System

Abstract

A system is provided.SOLUTION: A system comprising: means for capturing lip and tongue movements of a user; means for generating text based on the captured movements; and means for converting the generated text to speech.SELECTED DRAWING: Figure 1

Description

[Technical Field]

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

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

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] Remote work, an effective means of supporting diverse work styles in the context of teleworking and online meetings, faces environmental constraints. Voice communication is often difficult, particularly in open spaces, public places, and at home, posing a risk of privacy and confidential information leaks. The purpose of this invention is to overcome these constraints and enable smooth communication even in quiet environments and situations where privacy must be protected. [Means for solving the problem]

[0005] The present invention solves the problems by using the following means: A system is provided that includes means for capturing the movement of a user's lips and tongue, means for generating text based on the captured movement, and means for converting the generated text into speech. The system also includes means for improving the accuracy of text generation by utilizing the conversation context and the user's past learning data, means for transmitting the generated speech data to another terminal in real time to converse with the communication partner, means for converting speech received from the communication partner into text and displaying it to the user and / or playing it back as speech, and means for displaying training sample sentences that the user reads silently and saving the read data, thereby providing smooth and secure communication even in quiet environments.

[0006] "User's lip and tongue movements" refers to the visible changes that the lips and tongue make when the user speaks silently.

[0007] "Capture" refers to the process of capturing a user's lip and tongue movements using an in-camera or other visual sensor and storing them as digital data.

[0008] "Text generation" refers to the process of constructing meaningful sentences based on captured lip and tongue movement data.

[0009] "Speech conversion" refers to the process of converting generated text data into speech using speech synthesis technology.

[0010] "Past learning data" refers to a dataset containing lip and tongue movements and corresponding text data when a user reads silently during the training phase.

[0011] "Conversational context" refers to information including the current conversation flow, main themes, and related past conversation content.

[0012] "Voice data" refers to digital data of voice generated using voice synthesis technology.

[0013] "Call partner" refers to a party with whom a user makes a call.

[0014] "Training sample sentences" refer to example sentences or passages presented to a user for silent reading.

[0015] "Visual sensor" refers to a sensor device, such as an in-camera, used to capture the movements of a user's lips and tongue. [Brief explanation of the drawings]

[0016] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12]FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0017] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0018] First, the terms used in the following description will be explained.

[0019] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0020] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0021] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0022] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0023] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0024] [First embodiment]

[0025] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0026] 1, a 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.

[0027] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0028] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0029] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the 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.

[0030] 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 of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0031] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0032] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0033] 2, in the data processing device 12, a specific process 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" according to the technology of the present 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 process 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.

[0034] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0035] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0036] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0037] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. The system captures the movements of the user's lips and tongue, generates text based on the captured movements, and converts the text into speech. Below, we will provide an overview of the system's program and its specific processing in natural language.

[0038] System configuration

[0039] Capture Device

[0040] User: Speak silently.

[0041] On the device: The front camera is used to capture the user's lip and tongue movements in real time, which are then stored as visual data.

[0042] Data transmission and analysis

[0043] Terminal: Sends captured visual data to the server.

[0044] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[0045] Text-to-speech

[0046] Server: The generated text is converted into voice data using speech synthesis technology.

[0047] Server: Sends the generated audio data to other devices in real time.

[0048] Response processing from the other party

[0049] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0050] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[0051] Specific examples

[0052] 1. User training

[0053] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0054] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[0055] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[0056] 2. Actual call

[0057] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0058] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[0059] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[0060] Server: Converts text into speech and sends the speech data to the other device.

[0061] The other device: Plays the received voice data and transmits the user's words to the other device.

[0062] 3. Response from the other person

[0063] Other party: Reply verbally with "Got it, I'll join."

[0064] Remote device: Captures audio data and sends it to the server.

[0065] Server: Converts voice data into text and sends it to the user's device.

[0066] On the device: The received text is displayed to the user or played as audio.

[0067] This allows the system to provide a smooth and secure means of making calls even in quiet environments or places where privacy is required. By utilizing the user's past learning data and conversation context, the system can generate text with higher accuracy and realize natural-sounding voice calls.

[0068] The processing flow will be explained below.

[0069] Training Data Collection Processing Steps

[0070] Step 1: Training Initial Setup

[0071] User: Select "Silent Calling."

[0072] Terminal: Presents training sample sentences to the user and prepares them to begin training.

[0073] Step 2: Silent reading of sample sentences

[0074] User: Silently read the provided sample sentences aloud.

[0075] On your device: Use the front camera to capture the movements of your lips and tongue.

[0076] Step 3: Sending visual data

[0077] Terminal: Sends captured visual data to the server.

[0078] Step 4: Data analysis and training

[0079] Server: Analyzes the received visual data and stores the lip and tongue movements and corresponding text as training data.

[0080] Actual call processing steps

[0081] Step 1: Start the call

[0082] User: Select the person to call and start the call.

[0083] Device: Prepare for the call and activate the front camera.

[0084] Step 2: Silent speech

[0085] User: Speak silently.

[0086] On the device: Uses the front camera to capture the user's lip and tongue movements in real time.

[0087] Step 3: Sending visual data

[0088] Terminal: Sends captured visual data to the server.

[0089] Step 4: Text Generation

[0090] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[0091] Step 5: Audio conversion

[0092] Server: The generated text is converted into voice data using speech synthesis technology.

[0093] Step 6: Sending audio data

[0094] Server: Sends the generated voice data to the other device.

[0095] Step 7: Playing Audio

[0096] The other device: Plays the received voice data and transmits the user's words to the other device.

[0097] Steps for processing a reply from the other party

[0098] Step 1: Capture audio data

[0099] Recipient: Respond by voice as usual.

