation by delving deeper into related information.

[0365] Step 9:

[0366] The user continues the conversation based on the suggested topics. The device continues to monitor the user's emotions and the content of the conversation, and optimizes the conversation environment by suggesting new topics as needed.

[0367] (Example 2)

[0368] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal".

[0369] Conventional conversation support systems have struggled to achieve natural and smooth communication because they cannot accurately recognize the user's emotional state from their voice and facial expressions, and they do not adequately provide topics of interest or relevance. Therefore, there is a need for a system that can analyze the user's emotions and interests in real time and provide topics based on that analysis.

[0370] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0371] In this invention, the server includes recording means for acquiring audio data, speech recognition means for converting audio data into text data and classifying speakers, data analysis means for analyzing past conversation logs and data from external sources to generate selected topics, sentiment analysis means for acquiring facial expression data and voice features in real time and recognizing emotional states, and generative model utilization means for presenting topics that are highly relevant and interesting to the user. This makes it possible to achieve natural and effective communication by analyzing the user's emotional state and providing appropriate topics accordingly.

[0372] "Voice data" refers to information obtained by capturing a user's speech as a digital signal.

[0373] "Recording means" refers to devices or mechanisms for acquiring and storing audio data.

[0374] "Speech recognition means" refers to technologies and methods for converting speech data into text data and identifying the speaker.

[0375] "Data analysis methods" refer to techniques and methods for generating topics by analyzing information obtained from past conversation logs and external sources.

[0376] "Emotional analysis methods" refer to technologies and methods for analyzing a user's facial expressions and tone of voice in real time to determine their emotional state.

[0377] "Generative model utilization methods" refer to techniques and methods for generating and presenting relevant topics and prompts using generative models.

[0378] "Frequency response" refers to the acoustic characteristics used to analyze the properties of speech and identify the speaker.

[0379] In an embodiment of this invention, the system has a configuration that integrates multiple functions. First, the terminal is equipped with a microphone to acquire the user's voice data in real time. The terminal then transmits the acquired voice data to a server. The server uses speech recognition technology to convert the voice data into text data and identify the speaker. General speech recognition software and libraries are used here.

[0380] Furthermore, the device collects user facial expression data through its camera. This image data and audio data are combined and analyzed in real time by an emotion analysis engine on a server to evaluate the user's emotional state. In this process, the user's tone of voice and facial movements are important indicators.

[0381] The server performs data analysis based on sentiment analysis results, past conversation history, and external information sources. This analysis utilizes machine learning models to generate topics relevant to the user using an AI model. These generated topics are then presented to the user as prompts via the terminal.

[0382] Specifically, for example, when a user is discussing a new project with a colleague, the server detects the user's level of expectation and generates a prompt such as, "Have you ever brainstormed new ideas?" This prompt helps to stimulate the conversation and facilitate deeper discussion.

[0383] Such systems can accurately capture a user's emotional state and interests, and provide conversational support accordingly, thereby enabling natural and effective communication.

[0384] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0385] Step 1:

[0386] When a user speaks into the microphone, the device acquires audio data in real time. During this process, the audio signal is converted into digital data and sent to the server in a fixed buffer size. Analog audio signals are used as input, and digital audio data is obtained as output.

[0387] Step 2:

[0388] The server receives digital audio data and converts it into text data using speech recognition technology. Simultaneously, it analyzes the frequency characteristics to identify the speaker. The input is digital audio data, and the output consists of text data and speaker identification information. Speech recognition software is used in this process.

[0389] Step 3:

[0390] The device uses its camera to collect user facial expression data. This data, along with audio features, is sent to a server for sentiment analysis. An emotion engine processes this data to estimate the user's emotional state. The input consists of image data and audio feature data, and the output is the user's emotion category.

[0391] Step 4:

[0392] The server performs data analysis based on sentiment analysis results, past conversation logs, and data from external sources. This analysis utilizes machine learning algorithms and a generative AI model to determine topics highly relevant to the user. The input consists of sentiment analysis results and various data sources, and the output is a generated prompt message.

[0393] Step 5:

[0394] The terminal presents the user with a prompt message obtained from the server. This is done either as on-screen text or as voice guidance using speech synthesis. The input is the generated prompt message, and the output is the user experience provided. Through this process, the user can obtain the information and topics necessary for the next conversation.

[0395] (Application Example 2)

[0396] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the smart glasses 214 will be referred to as the "terminal."

[0397] In modern face-to-face retail, smooth communication between service staff and customers is crucial. However, instantly grasping a customer's emotions and interests and providing appropriate topics and services is not easy. Therefore, there is a need for effective customer service methods that improve customer satisfaction and increase purchasing intent.

[0398] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0399] In this invention, the server includes recording means for acquiring voice data, recognition means for converting voice data into text data and classifying speakers, analysis means for selecting topics by analyzing past conversation data and information from external sources, recognition means for acquiring image data in real time and analyzing emotions and levels of interest, presentation means for presenting selected topics, and auxiliary means aimed at analyzing the customer's emotional state in real time during customer service and presenting appropriate customer service methods or topics. This makes it possible for customer service employees to provide appropriate communication according to the customer's emotional state, thereby improving customer satisfaction and purchasing intent.

[0400] "Recording means" refers to devices for acquiring audio data, primarily equipment that uses microphones or recording devices to collect speech.

[0401] A "recognition means" is a device that converts acquired audio data into text data and has the function of identifying the speaker. It also includes the ability to analyze emotions and levels of interest from image data.

[0402] An "analysis tool" is a device that selects topics based on past conversation data and information obtained from external sources, and is primarily a device that performs data analysis and association.

[0403] A "presentation means" is a device that provides selected topics to the user and outputs information via a visual or auditory interface.

[0404] "Auxiliary tools" are devices that analyze a customer's emotional state in real time during customer service and suggest appropriate customer service methods or topics. They primarily operate in conjunction with emotion recognition technology.

[0405] Embodiments of the present invention are primarily realized as customer service support systems for physical retail stores. This system aims to enable employees to understand the customer's emotional state in real time during customer service and to communicate appropriately accordingly. Each component, including the server, terminals, and users, operates in proper coordination.

[0406] First, the terminal is equipped with a microphone to collect voice data, recording conversations during customer service in real time. The recorded voice data is sent to a server, where it is converted into text data using speech recognition technology. The server then performs speaker recognition, enabling individual customer identification.

[0407] Meanwhile, the terminal uses its camera to acquire customer facial expression data. This data is then used by an emotion recognition engine on the server to analyze the customer's emotional state and interests. The emotion engine uses the customer's voice tone and facial expressions to perform a real-time assessment of their emotions.

[0408] Next, the server analyzes past conversation history and external information, and uses a generative AI model to generate appropriate topics and customer service methods based on that analysis. These generated topics are then presented on the interface screen via the terminal, providing information to the employee.

[0409] For example, if a customer in a store shows interest in a new product but is hesitant to purchase it, the system will provide information such as, "This new product was very popular last week and is a limited production item." This allows employees to provide accurate guidance to customers and encourage them to make a purchase.

[0410] An example of a prompt would be, "Suggest topics based on the sentiment analysis results in your conversation with the customer. If the customer seems a little anxious, suggest topics that might help them feel more at ease." This allows the generative AI model to learn specific ways of responding to customers and improve the quality of customer service.

[0411] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0412] Step 1:

[0413] The terminal uses a microphone to acquire conversational audio data in real time within the physical store. The acquired audio data is transmitted from the terminal to the server. The input here is the audio signal, and the output is the audio data transmitted to the server.

[0414] Step 2:

[0415] The server converts the received audio data into text data using speech recognition software. During this process, it identifies the speaker and classifies the content of each identified conversation. The input here is audio data, and the output is text data and speaker identification information.

[0416] Step 3:

[0417] The terminal acquires images of the customer's facial expressions via its camera and sends them to the server. The input is image data, and the output is the transfer of real-time image data to the server. Image acquisition is performed to obtain a clear image focusing on a specific area of ​​the face.

[0418] Step 4:

[0419] The server integrates acquired image data and voice tone data and analyzes the customer's emotions and level of interest using an emotion recognition engine. The inputs here are image data and voice tone data, and the output is the customer's emotional state and level of interest. The data is analyzed by calculating the rate of change in facial expressions and the pitch variation of voice tone.

[0420] Step 5:

[0421] The server analyzes past conversation history and data from external sources, combines this with sentiment recognition results, and generates relevant topics via a generative AI model. The input for this step is past conversation data, sentiment analysis results, and external information, while the output is suggested topics. The server uses the generative AI model to generate topics according to the prompt.

[0422] Step 6:

[0423] The generated topics are presented to the user (employee) via a visual or audio interface through the terminal. The input here is topic suggestions, and the output is information delivered to the user as display or audio guidance. The displayed information is organized to facilitate specific customer service actions.

[0424] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0425] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0426] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the smart glasses 214.

[0427] [Third Embodiment]

[0428] Figure 5 shows an example of the configuration of the data processing system 310 according to the third embodiment.