[0100] Remote device: Captures the response audio data and sends it to the server.

[0101] Step 2: Analyzing the audio data

[0102] Server: Converts the received voice data into text using voice recognition technology.

[0103] Step 3: Send a text

[0104] Server: Sends the textual response to the user's device.

[0105] Step 4: Display text or play audio

[0106] Device: Displays the received text to the user, or synthesizes the text to speech and plays it through headphones.

[0107] This is a step-by-step explanation of the program's processing. Through this series of processes, the system enables smooth and safe calls even in quiet environments or places where privacy is required.

[0108] Example 1

[0109] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0110] Currently, voice communication relies on direct voice communication, which makes it difficult to use in quiet environments or places where privacy is required. Voice communication is also difficult for people with hearing impairments. Furthermore, conventional voice input systems have problems with misrecognition due to background noise and differences in speaking speed.

[0111] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0112] In this invention, the server includes means for capturing the user's lip and tongue movements, means for generating text based on the captured movements, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for converting the speech data received from the other terminal into text. This enables silent calls, allowing for smooth and safe calls even in quiet environments or places where privacy is required. It also provides an alternative means of voice calls for people with hearing impairments, enabling highly accurate calls unaffected by background noise or speech rate.

[0113] "Capturing means" refers to devices and software that detect the user's lip and tongue movements in real time and acquire and store them as visual data.

[0114] The "means for generating text" refers to a device and software for analyzing the captured visual data and generating text data based on the content of the user's silent utterances.

[0115] The "means for converting to speech" refers to devices and software for converting the generated text data into synthetic speech. Specifically, speech synthesis technology is used.

[0116] The "transmitting means" refers to a device and software for transmitting the generated voice data to another terminal via a network.

[0117] The "means for converting into text" refers to the device and software that analyzes the voice data received from the other party's terminal and converts it into text data. Specifically, it uses voice recognition technology.

[0118] "Conversational context" refers to information and processing methods for generating appropriate text based on the context of the user's statements and the flow of the dialogue.

[0119] "Past learning data" refers to data accumulated based on the user's past training and conversation content, and is the data and processing method used to improve the accuracy of text generation.

[0120] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. This system provides the function of capturing the movements of the user's lips and tongue, generating text based on the captured movements, and converting the text into speech. It is also possible to convert the speech of the other party into text and provide it to the user.

[0121] System configuration

[0122] Capture Device

[0123] The user speaks silently, and the device uses the front camera to capture the user's lip and tongue movements in real time. The captured movements are saved as visual data. The front camera uses a high-resolution camera and a dedicated chip for real-time image processing.

[0124] Data transmission and analysis

[0125] The device sends the captured visual data to a server, where it is encrypted and transmitted securely using the HTTPS protocol.

[0126] The server analyzes the received visual data and generates corresponding text using CNN (Convolutional Neural Network) and RNN (Recurrent Neural Network) technologies, generating highly accurate text based on the user's past learning data and conversational context.

[0127] Text-to-speech

[0128] The server converts the generated text into audio data using TTS (Text to Speech) technology, such as Google Cloud Text-to-Speech. This audio data is generated as a high-quality audio file (e.g., WAV format).

[0129] The server transmits the generated voice data to the other terminal in real time.

[0130] Response processing from the other party

[0131] The other device captures the voice data spoken by the other user and sends it to the server. The voice data is encrypted and sent using the HTTPS protocol.

[0132] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology. In this case, we use "Google Cloud Speech-to-Text."

[0133] The server sends the converted text to the user terminal.

[0134] The device will then display the received text to the user or play it as audio.

[0135] Specific examples

[0136] 1. User training

[0137] The user silently reads the training sample sentence, "I am currently in Tokyo."

[0138] The device's in-camera captures lip and tongue movements and sends the visual data to a server.

[0139] The server analyzes the received data and saves the text "I am in Tokyo now" along with the corresponding visual data as training data.

[0140] 2. Actual call

[0141] The user starts a call and silently says, "The meeting starts at 3:00 PM."

[0142] The device's in-camera captures the user's lip and tongue movements and sends the visual data to a server.

[0143] The server analyzes the visual data and generates the text "The meeting starts at 3pm" based on the training data and context.

[0144] The server converts the text into voice and transmits the voice data to the other terminal.

[0145] The other device plays the received voice data and transmits the user's words to the other device.

[0146] 3. Response from the other person

[0147] The other person will respond by voice, "Understood, I'll join."

[0148] The remote terminal captures the voice data and sends it to the server.

[0149] The server converts the voice data into text and sends it to the user's device.

[0150] The device will then display the received text to the user or play it as audio.

[0151] This allows the system to provide a smooth and secure means of making calls even in quiet environments or where privacy is required. By utilizing past training data and conversation context, the system can generate text with higher accuracy and achieve natural-sounding voice calls.

[0152] Prompt Sentence Examples

[0153] Here are some example prompts to present to a generative AI model:

[0154] User: Use the "Silent Call" system to say silently, "Thank you for your hard work, regarding tomorrow's meeting..."

[0155] Device: Capture lip and tongue movements with the front camera and send the data to the server.

[0156] Server: Analyze the received data, perform syntax analysis and speech synthesis, and send it to the other party as audio.

[0157] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0158] Program processing flow

[0159] Step 1: Capture process

[0160] The user speaks silently.

[0161] The device uses a built-in camera to capture the user's lip and tongue movements in real time, and specifically, it uses a high-resolution camera and a dedicated real-time image processing chip to acquire visual data.

[0162] Input: User's lip and tongue movements

[0163] Output: Visual data (images or video frames)

[0164] Step 2: Send data

[0165] The device encrypts the captured visual data and sends it to a server using the HTTPS protocol.