[0429] As shown in Figure 5, the data processing system 310 includes a data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0430] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0431] The headset terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a display 343. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and display 343 are also connected to the bus 52.

[0432] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0433] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0434] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0435] Figure 6 shows an example of the main functions of the data processing device 12 and the headset terminal 314. As shown in Figure 6, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0436] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0437] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0438] In the headset terminal 314, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0439] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the headset terminal 314 will be referred to as the "terminal".

[0440] This invention provides a system that automates the selection and provision of topics necessary for conversation, enabling users to communicate smoothly. This system consists of multiple modules that perform voice data acquisition, recognition, data analysis, topic generation, and topic updates based on real-time responses.

[0441] Regarding voice data acquisition and recognition

[0442] The device records audio at the start of a meeting or conversation and sends the data to a server. The server receives this audio data and converts it into text data using speech recognition technology. Simultaneously, it performs frequency analysis to identify individual speakers and records the statements of each identified speaker.

[0443] Data analysis and topic generation

[0444] The server collects past conversation logs and identifies each speaker's interests and concerns. Furthermore, it gathers the latest information through the internet and news APIs, selecting information highly relevant to each speaker. Based on this data, a generative AI model generates topics that are likely to interest each individual speaker.

[0445] Reaction analysis and topic presentation

[0446] The device uses its camera during the conversation to capture the other person's facial expressions and gestures. The server uses image recognition technology to evaluate the other person's emotions and level of interest in real time. Based on the speaker's responses, it assesses the appropriateness of the topic and adjusts the next topic offered as needed.

[0447] Specific example

[0448] This system is used by users participating in international conferences to support conversations with people they are meeting for the first time. The device records the conversation, and the server handles topic selection, freeing the user from the stress of choosing topics. Simultaneously, the system analyzes the other person's smiles and interesting reactions in real time, determining the next topic accordingly, ensuring a natural flow of conversation. Because the topics provided by the system adapt flexibly to the flow of the conversation, users can enjoy comfortable communication.

[0449] The following describes the processing flow.

[0450] Step 1:

[0451] The device activates the microphone at the start of a conversation and continuously records audio data. The recorded audio is sent to the server as packet data in real time.

[0452] Step 2:

[0453] The server inputs the received audio data into the speech recognition engine and converts it into text data. Simultaneously, it performs frequency analysis to extract voiceprint information and identify the speaker.

[0454] Step 3:

[0455] The server retrieves the speaker's past logs from the database. These logs record topics the speaker has shown interest in and keywords they have used to express interest, and the server uses this information to update the speaker's profile.

[0456] Step 4:

[0457] The server uses an external news API to collect the latest news data. The collected news data is analyzed and matched with speaker profile information to select highly relevant topics.

[0458] Step 5:

[0459] The server uses a generative AI model to generate topics that are likely to interest the speaker. These generated topics are then incorporated into the user's conversation in real time.

[0460] Step 6:

[0461] The device uses its camera to capture the facial expressions and gestures of the person it's talking to. The captured video data is then sent to a server.

[0462] Step 7:

[0463] The server analyzes the received video data using an image recognition engine and scores the emotions and level of interest of the person being spoken to in real time.

[0464] Step 8:

[0465] The server evaluates the effectiveness of the presented topic based on the scoring results. Based on the evaluation, it decides whether to continue with the topic or change it to a newly selected topic.

[0466] Step 9:

[0467] The user continues the conversation using the topic provided. The device prompts the user with the next topic as the conversation progresses, ensuring a smooth flow of conversation at all times.

[0468] (Example 1)

[0469] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0470] To facilitate smooth communication in conversations, it is necessary to automatically generate appropriate topics and dynamically adjust the flow of the conversation by evaluating the other person's responses in real time. Such a system would also be effective in overcoming language barriers when meeting someone for the first time or in cross-cultural exchange, but current technology does not yet offer sufficient automation or adaptability.

[0471] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0472] In this invention, the server includes acquisition means for acquiring audio signals, recognition means for converting the audio signals into strings and identifying the speaker, generation means for analyzing past dialogue history and information from external data sources to generate topics, evaluation means for acquiring video signals in real time and evaluating emotions and levels of interest, and notification means for notifying the generated topics. This enables the selection of appropriate topics that take into account the characteristics and interests of the other party, and the dynamic adjustment of conversation content based on reactions.

[0473] An "audio signal" is a representation of sound as an electrical signal, and it is the basis for acquiring and processing sound waves in digital or analog format.

[0474] "Acquisition means" refers to a device or method for collecting input data such as audio and video, and includes sensors, microphones, cameras, etc.

[0475] "Recognition means" refers to methods or devices that analyze audio signals or image data to identify specific patterns or information, and utilizes speech recognition or image recognition technologies.

[0476] "Generative means" refers to methods or devices for constructing and producing new information or content based on collected data, such as those used to automatically generate text or topics.

[0477] "Evaluation methods" refer to methods and devices for analyzing and judging qualitative information such as emotions and levels of interest from acquired data, and are particularly used for analyzing situations in real time.

[0478] "Notification means" refers to methods or devices for conveying information or results to the user, and includes providing information using screen displays, audio output, vibration, etc.

[0479] This invention is an automated system for assisting conversations, which uses audio and video signals to generate appropriate topics and dynamically adjusts the topics based on real-time responses.

[0480] Acquisition and conversion of audio signals

[0481] The device is equipped with a highly sensitive microphone, which captures the audio signal of the conversation in real time. This audio signal is transmitted to the server in digital format.

[0482] The server uses speech recognition software (e.g., a common speech recognition API) to convert the received audio signal into text. This organizes the conversation content as text data.

[0483] Speaker identification and data analysis

[0484] The server uses deep learning technology to identify speakers from the frequency characteristics of audio signals. It then analyzes stored past dialogue history to identify each speaker's interests and concerns.

[0485] When collecting the latest information from external data sources, the server uses a common data collection API.

[0486] Topic generation and adjustment

[0487] The server uses a generative AI model (e.g., a general generative AI platform) to generate topics highly relevant to each speaker. The generated topics are formatted with prompts and input into the model in the form of "Generate interesting news related to environmental issues."

[0488] Acquisition and evaluation of video signals

[0489] The camera on the device captures the video signal of the person you are talking to in real time. This allows the other person's facial expressions and gestures to be transmitted to the server.

[0490] The server uses this video signal and employs a common image recognition API to evaluate emotions and levels of interest in real time, and to confirm the appropriateness of the generated topic.

[0491] Specific example

[0492] Imagine a scenario where a user is participating in a cross-cultural exchange event. Using this system, the user can analyze the other person's smiles and surprised reactions, and then let the system choose the next topic accordingly. The system's dynamic adjustment function ensures that the conversation flows naturally and without interruption. In this way, the invention supports people's communication and provides a richer exchange experience.

[0493] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0494] Step 1:

[0495] The terminal acquires audio signals using a high-sensitivity microphone. The input is the audio signal from a conversation, which the terminal converts into a digital format before sending it to the server. Specifically, the microphone picks up the sound, and the built-in processing unit encodes it into a digital signal.

[0496] Step 2:

[0497] The server analyzes the received digital audio signal and performs speech recognition. The input is the acquired digital audio signal, and the output is a string generated based on it. The server uses a speech recognition API to convert the audio signal into text data while identifying the speaker. Specifically, the speech recognition algorithm maps the sound wave pattern to a linguistic structure.

[0498] Step 3:

[0499] The server analyzes past dialogue history using string data. The input for this step is the converted string data and past dialogue history, and the output is relevant data indicating the speaker's interests. The server uses natural language processing techniques to extract keywords from the text and identify the speaker's interests. Specifically, the text analysis engine evaluates each word and extracts the most important words.

[0500] Step 4:

[0501] The server collects information from external data sources and has a generative AI model generate topics. The input is the speaker's interest data and external data, and the output is a topic prompt sentence. By providing the prompt sentence to the generative AI model, the topics suggested by the model are obtained. Specifically, external data is collected via a RESTful API, and prompts such as "Please tell me the latest news on environmental issues" are sent to the generative AI model.

[0502] Step 5:

[0503] The terminal acquires video signals through the camera and transmits them to the server. The input is a real-time video signal, and the output is facial expression and movement data that the server uses for evaluation. The camera films the person being spoken to and sends the video to the server frame by frame in digital format.

[0504] Step 6:

[0505] The server analyzes the video signal and evaluates emotions and levels of interest. The input is real-time video data, and the output is the evaluation result of emotions and levels of interest. The server uses an image recognition API to determine emotions from facial expressions and movements. A visual information analysis algorithm analyzes image frames and performs actions to capture changes in facial expressions.

[0506] Step 7:

[0507] The server determines the appropriateness of the topic based on the evaluation results and updates the topic as needed. The output is the updated topic prompt. The server refers to the evaluation results and provides new input to the generating AI model to generate the next topic. Specifically, it feeds back the emotional data obtained in the previous step and adjusts the conversation content.