[0166] The server receives the received visual data.

[0167] Input: Visual data (terminal input data)

[0168] Output: Visual data (received by the server)

[0169] Step 3: Data analysis and text generation

[0170] The server analyzes the received visual data using a high-speed image processing algorithm, specifically, a convolutional neural network (CNN) to extract lip and tongue movements as features.

[0171] The server inputs the features into a recurrent neural network (RNN) model and generates the corresponding text, taking into account the user's past learning data and conversational context.

[0172] Input: Visual data (received data from the server)

[0173] Output: The generated text

[0174] Step 4: Text-to-speech

[0175] The server converts the generated text into voice data using TTS (Text to Speech) technology, for example, by using a "voice synthesis engine."

[0176] The server transmits the generated voice data to the other terminal in real time.

[0177] Input: Generated text

[0178] Output: Audio data

[0179] Step 5: Processing the other party's response

[0180] The other terminal captures the voice data uttered by the other user and transmits it to the server.

[0181] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology, for example, by using a "voice recognition engine."

[0182] The server sends the converted text to the user terminal.

[0183] Input: Voice data (the other person's input data)

[0184] Output: The converted text

[0185] Step 6: Display to the user

[0186] The device displays the received text to the user or plays it as audio using TTS technology.

[0187] Input: Translated text

[0188] Output: Text or audio displayed to the user

[0189] This allows the overall processing flow to be specified and the operations and data flow at each step to be clearly explained.

[0190] (Application example 1)

[0191] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0192] In security work, it is necessary to be able to make calls and share information smoothly even in quiet environments and when privacy is required. Current voice communication systems have the risk of information leaking to those around them due to the sound they emit, and are difficult to use in quiet environments. A system that can solve these problems is needed.

[0193] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0194] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for capturing speech data from the other party and converting it into text. This enables smooth and safe silent communication and information sharing even in security work that requires quiet environments and privacy.

[0195] "User" means an individual who uses the System to make silent calls.

[0196] "Lip and tongue movements" refers to the movements inside and outside the mouth that a user makes when silently speaking.

[0197] "Capture" refers to the act of recording visual data using an in-camera or other device.

[0198] "Text generation" refers to the process of analyzing captured visual data and creating corresponding string data.

[0199] "Text-to-speech" refers to the technique of converting generated text into audio data.

[0200] "Other device" refers to the device used by the person with whom you are making a call.

[0201] "Server" refers to a central device that performs processing such as data analysis, text generation, and speech conversion.

[0202] "Silent security call application" refers to software that enables silent calls in security work where quiet environments and privacy are required.

[0203] This invention relates to a "Silent Security Call Application" that enables users to make silent calls in situations where quiet environments and privacy protection are required in security work. The following is a detailed explanation of how to specifically implement this system.

[0204] System program generation

[0205] The system achieves the following main functions:

[0206] 1. Hardware Configuration

[0207] User device: Use a device equipped with an in-camera, such as a smartphone or smart glasses.

[0208] Server: Use a high-performance computer server to perform data analysis, text generation, and speech conversion.

[0209] 2. Software Configuration

[0210] OpenCV: A library used to capture the user's lip and tongue movements.

[0211] gTTS (Google Text-to-Speech): A library for converting generated text to speech.

[0212] speech_recognition: A library for converting speech into text.

[0213] Program processing

[0214] 1. Data Capture

[0215] On the device: The front camera captures the user's lip and tongue movements in real time.

[0216] Data transmission: The captured visual data is transmitted from the device to the server.

[0217] 2. Text Generation

[0218] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[0219] 3. Audio conversion

[0220] Server: Converts the generated text into speech data using the gTTS library.

[0221] Data transmission: The generated voice data is transmitted to other devices in real time.

[0222] 4. Processing replies from the other party

[0223] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0224] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[0225] Specific examples

[0226] 1. User training

[0227] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0228] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[0229] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[0230] 2. Actual call

[0231] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0232] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[0233] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[0234] Server: Converts text into speech and sends the speech data to the other device.

[0235] The other device: Plays the received voice data and transmits the user's words to the other device.

[0236] 3. Response from the other person

[0237] Other party: Reply verbally with "Got it, I'll join."

[0238] Remote device: Captures audio data and sends it to the server.

[0239] Server: Converts voice data into text and sends it to the user's device.

[0240] On the device: The received text is displayed to the user or played as audio.

[0241] Prompt Sentence Examples

[0242] Your application will analyze the user's lip movements to generate text and convert it to speech. The following requirements must be met:

[0243] The user makes a call without speaking.

[0244] The capture device uses the in-camera.

[0245] The data is sent to a server and converted into text and audio data.

[0246] The software used is Python's OpenCV, gTTS, and speech_recognition libraries.

[0247] This makes it possible for the system to provide a means for smooth, safe, and silent communication and information sharing, even in quiet environments or places where privacy is required.

[0248] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0249] Step 1:

[0250] The user speaks silently. The user's lip and tongue movements are captured in real time by the in-camera. The input is the user's lip and tongue movements, and the output is visual data. The device records this visual data.

[0251] Step 2:

[0252] The terminal transmits the captured visual data to the server, where it is sent in the form of a data package that reaches the server over the network. The input is the visual data, and the output is the data package received at the server side.

[0253] Step 3:

[0254] The server analyzes the received visual data and generates corresponding text. To analyze the visual data, it uses machine learning models, which utilize the user's past learning data and conversational context. The input is the visual data, and the output is the generated text.

[0255] Step 4:

[0256] The server converts the generated text into speech data using speech synthesis technology and the gTTS (Google Text-to-Speech) library. The input is text data and the output is speech data.

[0257] Step 5:

[0258] The server then transmits the generated audio data to other devices in real time. This must be done quickly over the network, in the form of a data package. The input is the audio data, and the output is an audio file that can be played on other devices.

[0259] Step 6:

[0260] The other device plays back the received voice data and conveys what the other party said to the user. The input is voice data, and the output is the played-back voice.

[0261] Step 7:

[0262] The other party responds by voice. The other party's device captures the voice data and sends it to the server. The input is the other party's voice, and the output is the voice data sent to the server.

[0263] Step 8:

[0264] The server converts the received voice data into text using voice recognition technology. The input is voice data and the output is text data.

[0265] Step 9:

[0266] The terminal displays the text data sent from the server to the user or plays it aloud. The input is the text data, and the output is the displayed text or played audio.

[0267] This allows the entire system to smoothly communicate and share information in security operations without making any sound.

[0268] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0269] This invention relates to a "Silent Call" system that allows users to make calls without speaking. This system achieves more natural conversations by capturing the user's lip and tongue movements and combining them with an emotion engine. Below, we will provide an overview of the system's program and its specific processing in natural language.

[0270] System configuration

[0271] Capture device and emotion recognition engine

[0272] User: Speak silently.

[0273] On-device: Uses the front camera to capture the user's lip and tongue movements in real time, and uses an emotion recognition engine to analyze the user's emotions from their facial expressions.

[0274] Device: The captured movement and facial expression data is stored as visual data.

[0275] Data transmission and analysis

[0276] Terminal: Sends captured visual data to the server.

[0277] Server: Analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, utilizing the user's past learning data and conversational context.

[0278] Text-to-speech

[0279] Server: Based on the generated text and the recognized emotion data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[0280] Server: Sends the generated audio data to other devices in real time.

[0281] Response processing from the other party

[0282] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0283] Server: Converts the received voice data into text using speech recognition technology and displays it to the user or plays it back as audio.

[0284] Specific examples

[0285] 1. User training

[0286] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0287] Device: The front camera captures lip and tongue movements and sends visual data to the server. It also captures facial expressions and sends them as emotional data.

[0288] Server: Analyzes the received data and saves the text "I am currently in Tokyo" and the corresponding visual and emotional data as training data.

[0289] 2. Actual call

[0290] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0291] Device: The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[0292] Server: Analyzes visual and emotional data and generates text such as "The meeting starts at 3 p.m." based on the training data and context.

[0293] Server: Converts text into speech using emotion-based speech synthesis technology and sends the speech data to the other device.

[0294] The other device: Plays the received voice data and transmits the user's words to the other device.

[0295] 3. Response from the other person

[0296] Other party: Reply verbally with "Got it, I'll join."

[0297] Remote device: Captures audio data and sends it to the server.

[0298] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0299] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[0300] This system provides a means for smooth and secure communication even in quiet environments or places where privacy is required. By analyzing emotional data from the user's facial expressions and reflecting emotions in text generation and speech conversion, it is possible to achieve more natural and human-like communication.

[0301] The processing flow will be explained below.

[0302] Training Data Collection Processing Steps

[0303] Step 1: Training Initial Setup

[0304] User: Select "Silent Call" and start training.

[0305] Terminal: Presents training example sentences to the user.

[0306] Step 2: Silent reading of sample sentences

[0307] User: Silently read the provided sample sentences aloud.

[0308] On the device: The front camera captures the user's lip and tongue movements, as well as facial expressions to obtain emotional data.

[0309] Step 3: Sending visual data

[0310] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[0311] Step 4: Data analysis and training

[0312] Server: Analyzes the received visual and emotional data and saves the training data as data corresponding to the text.

[0313] Actual call processing steps

[0314] Step 1: Start the call

[0315] User: Select the person to call and start the call.

[0316] Device: Activate the front camera and emotion recognition engine and prepare.

[0317] Step 2: Silent speech

[0318] User: Speak silently.

[0319] On the device: The front camera captures the user's lip and tongue movements in real time, while the emotion engine simultaneously captures facial expression data.

[0320] Step 3: Sending visual data

[0321] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[0322] Step 4: Text Generation

[0323] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[0324] Step 5: Analyze the sentiment data

[0325] Server: Analyzes the user's emotions from facial expression data and assigns the recognized emotions to text data.

[0326] Step 6: Audio conversion

[0327] Server: The text generated reflecting emotions is converted into voice data using speech synthesis technology.

[0328] Step 7: Sending audio data

[0329] Server: Sends the generated voice data to the other device in real time.

[0330] Step 8: Playing Audio

[0331] The other device: Plays the received voice data and transmits the user's words to the other device.

[0332] Steps for processing a reply from the other party

[0333] Step 1: Capture audio data

[0334] Other party: Speak as if you were making a normal voice call.

[0335] Remote device: Captures the remote device's voice data and sends it to the server.

[0336] Step 2: Analyzing the audio data

[0337] Server: Converts the received voice data into text using voice recognition technology.

[0338] Step 3: Send a text

[0339] Server: Sends the generated text to the user's device and also analyzes emotion data if necessary.

[0340] Step 4: Display text or play audio

[0341] Terminal: Receives text and emotion data and displays it to the user or plays it back as audio.

[0342] The above is the flow of specific processing steps of the invention that combines an emotion engine. This system enables smooth and safe calls even in quiet environments or places where privacy is required, and also enables natural communication that reflects emotions.

[0343] Example 2

[0344] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0345] Conventional voice communication systems have the problem that it is difficult for users to communicate smoothly in quiet environments or places where privacy is required, and they also have the problem that it is difficult to express emotions, making it difficult to realize natural conversations.