[0508] Step 8:

[0509] The terminal notifies the user of the generated topic. The input is the updated topic, and the output is the notification to the user. The terminal uses a display device and speech synthesis technology to convey the suggested topic to the user.

[0510] (Application Example 1)

[0511] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0512] To achieve smooth communication, selecting appropriate topics and adjusting conversations through real-time sentiment analysis are crucial. However, this is difficult to do with conventional systems, and there is a particular lack of technology to flexibly adapt the flow of dialogue while considering the user's emotions. Furthermore, even in customer support responses, there is a lack of technology to immediately generate and provide the most appropriate answers to user questions.

[0513] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0514] In this invention, the server includes means for converting audio information into text information and classifying the audio source, means for analyzing past conversation information and information from external sources to select a topic, and means for acquiring image information in real time and analyzing emotions and levels of interest. This enables adaptive communication based on the user's current emotions and interests.

[0515] "Audio information" refers to audio data, which is data provided through the user's voice.

[0516] "Recording means" refers to a device or system for collecting and recording audio information.

[0517] "Text information" refers to information obtained by converting audio information into written characters.

[0518] "Recognition means" refers to a system or technology that converts audio information into text information and classifies the audio source.

[0519] "Analysis means" refers to systems and technologies for analyzing past conversation information and information from external sources.

[0520] "External information sources" refer to external data providers such as the internet and news APIs.

[0521] "Image information" refers to visual data acquired in real time, including facial expressions and gestures.

[0522] "Means of delivery" refers to systems and devices used to present selected topics to users.

[0523] "Control means" refers to devices or systems that analyze the user's emotions from their facial expressions and adjust the conversation accordingly.

[0524] A system implementing this invention includes a program for acquiring voice information, converting it to text information, and controlling conversation adjustments. Specifically, a server receives voice information and converts it to text information using speech recognition software. The speech recognition technology used here is Google Cloud Speech-to-Text, which analyzes voice with high accuracy.

[0525] Next, the server uses Azure Machine Learning to generate user-related topics based on past conversation information and data collected from external sources. The generating AI model analyzes the data based on the prompt text and selects the most appropriate conversation content.

[0526] Meanwhile, the device acquires image information in real time and analyzes the user's emotions and interests using Amazon Rekognition. This information is interpreted by the server and used to verify whether the generated topics match the user's emotions. This process enables a flexible and engaging communication experience for the user.

[0527] For example, in customer support for an e-commerce site, it becomes possible to instantly generate the best answer to a user's question, scan the user's facial expression in real time, and provide additional information to enhance their sense of security. The following is an example of a prompt message.

[0528] "Please describe typical problems and solutions related to [product name] recently purchased by customers. Include any relevant recent news. Also, please address emotional feedback based on customer facial expressions."

[0529] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0530] Step 1:

[0531] When a user inputs voice information into a device, the device records the audio. This audio information is sent to a server. The server receives the audio information and converts the input audio data into text information using the Google Cloud Speech-to-Text API. This process yields the text data read from the audio.

[0532] Step 2:

[0533] The server analyzes the acquired text information and uses Azure Machine Learning to compare it with past conversation data and external information sources from the internet. In this analysis process, a generative AI model generates appropriate topics via prompt sentences based on the input text information, and content relevant to the user is selected. As a result, topics based on the user's interests are output.

[0534] Step 3:

[0535] While the user is operating the device, the device uses its camera in real time to acquire image information of the user. The server analyzes this image information using Amazon Rekognition and recognizes the user's emotions and level of interest by interpreting their facial expressions. Based on the emotion data obtained from the input image information, the conversation content is adjusted.

[0536] Step 4:

[0537] The server references the acquired sentiment data and generated topics to determine the optimal flow of conversation. It then presents appropriate topics to the user using the provided means. Topics are suggested to match the prompt text and sentiment data, allowing the user to experience more comfortable communication. This step completes the system's output to the user.

[0538] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0539] This invention is a system that integrates speech recognition, emotion recognition, and data analysis to facilitate smooth communication in conversations. This system analyzes the user's emotional state in real time and provides appropriate topics for conversation, thereby achieving effective communication.

[0540] About the basic configuration

[0541] The system primarily consists of a recording device for acquiring speaker voice data, a recognition device for converting voice data into text and classifying speakers, an analysis device for analyzing past conversation data and external information, an emotion engine for recognizing the user's facial expressions and tone of voice, and a presentation device for presenting selected topics. These devices work together to provide the user with an optimal conversational experience.

[0542] Regarding the acquisition and recognition of audio data

[0543] The device acquires audio data in real time via its microphone and sends it to the server. The server uses speech recognition technology to convert the audio into text data and simultaneously classify the speaker. Frequency characteristics are used for speaker classification to accurately identify each individual speaker.

[0544] About the Emotion Engine

[0545] The device uses its camera and microphone to collect the user's facial expressions and tone of voice during conversations. An emotion engine on the server analyzes this data to recognize the user's emotions in real time, and uses the results to influence the selection of topics in the conversation.

[0546] Regarding the selection and presentation of topics

[0547] The server integrates data obtained from past conversation logs and external sources with the results of the emotion engine's analysis, and generates topics that are interesting and relevant to the user through a generative AI model. The selected topics are presented to the user via the device and used as material to continue the conversation.

[0548] Specific example

[0549] For example, when a user is talking to a colleague they've just met at work, if the system's emotion engine determines the user is nervous, it will prioritize suggesting topics that will help them relax. Furthermore, if it's clear the user is interested in a particular topic, the system will provide more in-depth information on that topic. In this way, the system addresses the user's emotional needs, helping to build natural and smooth communication.

[0550] The following describes the processing flow.

[0551] Step 1:

[0552] The device activates the microphone at specific locations and times to record surrounding sounds. The recorded audio data is converted to a digital format in real time and sent to the server.

[0553] Step 2:

[0554] The server inputs the received audio data into the speech recognition engine, which converts it into text data. During this process, frequency analysis is performed simultaneously, and the speaker is identified and classified by analyzing the voiceprint.

[0555] Step 3:

[0556] The server retrieves the speaker's past conversation logs from the database. The logs contain past topics, interests, and frequently used words, and the server updates the speaker profile based on this information.

[0557] Step 4:

[0558] The server connects to an external news API to collect the latest topics and related information. This data is then matched with speaker profiles to create a list of topics that are highly relevant to the speaker.

[0559] Step 5:

[0560] The device uses its camera and microphone to capture the user's facial expressions and tone of voice during a conversation. This emotion data is then sent to the server's emotion engine.

[0561] Step 6:

[0562] The server's emotion engine analyzes the user's emotional data and determines their emotional state. The emotional results are fed back into topic selection in real time.

[0563] Step 7:

[0564] The server uses a generated AI model to integrate past logs, external information, and sentiment data to generate topics that will interest the user. These topics are then prioritized, and a refined list is created.

[0565] Step 8:

[0566] The terminal presents the user with topics from the server. If the user shows interest in a topic, the system facilitates continued conversation by delving deeper into related information.

[0567] Step 9:

[0568] The user continues the conversation based on the suggested topics. The device continues to monitor the user's emotions and the content of the conversation, and optimizes the conversation environment by suggesting new topics as needed.

[0569] (Example 2)

[0570] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0571] Conventional conversation support systems have struggled to achieve natural and smooth communication because they cannot accurately recognize the user's emotional state from their voice and facial expressions, and they do not adequately provide topics of interest or relevance. Therefore, there is a need for a system that can analyze the user's emotions and interests in real time and provide topics based on that analysis.

[0572] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0573] In this invention, the server includes recording means for acquiring audio data, speech recognition means for converting audio data into text data and classifying speakers, data analysis means for analyzing past conversation logs and data from external sources to generate selected topics, sentiment analysis means for acquiring facial expression data and voice features in real time and recognizing emotional states, and generative model utilization means for presenting topics that are highly relevant and interesting to the user. This makes it possible to achieve natural and effective communication by analyzing the user's emotional state and providing appropriate topics accordingly.

[0574] "Voice data" refers to information obtained by capturing a user's speech as a digital signal.

[0575] "Recording means" refers to devices or mechanisms for acquiring and storing audio data.

[0576] "Speech recognition means" refers to technologies and methods for converting speech data into text data and identifying the speaker.

[0577] "Data analysis methods" refer to techniques and methods for generating topics by analyzing information obtained from past conversation logs and external sources.

[0578] "Emotional analysis methods" refer to technologies and methods for analyzing a user's facial expressions and tone of voice in real time to determine their emotional state.

[0579] "Generative model utilization methods" refer to techniques and methods for generating and presenting relevant topics and prompts using generative models.

[0580] "Frequency response" refers to the acoustic characteristics used to analyze the properties of speech and identify the speaker.

[0581] In an embodiment of this invention, the system has a configuration that integrates multiple functions. First, the terminal is equipped with a microphone to acquire the user's voice data in real time. The terminal then transmits the acquired voice data to a server. The server uses speech recognition technology to convert the voice data into text data and identify the speaker. General speech recognition software and libraries are used here.

[0582] Furthermore, the device collects user facial expression data through its camera. This image data and audio data are combined and analyzed in real time by an emotion analysis engine on a server to evaluate the user's emotional state. In this process, the user's tone of voice and facial movements are important indicators.

[0583] The server performs data analysis based on sentiment analysis results, past conversation history, and external information sources. This analysis utilizes machine learning models to generate topics relevant to the user using an AI model. These generated topics are then presented to the user as prompts via the terminal.

[0584] Specifically, for example, when a user is discussing a new project with a colleague, the server detects the user's level of expectation and generates a prompt such as, "Have you ever brainstormed new ideas?" This prompt helps to stimulate the conversation and facilitate deeper discussion.

[0585] Such systems can accurately capture a user's emotional state and interests, and provide conversational support accordingly, thereby enabling natural and effective communication.

[0586] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0587] Step 1:

[0588] When a user speaks into the microphone, the device acquires audio data in real time. During this process, the audio signal is converted into digital data and sent to the server in a fixed buffer size. Analog audio signals are used as input, and digital audio data is obtained as output.

[0589] Step 2:

[0590] The server receives digital audio data and converts it into text data using speech recognition technology. Simultaneously, it analyzes the frequency characteristics to identify the speaker. The input is digital audio data, and the output consists of text data and speaker identification information. Speech recognition software is used in this process.

[0591] Step 3:

[0592] The device uses its camera to collect user facial expression data. This data, along with audio features, is sent to a server for sentiment analysis. An emotion engine processes this data to estimate the user's emotional state. The input consists of image data and audio feature data, and the output is the user's emotion category.

[0593] Step 4:

[0594] The server performs data analysis based on sentiment analysis results, past conversation logs, and data from external sources. This analysis utilizes machine learning algorithms and a generative AI model to determine topics highly relevant to the user. The input consists of sentiment analysis results and various data sources, and the output is a generated prompt message.

[0595] Step 5:

[0596] The terminal presents the user with a prompt message obtained from the server. This is done either as on-screen text or as voice guidance using speech synthesis. The input is the generated prompt message, and the output is the user experience provided. Through this process, the user can obtain the information and topics necessary for the next conversation.

[0597] (Application Example 2)

[0598] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server," and the headset-type terminal 314 will be referred to as the "terminal."

[0599] In modern face-to-face retail, smooth communication between service staff and customers is crucial. However, instantly grasping a customer's emotions and interests and providing appropriate topics and services is not easy. Therefore, there is a need for effective customer service methods that improve customer satisfaction and increase purchasing intent.

[0600] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0601] In this invention, the server includes recording means for acquiring voice data, recognition means for converting voice data into text data and classifying speakers, analysis means for selecting topics by analyzing past conversation data and information from external sources, recognition means for acquiring image data in real time and analyzing emotions and levels of interest, presentation means for presenting selected topics, and auxiliary means aimed at analyzing the customer's emotional state in real time during customer service and presenting appropriate customer service methods or topics. This makes it possible for customer service employees to provide appropriate communication according to the customer's emotional state, thereby improving customer satisfaction and purchasing intent.

[0602] "Recording means" refers to devices for acquiring audio data, primarily equipment that uses microphones or recording devices to collect speech.

[0603] A "recognition means" is a device that converts acquired audio data into text data and has the function of identifying the speaker. It also includes the ability to analyze emotions and levels of interest from image data.

[0604] An "analysis tool" is a device that selects topics based on past conversation data and information obtained from external sources, and is primarily a device that performs data analysis and association.

[0605] A "presentation means" is a device that provides selected topics to the user and outputs information via a visual or auditory interface.

[0606] "Auxiliary tools" are devices that analyze a customer's emotional state in real time during customer service and suggest appropriate customer service methods or topics. They primarily operate in conjunction with emotion recognition technology.

[0607] Embodiments of the present invention are primarily realized as customer service support systems for physical retail stores. This system aims to enable employees to understand the customer's emotional state in real time during customer service and to communicate appropriately accordingly. Each component, including the server, terminals, and users, operates in proper coordination.

[0608] First, the terminal is equipped with a microphone to collect voice data, recording conversations during customer service in real time. The recorded voice data is sent to a server, where it is converted into text data using speech recognition technology. The server then performs speaker recognition, enabling individual customer identification.

[0609] Meanwhile, the terminal uses its camera to acquire customer facial expression data. This data is then used by an emotion recognition engine on the server to analyze the customer's emotional state and interests. The emotion engine uses the customer's voice tone and facial expressions to perform a real-time assessment of their emotions.

[0610] Next, the server analyzes past conversation history and external information, and uses a generative AI model to generate appropriate topics and customer service methods based on that analysis. These generated topics are then presented on the interface screen via the terminal, providing information to the employee.

[0611] For example, if a customer in a store shows interest in a new product but is hesitant to purchase it, the system will provide information such as, "This new product was very popular last week and is a limited production item." This allows employees to provide accurate guidance to customers and encourage them to make a purchase.

[0612] An example of a prompt would be, "Suggest topics based on the sentiment analysis results in your conversation with the customer. If the customer seems a little anxious, suggest topics that might help them feel more at ease." This allows the generative AI model to learn specific ways of responding to customers and improve the quality of customer service.

[0613] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0614] Step 1:

[0615] The terminal uses a microphone to acquire conversational audio data in real time within the physical store. The acquired audio data is transmitted from the terminal to the server. The input here is the audio signal, and the output is the audio data transmitted to the server.

[0616] Step 2:

[0617] The server converts the received audio data into text data using speech recognition software. During this process, it identifies the speaker and classifies the content of each identified conversation. The input here is audio data, and the output is text data and speaker identification information.

[0618] Step 3:

[0619] The terminal acquires images of the customer's facial expressions via its camera and sends them to the server. The input is image data, and the output is the transfer of real-time image data to the server. Image acquisition is performed to obtain a clear image focusing on a specific area of ​​the face.

[0620] Step 4:

[0621] The server integrates acquired image data and voice tone data and analyzes the customer's emotions and level of interest using an emotion recognition engine. The inputs here are image data and voice tone data, and the output is the customer's emotional state and level of interest. The data is analyzed by calculating the rate of change in facial expressions and the pitch variation of voice tone.

[0622] Step 5:

[0623] The server analyzes past conversation history and data from external sources, combines this with sentiment recognition results, and generates relevant topics via a generative AI model. The input for this step is past conversation data, sentiment analysis results, and external information, while the output is suggested topics. The server uses the generative AI model to generate topics according to the prompt.

[0624] Step 6:

[0625] The generated topics are presented to the user (employee) via a visual or audio interface through the terminal. The input here is topic suggestions, and the output is information delivered to the user as display or audio guidance. The displayed information is organized to facilitate specific customer service actions.

[0626] The specific processing unit 290 transmits the result of the specific processing to the headset terminal 314. In the headset terminal 314, the control unit 46A causes the speaker 240 and display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0627] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0628] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and specific processing may also be performed by the headset terminal 314.

[0629] [Fourth Embodiment]

[0630] Figure 7 shows an example of the configuration of the data processing system 410 according to the fourth embodiment.

[0631] As shown in Figure 7, the data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[0632] The data processing device 12 comprises a computer 22, a database 24, and a communication interface 26. The computer 22 is an example of a "computer" related to the technology of this disclosure. The computer 22 comprises a processor 28, RAM 30, and storage 32. The processor 28, RAM 30, and storage 32 are connected to a bus 34. The database 24 and the communication interface 26 are also connected to the bus 34. The communication interface 26 is connected to a network 54. An example of the network 54 is a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0633] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication interface 44, and a controlled object 443. The computer 36 includes a processor 46, RAM 48, and storage 50. The processor 46, RAM 48, and storage 50 are connected to a bus 52. The microphone 238, speaker 240, camera 42, and controlled object 443 are also connected to the bus 52.

[0634] The microphone 238 receives voice signals from the user 20 and receives instructions from the user 20. The microphone 238 captures the voice signals from the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio according to the instructions from the processor 46.

[0635] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an image sensor such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the area around the user 20 (for example, an imaging range defined by a field of view equivalent to the width of a typical healthy person's field of vision).

[0636] Communication interface 44 is connected to network 54. Communication interfaces 44 and 26 are responsible for the exchange of various information between processor 46 and processor 28 via network 54. The exchange of various information between processor 46 and processor 28 using communication interfaces 44 and 26 is performed in a secure manner.

[0637] The controlled object 443 includes a display device, LEDs in the eyes, and motors that drive the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the robot 414's emotions can be expressed by controlling these motors. Furthermore, the robot 414's facial expressions can also be expressed by controlling the illumination state of the LEDs in its eyes.

[0638] Figure 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Figure 8, the data processing device 12 performs specific processing using the processor 28. The storage 32 stores the specific processing program 56.

[0639] The specific processing program 56 is an example of a "program" relating to the technology of this disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0640] The storage 32 stores the data generation model 58 and the emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0641] In robot 414, the processor 46 performs the reception output processing. The storage 50 stores the reception output program 60. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output processing is realized by the processor 46 operating as a control unit 46A according to the reception output program 60 executed on the RAM 48.