[0346] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0347] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for analyzing the user's facial expression to obtain emotional data, means for generating speech data reflecting the emotion based on the generated text and the obtained emotional data, and means for transmitting the generated speech data to another terminal in real time, thereby enabling smooth and natural conversation even in quiet environments or places where privacy is required.

[0348] "User" refers to an individual who uses the system to make silent speech.

[0349] "Means for capturing lip and tongue movements" refers to a device equipped with a camera and / or sensors for capturing the user's lip and tongue movements in real time.

[0350] "Means for generating text based on captured movements" refers to algorithms or software that analyze captured lip and tongue movements and generate corresponding text data.

[0351] "Generated text-to-speech means" refers to an algorithm or program that uses speech synthesis technology to convert text data into speech data.

[0352] "Means of analyzing facial expressions to obtain emotional data" refers to technology and software that captures a user's facial expressions and recognizes and analyzes emotions from those expressions.

[0353] "Means for generating voice data that reflects emotions based on acquired emotional data" refers to algorithms and programs for synthesizing voice that reflects emotional data.

[0354] "Means for transmitting generated voice data to other terminals in real time" refers to communication technologies and protocols for delivering voice data to other terminals without delay.

[0355] The present invention relates to a "Silent Call" system that enables users to have natural conversations without vocalizations in quiet environments or situations where privacy is required. This system acquires emotional data by capturing the movements of the user's lips and tongue and analyzing their facial expressions, and combines them to generate text and voice data for voice calls.

[0356] System configuration

[0357] Capture device and emotion recognition engine

[0358] User: Speaks silently. Imagine a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[0359] Device:

[0360] An in-camera (e.g., a commercially available high-resolution camera) is used to capture the user's lip and tongue movements in real time, and an emotion recognition engine (e.g., emotion analysis software) is used to analyze the user's emotions from their facial expressions and obtain emotion data.

[0361] The captured movement and emotion data is stored as visual data.

[0362] Hardware used:

[0363] High-resolution in-camera, microprocessor

[0364] Software used:

[0365] Sentiment analysis software, data analysis algorithms

[0366] Text generation and speech synthesis

[0367] Device:

[0368] The captured visual and emotional data is sent to a server, using an internet connection for data communication.

[0369] server:

[0370] The received visual data is analyzed and the text "Hello, how are you?" is generated from lip and tongue movements and facial expressions using a generative AI model (e.g., a machine learning model, as the term is generally used).

[0371] Based on the generated text and the acquired emotional data, it is converted into voice data using emotion-reflecting voice synthesis technology (commonly known as voice synthesis software).

[0372] Hardware used:

[0373] Server computers, high-speed data processing units

[0374] Software used:

[0375] Machine learning models, speech synthesis software

[0376] Communication and playback

[0377] server:

[0378] The generated voice data is sent to other devices in real time using a real-time protocol.

[0379] Remote device:

[0380] The received voice data is played back and the user's silent speech is conveyed to the other party.

[0381] Hardware used:

[0382] Communication modem, speaker

[0383] Specific examples

[0384] 1. User training

[0385] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0386] Device:

[0387] The front camera captures lip and tongue movements and sends visual data to the server, as well as facial expressions, which are then sent as emotional data.

[0388] server:

[0389] The received data is analyzed, and the visual and emotional data corresponding to the text "I am in Tokyo now" are saved as training data.

[0390] 2. Actual call

[0391] User: Starts a call and silently says "My meeting starts at 3 PM."

[0392] Device:

[0393] The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[0394] server:

[0395] It analyzes visual and emotional data and generates the text "The meeting starts at 3 p.m." based on the training data and context.

[0396] server:

[0397] Text is converted into speech using emotion-reflecting speech synthesis technology, and the speech data is sent to the other party's terminal.

[0398] Remote device:

[0399] The received voice data is played back and the user's words are conveyed to the other party.

[0400] 3. Response from the other person

[0401] Other party: Reply verbally with "Got it, I'll join."

[0402] Remote device: Captures audio data and sends it to the server.

[0403] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0404] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[0405] This system enables smooth and natural conversations even in quiet environments or places where privacy is required. It also analyzes emotional data from the user's facial expressions and reflects those emotions in text generation and speech conversion, enabling more human-like communication.

[0406] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0407] Step 1:

[0408] User: Speaks silently. For example, consider a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[0409] Input: The user's lip and tongue movements.

[0410] Output: Real-time lip and tongue movement data.

[0411] Specific actions: The user speaks silently in front of the camera.

[0412] Step 2:

[0413] On the device: The front camera captures the user's lip and tongue movements in real time, and an emotion recognition engine is used to analyze facial expressions and obtain emotional data.

[0414] Input: Real-time lip and tongue movement data, facial expression data.

[0415] Output: Captured movement and emotion data are saved as visual data.

[0416] What it does: The camera and emotion recognition engine work together to record the user's lips, tongue, and facial expressions.

[0417] Step 3:

[0418] Device: The saved visual and emotional data is sent to the server. Data communication is via an internet connection.

[0419] Input: visual and emotional data.

[0420] Output: Sending data to the server.

[0421] What happens: Your device uploads data to a server via your internet connection.

[0422] Step 4:

[0423] Server: Analyzes the received visual data and generates text from lip and tongue movements.

[0424] Input: Captured visual data, emotion data.

[0425] Output: The generated text.

[0426] What it does: Uses a generative AI model to generate the text "Hello, how are you?" from lip and tongue movements.

[0427] Step 5:

[0428] Server: Based on the generated text and the acquired emotional data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[0429] Input: Text data, emotion data.

[0430] Output: Emotionally-informed audio data.

[0431] Specific operation: Create emotionally-charged speech data using a text generation model.