[0642] Next, the specific processing performed by the specific processing unit 290 of the data processing device 12 will be described. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0643] This invention provides a system that automates the selection and provision of topics necessary for conversation, enabling users to communicate smoothly. This system consists of multiple modules that perform voice data acquisition, recognition, data analysis, topic generation, and topic updates based on real-time responses.

[0644] Regarding voice data acquisition and recognition

[0645] The device records audio at the start of a meeting or conversation and sends the data to a server. The server receives this audio data and converts it into text data using speech recognition technology. Simultaneously, it performs frequency analysis to identify individual speakers and records the statements of each identified speaker.

[0646] Data analysis and topic generation

[0647] The server collects past conversation logs and identifies each speaker's interests and concerns. Furthermore, it gathers the latest information through the internet and news APIs, selecting information highly relevant to each speaker. Based on this data, a generative AI model generates topics that are likely to interest each individual speaker.

[0648] Reaction analysis and topic presentation

[0649] The device uses its camera during the conversation to capture the other person's facial expressions and gestures. The server uses image recognition technology to evaluate the other person's emotions and level of interest in real time. Based on the speaker's responses, it assesses the appropriateness of the topic and adjusts the next topic offered as needed.

[0650] Specific example

[0651] This system is used by users participating in international conferences to support conversations with people they are meeting for the first time. The device records the conversation, and the server handles topic selection, freeing the user from the stress of choosing topics. Simultaneously, the system analyzes the other person's smiles and interesting reactions in real time, determining the next topic accordingly, ensuring a natural flow of conversation. Because the topics provided by the system adapt flexibly to the flow of the conversation, users can enjoy comfortable communication.

[0652] The following describes the processing flow.

[0653] Step 1:

[0654] The device activates the microphone at the start of a conversation and continuously records audio data. The recorded audio is sent to the server as packet data in real time.

[0655] Step 2:

[0656] The server inputs the received audio data into the speech recognition engine and converts it into text data. Simultaneously, it performs frequency analysis to extract voiceprint information and identify the speaker.

[0657] Step 3:

[0658] The server retrieves the speaker's past logs from the database. These logs record topics the speaker has shown interest in and keywords they have used to express interest, and the server uses this information to update the speaker's profile.

[0659] Step 4:

[0660] The server uses an external news API to collect the latest news data. The collected news data is analyzed and matched with speaker profile information to select highly relevant topics.

[0661] Step 5:

[0662] The server uses a generative AI model to generate topics that are likely to interest the speaker. These generated topics are then incorporated into the user's conversation in real time.

[0663] Step 6:

[0664] The device uses its camera to capture the facial expressions and gestures of the person it's talking to. The captured video data is then sent to a server.

[0665] Step 7:

[0666] The server analyzes the received video data using an image recognition engine and scores the emotions and level of interest of the person being spoken to in real time.

[0667] Step 8:

[0668] The server evaluates the effectiveness of the presented topic based on the scoring results. Based on the evaluation, it decides whether to continue with the topic or change it to a newly selected topic.

[0669] Step 9:

[0670] The user continues the conversation using the topic provided. The device prompts the user with the next topic as the conversation progresses, ensuring a smooth flow of conversation at all times.

[0671] (Example 1)

[0672] Next, we will describe Example 1. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0673] To facilitate smooth communication in conversations, it is necessary to automatically generate appropriate topics and dynamically adjust the flow of the conversation by evaluating the other person's responses in real time. Such a system would also be effective in overcoming language barriers when meeting someone for the first time or in cross-cultural exchange, but current technology does not yet offer sufficient automation or adaptability.

[0674] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 1 is realized by the following means.

[0675] In this invention, the server includes acquisition means for acquiring audio signals, recognition means for converting the audio signals into strings and identifying the speaker, generation means for analyzing past dialogue history and information from external data sources to generate topics, evaluation means for acquiring video signals in real time and evaluating emotions and levels of interest, and notification means for notifying the generated topics. This enables the selection of appropriate topics that take into account the characteristics and interests of the other party, and the dynamic adjustment of conversation content based on reactions.

[0676] An "audio signal" is a representation of sound as an electrical signal, and it is the basis for acquiring and processing sound waves in digital or analog format.

[0677] "Acquisition means" refers to a device or method for collecting input data such as audio and video, and includes sensors, microphones, cameras, etc.

[0678] "Recognition means" refers to methods or devices that analyze audio signals or image data to identify specific patterns or information, and utilizes speech recognition or image recognition technologies.

[0679] "Generative means" refers to methods or devices for constructing and producing new information or content based on collected data, such as those used to automatically generate text or topics.

[0680] "Evaluation methods" refer to methods and devices for analyzing and judging qualitative information such as emotions and levels of interest from acquired data, and are particularly used for analyzing situations in real time.

[0681] "Notification means" refers to methods or devices for conveying information or results to the user, and includes providing information using screen displays, audio output, vibration, etc.

[0682] This invention is an automated system for assisting conversations, which uses audio and video signals to generate appropriate topics and dynamically adjusts the topics based on real-time responses.

[0683] Acquisition and conversion of audio signals

[0684] The device is equipped with a highly sensitive microphone, which captures the audio signal of the conversation in real time. This audio signal is transmitted to the server in digital format.

[0685] The server uses speech recognition software (e.g., a common speech recognition API) to convert the received audio signal into text. This organizes the conversation content as text data.

[0686] Speaker identification and data analysis

[0687] The server uses deep learning technology to identify speakers from the frequency characteristics of audio signals. It then analyzes stored past dialogue history to identify each speaker's interests and concerns.

[0688] When collecting the latest information from external data sources, the server uses a common data collection API.

[0689] Topic generation and adjustment

[0690] The server uses a generative AI model (e.g., a general generative AI platform) to generate topics highly relevant to each speaker. The generated topics are formatted with prompts and input into the model in the form of "Generate interesting news related to environmental issues."

[0691] Acquisition and evaluation of video signals

[0692] The camera on the device captures the video signal of the person you are talking to in real time. This allows the other person's facial expressions and gestures to be transmitted to the server.

[0693] The server uses this video signal and employs a common image recognition API to evaluate emotions and levels of interest in real time, and to confirm the appropriateness of the generated topic.

[0694] Specific example

[0695] Imagine a scenario where a user is participating in a cross-cultural exchange event. Using this system, the user can analyze the other person's smiles and surprised reactions, and then let the system choose the next topic accordingly. The system's dynamic adjustment function ensures that the conversation flows naturally and without interruption. In this way, the invention supports people's communication and provides a richer exchange experience.

[0696] The flow of the specific processing in Example 1 will be explained using Figure 11.

[0697] Step 1:

[0698] The terminal acquires audio signals using a high-sensitivity microphone. The input is the audio signal from a conversation, which the terminal converts into a digital format before sending it to the server. Specifically, the microphone picks up the sound, and the built-in processing unit encodes it into a digital signal.

[0699] Step 2:

[0700] The server analyzes the received digital audio signal and performs speech recognition. The input is the acquired digital audio signal, and the output is a string generated based on it. The server uses a speech recognition API to convert the audio signal into text data while identifying the speaker. Specifically, the speech recognition algorithm maps the sound wave pattern to a linguistic structure.

[0701] Step 3:

[0702] The server analyzes past dialogue history using string data. The input for this step is the converted string data and past dialogue history, and the output is relevant data indicating the speaker's interests. The server uses natural language processing techniques to extract keywords from the text and identify the speaker's interests. Specifically, the text analysis engine evaluates each word and extracts the most important words.

[0703] Step 4:

[0704] The server collects information from external data sources and has a generative AI model generate topics. The input is the speaker's interest data and external data, and the output is a topic prompt sentence. By providing the prompt sentence to the generative AI model, the topics suggested by the model are obtained. Specifically, external data is collected via a RESTful API, and prompts such as "Please tell me the latest news on environmental issues" are sent to the generative AI model.

[0705] Step 5:

[0706] The terminal acquires video signals through the camera and transmits them to the server. The input is a real-time video signal, and the output is facial expression and movement data that the server uses for evaluation. The camera films the person being spoken to and sends the video to the server frame by frame in digital format.

[0707] Step 6:

[0708] The server analyzes the video signal and evaluates emotions and levels of interest. The input is real-time video data, and the output is the evaluation result of emotions and levels of interest. The server uses an image recognition API to determine emotions from facial expressions and movements. A visual information analysis algorithm analyzes image frames and performs actions to capture changes in facial expressions.

[0709] Step 7:

[0710] The server determines the appropriateness of the topic based on the evaluation results and updates the topic as needed. The output is the updated topic prompt. The server refers to the evaluation results and provides new input to the generating AI model to generate the next topic. Specifically, it feeds back the emotional data obtained in the previous step and adjusts the conversation content.

[0711] Step 8:

[0712] The terminal notifies the user of the generated topic. The input is the updated topic, and the output is the notification to the user. The terminal uses a display device and speech synthesis technology to convey the suggested topic to the user.