[0432] Step 6:

[0433] Server: The generated voice data is sent to other devices in real time using a real-time protocol.

[0434] Input: Emotionally-charged speech data.

[0435] Output: Audio data sent to other devices.

[0436] Specific operation: Sends voice data to a remote terminal using a real-time protocol.

[0437] Step 7:

[0438] The other party's terminal: Plays back the received voice data and conveys the user's silent speech to the other party.

[0439] Input: Audio data sent from the server.

[0440] Output: The audio played.

[0441] Specific operation: Audio is output from the speaker and the receiver listens to the user's speech.

[0442] Step 8:

[0443] Recipient: "Yes, I'm fine" replies in a normal voice.

[0444] Input: The other person's speech.

[0445] Output: Audio data.

[0446] Specific operation: Convert analog voice into digital data and input it into the terminal.

[0447] Step 9:

[0448] Remote device: Captures the remote device's voice data and sends it to the server.

[0449] Input: The other party's voice data.

[0450] Output: Sends audio data to the server.

[0451] Specific operation: After capturing the audio, send the data to the server.

[0452] Step 10:

[0453] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0454] Input: The other party's voice data.

[0455] Output: Text data, emotion data.

[0456] Specific operation: Converts speech into text using speech recognition technology and analyzes emotional data.

[0457] Step 11:

[0458] Device: Receives text and emotion data and displays it to the user, or plays back a synthesized voice that reflects the emotion.

[0459] Input: Text data, emotion data.

[0460] Output: The displayed text or played audio.

[0461] Specific Action: Display the text to the user or play it as emotional speech.

[0462] (Application example 2)

[0463] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0464] In modern society, there is a growing demand for calls and communication in situations where confidentiality and privacy are essential. However, conventional voice calls pose the risk of sound leakage and eavesdropping, making them unsuitable for use in quiet environments or sensitive situations. There is also a need for a method to convey text information to the other party more accurately and emotionally. Technology is needed to solve this problem and enable emotionally charged, natural communication even in silent environments.

[0465] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for capturing the movements of the user's lips and tongue, means for generating text based on the captured movements, means for recognizing the user's emotions and reflecting them in the generated text, and means for converting the generated text into speech. This makes it possible to accurately generate text from the movements of the lips and tongue even in a silent environment, enabling voice calls that reflect the user's emotions.

[0466] "Means for capturing the movement of a user's lips and tongue" refers to devices or technologies that use an in-camera or specific sensors to capture the movement and shape of the area around a user's mouth in real time and record it as digital data.

[0467] "Means for generating text from captured movements" means algorithms and software for analyzing captured lip and tongue movement data and converting it into corresponding characters, words, or sentences.

[0468] "Means for recognizing user emotions and reflecting them in text generation" refers to technology that uses image analysis technology and facial expression recognition engines to detect emotions from the user's facial expressions and incorporate that emotional information into the text generation process.

[0469] "Means for converting generated text into speech" refers to software and hardware that utilizes speech synthesis technology using artificial intelligence to convert text data into speech data and generate natural-sounding speech that reflects emotion.

[0470] "Means for improving the accuracy of text generation by utilizing conversational context and the user's past learning data" refers to algorithms and databases that utilize the user's past conversation history and contextual information from dialogue to improve the content and accuracy of the generated text.

[0471] "Means for transmitting the generated voice data and emotional data to another terminal in real time and conversing with the other party" refers to communication technologies and protocols that enable real-time communication by quickly transmitting the generated voice data and emotional information to another device via a network.

[0472] The "Secure Silent Call System" that realizes this application example captures the movement of the user's lips and tongue, generates text based on that movement, recognizes the user's emotions and reflects them in the generated text, converts the generated text into speech, and transmits it to other devices in real time, enabling communication.

[0473] System configuration

[0474] 1. Capture device and emotion recognition engine

[0475] user:

[0476] This feature is intended for use in quiet environments such as conference rooms and libraries.

[0477] Device:

[0478] The system uses a built-in camera and specific sensors to capture the user's lip and tongue movements in real time. It also detects facial expressions and analyzes the user's emotions using an emotion recognition engine. The hardware used includes a standard webcam and a highly accurate facial recognition sensor.

[0479] Device:

[0480] The captured movement and facial expression data are stored as visual data and sent to a server.

[0481] 2. Data submission and analysis

[0482] server:

[0483] The software analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, leveraging the user's past learning data and conversational context. The software includes a lip movement recognition engine (LipMovementRecognizer) and an emotion recognition engine (EmotionRecognizer).

[0484] 3. Text-to-speech

[0485] server:

[0486] The generated text and emotion data are then converted into emotion-reflecting voice data using speech synthesis technology. This process is carried out using TextToSpeech software.

[0487] server:

[0488] The generated voice data is transmitted to another terminal in real time.

[0489] 4. Processing replies from the other party

[0490] Remote device:

[0491] As with a normal voice call, the voice data from the other party is captured and sent to the server.

[0492] server:

[0493] The received voice data is converted into text using voice recognition technology, and the text and emotional data are sent to the user's device.

[0494] Device:

[0495] It receives text and emotion data and displays it to the user. It also synthesizes the text into speech and plays it back with the emotion reflected.

[0496] Specific examples

[0497] 1. User training:

[0498] The user silently reads the training sample sentence, "I am in Tokyo now." The device captures lip and tongue movements with the front camera and sends the visual data to the server. It also captures facial expressions and sends them as emotional data. The server analyzes the received data and saves the visual and emotional data corresponding to the text "I am in Tokyo now" as training data.

[0499] 2. Actual call:

[0500] The user silently utters, "The meeting will start at 3 PM." The device captures lip and tongue movements with the in-camera and sends the visual data to the server. It also sends facial expression data for emotion recognition. The server analyzes the visual and emotional data and generates the text "The meeting will start at 3 PM" based on the learning data and context. It then converts this into speech using emotion-reflecting speech synthesis technology and sends it to the other device. The other device receives the voice data and conveys the user's remarks to the other party.

[0501] 3. Response from the other person:

[0502] The other party responds by saying "Got it. I'll join." The other party's device captures the voice data and sends it to the server. The server converts the voice data into text and sends the text and emotion data to the user's device. The device receives the text and emotion data and displays it to the user or plays it back as voice.

[0503] Example prompt for a generative AI model:

[0504] "Analyze the user's lip and tongue movements and generate text. Use the following video stream data. Also, recognize emotions from facial expressions and generate audio data that reflects them. Video stream data example: [Image data] Emotion data example: [Emotion label based on image data]"

[0505] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0506] Step 1:

[0507] The user silently speaks. For example, in a quiet environment, the user silently utters, "The meeting will start at 3:00 PM." At this time, the user moves their mouth in the same way as they would during a normal phone call.

[0508] Step 2:

[0509] The device uses an in-camera and specific sensors to capture the user's lip and tongue movements in real time. The video data captured by the in-camera is input, and visual data for analyzing lip and tongue movements is output. The device also captures the user's facial expressions. This facial expression data is also output as visual data.

[0510] Step 3:

[0511] The terminal transmits the captured visual data to the server. The input is the captured visual data, and the output is the data transmission to the server. Network communication technology is used for data transmission.

[0512] Step 4:

[0513] The server analyzes the received visual data. Specifically, it analyzes lip and tongue movement data and facial expression data to generate text data. The input is the received visual data, and the output is the text data generated by the data analysis process. LipMovementRecognizer and EmotionRecognizer are used for the analysis.

[0514] Step 5:

[0515] The server uses the conversational context and the user's past learning data to improve the accuracy of the generated text. The input is the initial text data, the user's past learning data, and the conversational context, and the output is the final text data with high accuracy. Natural language processing technology is used for data processing.

[0516] Step 6:

[0517] The server converts the final text data and emotion data into speech. The input is the final text data and emotion data, and the output is speech data that reflects the emotion expression. TextToSpeech software is used for speech synthesis.

[0518] Step 7:

[0519] The server transmits the generated voice data to the remote terminal in real time. The input is voice data, and the output is voice data transmitted over the network. A communication protocol is used for transmission.

[0520] Step 8:

[0521] The other device plays back the received voice data and transmits the user's speech to the other device. The input is the received voice data, and the output is the played voice.

[0522] Step 9:

[0523] The other party responds with voice, just like in a normal voice call. For example, they might respond with, "Got it. I'll join." This voice is captured by the other party's device.

[0524] Step 10:

[0525] The remote terminal transmits the captured voice data to the server. The input is the voice data, and the output is the data transmission to the server.

[0526] Step 11:

[0527] The server analyzes the received voice data and converts it into text data. The input is voice data and the output is text data. The analysis is performed using voice recognition technology.

[0528] Step 12:

[0529] The server sends the generated text data and emotion data to the user terminal. The input is the text data and emotion data, and the output is the data transmission to the user terminal.

[0530] Step 13:

[0531] The terminal displays the received text data and emotion data to the user, and also plays the text as speech that reflects the emotion. The input is the text data and emotion data, and the output is the displayed text and the played speech.

[0532] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the 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.

[0533] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0534] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0535] [Second embodiment]

[0536] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0537] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0538] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0539] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0540] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0541] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0542] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0543] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0544] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[0545] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0546] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0547] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[0548] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. The system captures the movements of the user's lips and tongue, generates text based on the captured movements, and converts the text into speech. Below, we will provide an overview of the system's program and its specific processing in natural language.

[0549] System configuration

[0550] Capture Device

[0551] User: Speak silently.

[0552] On the device: The front camera is used to capture the user's lip and tongue movements in real time, which are then stored as visual data.

[0553] Data transmission and analysis

[0554] Terminal: Sends captured visual data to the server.

[0555] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[0556] Text-to-speech

[0557] Server: The generated text is converted into voice data using speech synthesis technology.

[0558] Server: Sends the generated audio data to other devices in real time.

[0559] Response processing from the other party

[0560] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0561] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[0562] Specific examples

[0563] 1. User training

[0564] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0565] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[0566] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[0567] 2. Actual call

[0568] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0569] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[0570] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[0571] Server: Converts text into speech and sends the speech data to the other device.

[0572] The other device: Plays the received voice data and transmits the user's words to the other device.

[0573] 3. Response from the other person

[0574] Other party: Reply verbally with "Got it, I'll join."

[0575] Remote device: Captures audio data and sends it to the server.

[0576] Server: Converts voice data into text and sends it to the user's device.

[0577] On the device: The received text is displayed to the user or played as audio.

[0578] This allows the system to provide a smooth and secure means of making calls even in quiet environments or places where privacy is required. By utilizing the user's past learning data and conversation context, the system can generate text with higher accuracy and realize natural-sounding voice calls.

[0579] The processing flow will be explained below.

[0580] Training Data Collection Processing Steps

[0581] Step 1: Training Initial Setup

[0582] User: Select "Silent Calling."

[0583] Terminal: Presents training sample sentences to the user and prepares them to begin training.

[0584] Step 2: Silent reading of sample sentences

[0585] User: Silently read the provided sample sentences aloud.

[0586] On your device: Use the front camera to capture the movements of your lips and tongue.

[0587] Step 3: Sending visual data

[0588] Terminal: Sends captured visual data to the server.