[0713] (Application Example 1)

[0714] Next, we will explain Application Example 1. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0715] To achieve smooth communication, selecting appropriate topics and adjusting conversations through real-time sentiment analysis are crucial. However, this is difficult to do with conventional systems, and there is a particular lack of technology to flexibly adapt the flow of dialogue while considering the user's emotions. Furthermore, even in customer support responses, there is a lack of technology to immediately generate and provide the most appropriate answers to user questions.

[0716] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 1 is realized by the following means.

[0717] In this invention, the server includes means for converting audio information into text information and classifying the audio source, means for analyzing past conversation information and information from external sources to select a topic, and means for acquiring image information in real time and analyzing emotions and levels of interest. This enables adaptive communication based on the user's current emotions and interests.

[0718] "Audio information" refers to audio data, which is data provided through the user's voice.

[0719] "Recording means" refers to a device or system for collecting and recording audio information.

[0720] "Text information" refers to information obtained by converting audio information into written characters.

[0721] "Recognition means" refers to a system or technology that converts audio information into text information and classifies the audio source.

[0722] "Analysis means" refers to systems and technologies for analyzing past conversation information and information from external sources.

[0723] "External information sources" refer to external data providers such as the internet and news APIs.

[0724] "Image information" refers to visual data acquired in real time, including facial expressions and gestures.

[0725] "Means of delivery" refers to systems and devices used to present selected topics to users.

[0726] "Control means" refers to devices or systems that analyze the user's emotions from their facial expressions and adjust the conversation accordingly.

[0727] A system implementing this invention includes a program for acquiring voice information, converting it to text information, and controlling conversation adjustments. Specifically, a server receives voice information and converts it to text information using speech recognition software. The speech recognition technology used here is Google Cloud Speech-to-Text, which analyzes voice with high accuracy.

[0728] Next, the server uses Azure Machine Learning to generate user-related topics based on past conversation information and data collected from external sources. The generating AI model analyzes the data based on the prompt text and selects the most appropriate conversation content.

[0729] Meanwhile, the device acquires image information in real time and analyzes the user's emotions and interests using Amazon Rekognition. This information is interpreted by the server and used to verify whether the generated topics match the user's emotions. This process enables a flexible and engaging communication experience for the user.

[0730] For example, in customer support for an e-commerce site, it becomes possible to instantly generate the best answer to a user's question, scan the user's facial expression in real time, and provide additional information to enhance their sense of security. The following is an example of a prompt message.

[0731] "Please describe typical problems and solutions related to [product name] recently purchased by customers. Include any relevant recent news. Also, please address emotional feedback based on customer facial expressions."

[0732] The flow of a specific process in Application Example 1 will be explained using Figure 12.

[0733] Step 1:

[0734] When a user inputs voice information into a device, the device records the audio. This audio information is sent to a server. The server receives the audio information and converts the input audio data into text information using the Google Cloud Speech-to-Text API. This process yields the text data read from the audio.

[0735] Step 2:

[0736] The server analyzes the acquired text information and uses Azure Machine Learning to compare it with past conversation data and external information sources from the internet. In this analysis process, a generative AI model generates appropriate topics via prompt sentences based on the input text information, and content relevant to the user is selected. As a result, topics based on the user's interests are output.

[0737] Step 3:

[0738] While the user is operating the device, the device uses its camera in real time to acquire image information of the user. The server analyzes this image information using Amazon Rekognition and recognizes the user's emotions and level of interest by interpreting their facial expressions. Based on the emotion data obtained from the input image information, the conversation content is adjusted.

[0739] Step 4:

[0740] The server references the acquired sentiment data and generated topics to determine the optimal flow of conversation. It then presents appropriate topics to the user using the provided means. Topics are suggested to match the prompt text and sentiment data, allowing the user to experience more comfortable communication. This step completes the system's output to the user.

[0741] Furthermore, an emotion engine that estimates the user's emotions may be incorporated. That is, the identification processing unit 290 may use the emotion identification model 59 to estimate the user's emotions and perform identification processing using the user's emotions.

[0742] This invention is a system that integrates speech recognition, emotion recognition, and data analysis to facilitate smooth communication in conversations. This system analyzes the user's emotional state in real time and provides appropriate topics for conversation, thereby achieving effective communication.

[0743] About the basic configuration

[0744] The system primarily consists of a recording device for acquiring speaker voice data, a recognition device for converting voice data into text and classifying speakers, an analysis device for analyzing past conversation data and external information, an emotion engine for recognizing the user's facial expressions and tone of voice, and a presentation device for presenting selected topics. These devices work together to provide the user with an optimal conversational experience.

[0745] Regarding the acquisition and recognition of audio data

[0746] The device acquires audio data in real time via its microphone and sends it to the server. The server uses speech recognition technology to convert the audio into text data and simultaneously classify the speaker. Frequency characteristics are used for speaker classification to accurately identify each individual speaker.

[0747] About the Emotion Engine

[0748] The device uses its camera and microphone to collect the user's facial expressions and tone of voice during conversations. An emotion engine on the server analyzes this data to recognize the user's emotions in real time, and uses the results to influence the selection of topics in the conversation.

[0749] Regarding the selection and presentation of topics

[0750] The server integrates data obtained from past conversation logs and external sources with the results of the emotion engine's analysis, and generates topics that are interesting and relevant to the user through a generative AI model. The selected topics are presented to the user via the device and used as material to continue the conversation.

[0751] Specific example

[0752] For example, when a user is talking to a colleague they've just met at work, if the system's emotion engine determines the user is nervous, it will prioritize suggesting topics that will help them relax. Furthermore, if it's clear the user is interested in a particular topic, the system will provide more in-depth information on that topic. In this way, the system addresses the user's emotional needs, helping to build natural and smooth communication.

[0753] The following describes the processing flow.

[0754] Step 1:

[0755] The device activates the microphone at specific locations and times to record surrounding sounds. The recorded audio data is converted to a digital format in real time and sent to the server.

[0756] Step 2:

[0757] The server inputs the received audio data into the speech recognition engine, which converts it into text data. During this process, frequency analysis is performed simultaneously, and the speaker is identified and classified by analyzing the voiceprint.

[0758] Step 3:

[0759] The server retrieves the speaker's past conversation logs from the database. The logs contain past topics, interests, and frequently used words, and the server updates the speaker profile based on this information.

[0760] Step 4:

[0761] The server connects to an external news API to collect the latest topics and related information. This data is then matched with speaker profiles to create a list of topics that are highly relevant to the speaker.

[0762] Step 5:

[0763] The device uses its camera and microphone to capture the user's facial expressions and tone of voice during a conversation. This emotion data is then sent to the server's emotion engine.

[0764] Step 6:

[0765] The server's emotion engine analyzes the user's emotional data and determines their emotional state. The emotional results are fed back into topic selection in real time.

[0766] Step 7:

[0767] The server uses a generated AI model to integrate past logs, external information, and sentiment data to generate topics that will interest the user. These topics are then prioritized, and a refined list is created.

[0768] Step 8:

[0769] The terminal presents the user with topics from the server. If the user shows interest in a topic, the system facilitates continued conversation by delving deeper into related information.

[0770] Step 9:

[0771] The user continues the conversation based on the suggested topics. The device continues to monitor the user's emotions and the content of the conversation, and optimizes the conversation environment by suggesting new topics as needed.

[0772] (Example 2)

[0773] Next, we will describe Example 2. In the following description, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0774] Conventional conversation support systems have struggled to achieve natural and smooth communication because they cannot accurately recognize the user's emotional state from their voice and facial expressions, and they do not adequately provide topics of interest or relevance. Therefore, there is a need for a system that can analyze the user's emotions and interests in real time and provide topics based on that analysis.

[0775] The identification process performed by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means.

[0776] In this invention, the server includes recording means for acquiring audio data, speech recognition means for converting audio data into text data and classifying speakers, data analysis means for analyzing past conversation logs and data from external sources to generate selected topics, sentiment analysis means for acquiring facial expression data and voice features in real time and recognizing emotional states, and generative model utilization means for presenting topics that are highly relevant and interesting to the user. This makes it possible to achieve natural and effective communication by analyzing the user's emotional state and providing appropriate topics accordingly.

[0777] "Voice data" refers to information obtained by capturing a user's speech as a digital signal.

[0778] "Recording means" refers to devices or mechanisms for acquiring and storing audio data.

[0779] "Speech recognition means" refers to technologies and methods for converting speech data into text data and identifying the speaker.

[0780] "Data analysis methods" refer to techniques and methods for generating topics by analyzing information obtained from past conversation logs and external sources.

[0781] "Emotional analysis methods" refer to technologies and methods for analyzing a user's facial expressions and tone of voice in real time to determine their emotional state.

[0782] "Generative model utilization methods" refer to techniques and methods for generating and presenting relevant topics and prompts using generative models.

[0783] "Frequency response" refers to the acoustic characteristics used to analyze the properties of speech and identify the speaker.

[0784] In an embodiment of this invention, the system has a configuration that integrates multiple functions. First, the terminal is equipped with a microphone to acquire the user's voice data in real time. The terminal then transmits the acquired voice data to a server. The server uses speech recognition technology to convert the voice data into text data and identify the speaker. General speech recognition software and libraries are used here.

[0785] Furthermore, the device collects user facial expression data through its camera. This image data and audio data are combined and analyzed in real time by an emotion analysis engine on a server to evaluate the user's emotional state. In this process, the user's tone of voice and facial movements are important indicators.

[0786] The server performs data analysis based on sentiment analysis results, past conversation history, and external information sources. This analysis utilizes machine learning models to generate topics relevant to the user using an AI model. These generated topics are then presented to the user as prompts via the terminal.

[0787] Specifically, for example, when a user is discussing a new project with a colleague, the server detects the user's level of expectation and generates a prompt such as, "Have you ever brainstormed new ideas?" This prompt helps to stimulate the conversation and facilitate deeper discussion.

[0788] Such systems can accurately capture a user's emotional state and interests, and provide conversational support accordingly, thereby enabling natural and effective communication.

[0789] The flow of the specific processing in Example 2 will be explained using Figure 13.

[0790] Step 1:

[0791] When a user speaks into the microphone, the device acquires audio data in real time. During this process, the audio signal is converted into digital data and sent to the server in a fixed buffer size. Analog audio signals are used as input, and digital audio data is obtained as output.

[0792] Step 2:

[0793] The server receives digital audio data and converts it into text data using speech recognition technology. Simultaneously, it analyzes the frequency characteristics to identify the speaker. The input is digital audio data, and the output consists of text data and speaker identification information. Speech recognition software is used in this process.

[0794] Step 3:

[0795] The device uses its camera to collect user facial expression data. This data, along with audio features, is sent to a server for sentiment analysis. An emotion engine processes this data to estimate the user's emotional state. The input consists of image data and audio feature data, and the output is the user's emotion category.

[0796] Step 4:

[0797] The server performs data analysis based on sentiment analysis results, past conversation logs, and data from external sources. This analysis utilizes machine learning algorithms and a generative AI model to determine topics highly relevant to the user. The input consists of sentiment analysis results and various data sources, and the output is a generated prompt message.

[0798] Step 5:

[0799] The terminal presents the user with a prompt message obtained from the server. This is done either as on-screen text or as voice guidance using speech synthesis. The input is the generated prompt message, and the output is the user experience provided. Through this process, the user can obtain the information and topics necessary for the next conversation.

[0800] (Application Example 2)

[0801] Next, we will explain application example 2. In the following explanation, the data processing device 12 will be referred to as the "server" and the robot 414 as the "terminal".

[0802] In modern face-to-face retail, smooth communication between service staff and customers is crucial. However, instantly grasping a customer's emotions and interests and providing appropriate topics and services is not easy. Therefore, there is a need for effective customer service methods that improve customer satisfaction and increase purchasing intent.

[0803] The specific processing performed by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means.

[0804] In this invention, the server includes recording means for acquiring voice data, recognition means for converting voice data into text data and classifying speakers, analysis means for selecting topics by analyzing past conversation data and information from external sources, recognition means for acquiring image data in real time and analyzing emotions and levels of interest, presentation means for presenting selected topics, and auxiliary means aimed at analyzing the customer's emotional state in real time during customer service and presenting appropriate customer service methods or topics. This makes it possible for customer service employees to provide appropriate communication according to the customer's emotional state, thereby improving customer satisfaction and purchasing intent.

[0805] "Recording means" refers to devices for acquiring audio data, primarily equipment that uses microphones or recording devices to collect speech.

[0806] A "recognition means" is a device that converts acquired audio data into text data and has the function of identifying the speaker. It also includes the ability to analyze emotions and levels of interest from image data.

[0807] An "analysis tool" is a device that selects topics based on past conversation data and information obtained from external sources, and is primarily a device that performs data analysis and association.

[0808] A "presentation means" is a device that provides selected topics to the user and outputs information via a visual or auditory interface.

[0809] "Auxiliary tools" are devices that analyze a customer's emotional state in real time during customer service and suggest appropriate customer service methods or topics. They primarily operate in conjunction with emotion recognition technology.

[0810] Embodiments of the present invention are primarily realized as customer service support systems for physical retail stores. This system aims to enable employees to understand the customer's emotional state in real time during customer service and to communicate appropriately accordingly. Each component, including the server, terminals, and users, operates in proper coordination.

[0811] First, the terminal is equipped with a microphone to collect voice data, recording conversations during customer service in real time. The recorded voice data is sent to a server, where it is converted into text data using speech recognition technology. The server then performs speaker recognition, enabling individual customer identification.

[0812] Meanwhile, the terminal uses its camera to acquire customer facial expression data. This data is then used by an emotion recognition engine on the server to analyze the customer's emotional state and interests. The emotion engine uses the customer's voice tone and facial expressions to perform a real-time assessment of their emotions.

[0813] Next, the server analyzes past conversation history and external information, and uses a generative AI model to generate appropriate topics and customer service methods based on that analysis. These generated topics are then presented on the interface screen via the terminal, providing information to the employee.

[0814] For example, if a customer in a store shows interest in a new product but is hesitant to purchase it, the system will provide information such as, "This new product was very popular last week and is a limited production item." This allows employees to provide accurate guidance to customers and encourage them to make a purchase.

[0815] An example of a prompt would be, "Suggest topics based on the sentiment analysis results in your conversation with the customer. If the customer seems a little anxious, suggest topics that might help them feel more at ease." This allows the generative AI model to learn specific ways of responding to customers and improve the quality of customer service.

[0816] The flow of a specific process in Application Example 2 will be explained using Figure 14.

[0817] Step 1:

[0818] The terminal uses a microphone to acquire conversational audio data in real time within the physical store. The acquired audio data is transmitted from the terminal to the server. The input here is the audio signal, and the output is the audio data transmitted to the server.

[0819] Step 2:

[0820] The server converts the received audio data into text data using speech recognition software. During this process, it identifies the speaker and classifies the content of each identified conversation. The input here is audio data, and the output is text data and speaker identification information.

[0821] Step 3:

[0822] The terminal acquires images of the customer's facial expressions via its camera and sends them to the server. The input is image data, and the output is the transfer of real-time image data to the server. Image acquisition is performed to obtain a clear image focusing on a specific area of ​​the face.

[0823] Step 4:

[0824] The server integrates acquired image data and voice tone data and analyzes the customer's emotions and level of interest using an emotion recognition engine. The inputs here are image data and voice tone data, and the output is the customer's emotional state and level of interest. The data is analyzed by calculating the rate of change in facial expressions and the pitch variation of voice tone.

[0825] Step 5:

[0826] The server analyzes past conversation history and data from external sources, combines this with sentiment recognition results, and generates relevant topics via a generative AI model. The input for this step is past conversation data, sentiment analysis results, and external information, while the output is suggested topics. The server uses the generative AI model to generate topics according to the prompt.

[0827] Step 6:

[0828] The generated topics are presented to the user (employee) via a visual or audio interface through the terminal. The input here is topic suggestions, and the output is information delivered to the user as display or audio guidance. The displayed information is organized to facilitate specific customer service actions.

[0829] The specific processing unit 290 transmits the result of the specific processing to the robot 414. In the robot 414, the control unit 46A causes the speaker 240 and the controlled object 443 to output the result of the specific processing. The microphone 238 acquires audio indicating user input for the result of the specific processing. The control unit 46A transmits the audio data indicating user input acquired by the microphone 238 to the data processing unit 12. In the data processing unit 12, the specific processing unit 290 acquires the audio data.

[0830] Data generation model 58 is a type of so-called generative AI (Artificial Intelligence). One example of data generation model 58 is ChatGPT (Internet Search).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search) <url: https: gemini.google.com ?hl="ja">Examples of generative AI include the following. The data generation model 58 is obtained by performing deep learning on a neural network. The data generation model 58 is input with prompts containing instructions, and with inference data such as audio data representing speech, text data representing text, and image data representing images. The data generation model 58 infers from the input inference data according to the instructions indicated by the prompts, and outputs the inference results in data formats such as audio data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0831] In the above embodiment, an example was given in which specific processing is performed by the data processing device 12, but the technology of this disclosure is not limited thereto, and the specific processing may also be performed by the robot 414.

[0832] Furthermore, the emotion identification model 59, acting as an emotion engine, may determine the user's emotion according to a specific mapping. Specifically, the emotion identification model 59 may determine the user's emotion according to a specific mapping, which is an emotion map (see Figure 9). Similarly, the emotion identification model 59 may also determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[0833] Figure 9 shows an emotion map 400 in which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. The closer to the center of the concentric circles, the more primitive the emotions are located. Further out of the concentric circles, emotions representing states and actions arising from mental states are located. Emotion is a concept that includes feelings and mental states. On the left side of the concentric circles, emotions that are generally generated from reactions occurring in the brain are located. On the right side of the concentric circles, emotions that are generally induced by situational judgment are located. Above and below the concentric circles, emotions that are generally generated from reactions occurring in the brain and induced by situational judgment are located. In addition, the emotion of "pleasure" is located on the upper side of the concentric circles, and the emotion of "displeasure" is located on the lower side. Thus, in the emotion map 400, multiple emotions are mapped based on the structure in which emotions arise, and emotions that are likely to occur simultaneously are mapped close together.

[0834] These emotions are distributed at the 3 o'clock position on the Emotion Map 400, and usually fluctuate between feelings of security and anxiety. In the right half of the Emotion Map 400, situational awareness takes precedence over internal feelings, resulting in a calm impression.

[0835] The inside of the Emotion Map 400 represents inner thoughts, while the outside represents actions. Therefore, the further you go from the outside of the Emotion Map 400, the more visible (expressed in actions) your emotions become.

[0836] Here, human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. Similarly, in robots, cars, motorcycles, etc., emotions can be created based on various balances, such as posture and battery level. When these balances deviate from the ideal, it results in discomfort, and when they approach the ideal, it results in pleasure. The emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on a system for analyzing brain physiological signals of speech emotion recognition and emotion, Tokushima University, doctoral dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map contains emotions belonging to a region called "response," where sensation is dominant. The right half of the emotion map contains emotions belonging to a region called "situation," where situational awareness is dominant.

[0837] The emotion map defines two emotions that promote learning. One is the emotion around the middle of the negative "repentance" and "reflection" on the situation side. In other words, it is when the robot experiences negative emotions such as "I never want to feel this way again" or "I don't want to be scolded again." The other is the emotion around the positive "desire" on the reaction side. In other words, it is when the robot has positive feelings such as "I want more" or "I want to know more."

[0838] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​representing each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple training data sets, which are combinations of user input and emotion values ​​representing each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions located close together have similar values, as shown in the emotion map 900 in Figure 10. Figure 10 shows an example where multiple emotions such as "reassured," "calm," and "confident" have similar emotion values.

[0839] The above description primarily focuses on the functions of the data processing device 12 in relation to this disclosure. However, the system related to this disclosure is not necessarily implemented on a server. The system related to this disclosure may be implemented as a general information processing system. This disclosure may be implemented, for example, as a software program that runs on a personal computer or as an application that runs on a smartphone. The method related to this disclosure may be provided to users in SaaS (Software as a Service) format.

[0840] In the above embodiment, an example was given in which a specific process is performed by a single computer 22. However, the technology of this disclosure is not limited thereto, and a distributed processing of the specific process may be performed by multiple computers, including computer 22. For example, a data generation model 58 may be provided in an external device of the data processing device 12, and the external device may generate data according to the input data.

[0841] In the above embodiment, an example was given in which the specific processing program 56 is stored in the storage 32, but the technology of this disclosure is not limited thereto. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-temporary storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-temporary storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes specific processing according to the specific processing program 56.

[0842] Alternatively, the specific processing program 56 may be stored in a storage device such as a server connected to the data processing device 12 via the network 54, and the specific processing program 56 may be downloaded and installed on the computer 22 in response to a request from the data processing device 12.

[0843] Furthermore, it is not necessary to store the entirety of the specific processing program 56 in a storage device such as a server connected to the data processing device 12 via the network 54, or to store the entirety of the specific processing program 56 in the storage 32; it is acceptable to store only a portion of the specific processing program 56.

[0844] The following types of processors can be used as hardware resources to perform specific processing. Examples of processors include a CPU, a general-purpose processor that functions as a hardware resource to perform specific processing by executing software, i.e., a program. Other examples of processors include dedicated electrical circuits, such as FPGAs (Field-Programmable Gate Arrays), PLDs (Programmable Logic Devices), or ASICs (Application Specific Integrated Circuits), which have circuit configurations specifically designed to perform specific processing. All of these processors have built-in or connected memory, and all of them perform specific processing by using memory.

[0845] The hardware resource that performs a specific process may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, the hardware resource that performs a specific process may consist of a single processor.

[0846] Examples of configurations using a single processor include, firstly, a configuration in which one or more CPUs and software are combined to form a single processor, and this processor functions as a hardware resource that performs a specific process. Secondly, there is a configuration using a processor that realizes the functions of the entire system, including multiple hardware resources that perform a specific process, on a single IC chip, as exemplified by SoCs (System-on-a-chip). In this way, a specific process is realized using one or more of the above types of processors as hardware resources.

[0847] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits that combine circuit elements such as semiconductor devices. Also, the specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps can be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0848] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0849] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted as being incorporated by reference.

[0850] The following is further disclosed regarding the embodiments described above.

[0851] (Claim 1)

[0852] A recording device for acquiring audio data,

[0853] A recognition device for converting the aforementioned audio data into text data and classifying the speaker,

[0854] An analytical device for selecting topics by analyzing past conversation data and information from external sources,

[0855] A recognition device that acquires image data in real time and analyzes emotions and levels of interest,

[0856] A presentation device for presenting selected topics,

[0857] A system that includes this.

[0858] (Claim 2)

[0859] The system according to claim 1, which identifies a speaker by utilizing the speaker's frequency characteristics.

[0860] (Claim 3)

[0861] The system according to claim 1, wherein the topic is updated based on the image data acquired by the recognition device.

[0862] "Example 1"

[0863] (Claim 1)

[0864] A means for acquiring audio signals,

[0865] A recognition means for converting the aforementioned audio signal into a string and identifying the speaker,

[0866] A generation method for generating topics by analyzing past dialogue history and information from external data sources,

[0867] An evaluation method for acquiring video signals in real time and evaluating emotions and levels of interest,

[0868] A notification means for notifying about generated topics,

[0869] A system that includes this.

[0870] (Claim 2)

[0871] The system according to claim 1, which identifies a speaker by utilizing the acoustic characteristics of the speaker.

[0872] (Claim 3)

[0873] The system according to claim 1, wherein the evaluation means adjusts the topic based on the video signal acquired.

[0874] "Application Example 1"

[0875] (Claim 1)

[0876] A recording means for acquiring audio information,

[0877] A recognition means for converting the aforementioned audio information into text information and classifying the audio source,

[0878] An analytical method for selecting topics by analyzing past conversation information and information from external sources,

[0879] A recognition method that acquires image information in real time and analyzes emotions and levels of interest,

[0880] A means of presenting selected topics,

[0881] A control mechanism for analyzing emotions from the user's facial expressions and adjusting the flow of conversation,

[0882] A system that includes this.

[0883] (Claim 2)

[0884] The system according to claim 1, which identifies a sound source by utilizing the frequency characteristics of the sound source.

[0885] (Claim 3)

[0886] The system according to claim 1, wherein the topic is updated based on the image information acquired by the recognition means.

[0887] "Example 2 of combining an emotion engine"

[0888] (Claim 1)

[0889] A recording means for acquiring audio data,

[0890] A speech recognition means for converting the aforementioned audio data into text data and classifying the speaker,

[0891] A data analysis method for generating selected topics by analyzing past conversation logs and data from external sources,

[0892] A means of emotion analysis for recognizing emotional states by acquiring facial expression data and voice characteristics in real time,

[0893] A method for utilizing generative models to present users with highly relevant and interesting topics,

[0894] A system that includes this.

[0895] (Claim 2)

[0896] The system according to claim 1, comprising a method for identifying a speaker using the speaker's frequency characteristics.

[0897] (Claim 3)

[0898] The system according to claim 1, comprising a method of updating a topic based on data obtained by the sentiment analysis means and presenting it via a generative model.

[0899] "Application example 2 of combining emotional engines"

[0900] (Claim 1)

[0901] A recording means for acquiring audio data,

[0902] A recognition means for converting the aforementioned audio data into text data and classifying the speaker,

[0903] An analytical method for selecting topics by analyzing past conversation data and information from external sources,

[0904] A recognition method for acquiring image data in real time and analyzing emotions and levels of interest,

[0905] A means of presenting the selected topic,

[0906] An auxiliary means aimed at analyzing the customer's emotional state in real time during customer service and suggesting appropriate customer service methods or topics,

[0907] A system that includes this.

[0908] (Claim 2)

[0909] The system according to claim 1, which identifies a speaker by utilizing the speaker's frequency characteristics.

[0910] (Claim 3)

[0911] The system according to claim 1, wherein the recognition means updates the topic based on the image data acquired and optimizes the customer service method. [Explanation of Symbols]

[0912] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Devices 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robots< / url:> < / url:> < / url:> < / url:>

Claims

1. A recording device for acquiring audio data, A recognition device for converting the aforementioned audio data into text data and classifying the speaker, An analytical device for selecting topics by analyzing past conversation data and information from external sources, A recognition device that acquires image data in real time and analyzes emotions and levels of interest, A presentation device for presenting selected topics, A system that includes this.

2. The system according to claim 1, which identifies a speaker by utilizing the speaker's frequency characteristics.

3. The system according to claim 1, wherein the topic is updated based on the image data acquired by the recognition device.

Images