[0589] Step 4: Data analysis and training

[0590] Server: Analyzes the received visual data and stores the lip and tongue movements and corresponding text as training data.

[0591] Actual call processing steps

[0592] Step 1: Start the call

[0593] User: Select the person to call and start the call.

[0594] Device: Prepare for the call and activate the front camera.

[0595] Step 2: Silent speech

[0596] User: Speak silently.

[0597] On the device: Uses the front camera to capture the user's lip and tongue movements in real time.

[0598] Step 3: Sending visual data

[0599] Terminal: Sends captured visual data to the server.

[0600] Step 4: Text Generation

[0601] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[0602] Step 5: Audio conversion

[0603] Server: The generated text is converted into voice data using speech synthesis technology.

[0604] Step 6: Sending audio data

[0605] Server: Sends the generated voice data to the other device.

[0606] Step 7: Playing Audio

[0607] The other device: Plays the received voice data and transmits the user's words to the other device.

[0608] Steps for processing a reply from the other party

[0609] Step 1: Capture audio data

[0610] Recipient: Respond by voice as usual.

[0611] Remote device: Captures the response audio data and sends it to the server.

[0612] Step 2: Analyzing the audio data

[0613] Server: Converts the received voice data into text using voice recognition technology.

[0614] Step 3: Send a text

[0615] Server: Sends the textual response to the user's device.

[0616] Step 4: Display text or play audio

[0617] Device: Displays the received text to the user, or synthesizes the text to speech and plays it through headphones.

[0618] This is a step-by-step explanation of the program's processing. Through this series of processes, the system enables smooth and safe calls even in quiet environments or places where privacy is required.

[0619] Example 1

[0620] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0621] Currently, voice communication relies on direct voice communication, which makes it difficult to use in quiet environments or places where privacy is required. Voice communication is also difficult for people with hearing impairments. Furthermore, conventional voice input systems have problems with misrecognition due to background noise and differences in speaking speed.

[0622] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0623] In this invention, the server includes means for capturing the user's lip and tongue movements, means for generating text based on the captured movements, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for converting the speech data received from the other terminal into text. This enables silent calls, allowing for smooth and safe calls even in quiet environments or places where privacy is required. It also provides an alternative means of voice calls for people with hearing impairments, enabling highly accurate calls unaffected by background noise or speech rate.

[0624] "Capturing means" refers to devices and software that detect the user's lip and tongue movements in real time and acquire and store them as visual data.

[0625] The "means for generating text" refers to a device and software for analyzing the captured visual data and generating text data based on the content of the user's silent utterances.

[0626] The "means for converting to speech" refers to devices and software for converting the generated text data into synthetic speech. Specifically, speech synthesis technology is used.

[0627] The "transmitting means" refers to a device and software for transmitting the generated voice data to another terminal via a network.

[0628] The "means for converting into text" refers to the device and software that analyzes the voice data received from the other party's terminal and converts it into text data. Specifically, it uses voice recognition technology.

[0629] "Conversational context" refers to information and processing methods for generating appropriate text based on the context of the user's statements and the flow of the dialogue.

[0630] "Past learning data" refers to data accumulated based on the user's past training and conversation content, and is the data and processing method used to improve the accuracy of text generation.

[0631] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. This system provides the function of capturing the movements of the user's lips and tongue, generating text based on the captured movements, and converting the text into speech. It is also possible to convert the speech of the other party into text and provide it to the user.

[0632] System configuration

[0633] Capture Device

[0634] The user speaks silently, and the device uses the front camera to capture the user's lip and tongue movements in real time. The captured movements are saved as visual data. The front camera uses a high-resolution camera and a dedicated chip for real-time image processing.

[0635] Data transmission and analysis

[0636] The device sends the captured visual data to a server, where it is encrypted and transmitted securely using the HTTPS protocol.

[0637] The server analyzes the received visual data and generates corresponding text using CNN (Convolutional Neural Network) and RNN (Recurrent Neural Network) technologies, generating highly accurate text based on the user's past learning data and conversational context.

[0638] Text-to-speech

[0639] The server converts the generated text into audio data using TTS (Text to Speech) technology, such as Google Cloud Text-to-Speech. This audio data is generated as a high-quality audio file (e.g., WAV format).

[0640] The server transmits the generated voice data to the other terminal in real time.

[0641] Response processing from the other party

[0642] The other device captures the voice data spoken by the other user and sends it to the server. The voice data is encrypted and sent using the HTTPS protocol.

[0643] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology. In this case, we use "Google Cloud Speech-to-Text."

[0644] The server sends the converted text to the user terminal.

[0645] The device will then display the received text to the user or play it as audio.

[0646] Specific examples

[0647] 1. User training

[0648] The user silently reads the training sample sentence, "I am currently in Tokyo."

[0649] The device's in-camera captures lip and tongue movements and sends the visual data to a server.

[0650] The server analyzes the received data and saves the text "I am in Tokyo now" along with the corresponding visual data as training data.

[0651] 2. Actual call

[0652] The user starts a call and silently says, "The meeting starts at 3:00 PM."

[0653] The device's in-camera captures the user's lip and tongue movements and sends the visual data to a server.

[0654] The server analyzes the visual data and generates the text "The meeting starts at 3pm" based on the training data and context.

[0655] The server converts the text into voice and transmits the voice data to the other terminal.

[0656] The other device plays the received voice data and transmits the user's words to the other device.

[0657] 3. Response from the other person

[0658] The other person will respond by voice, "Understood, I'll join."

[0659] The remote terminal captures the voice data and sends it to the server.

[0660] The server converts the voice data into text and sends it to the user's device.

[0661] The device will then display the received text to the user or play it as audio.

[0662] This allows the system to provide a smooth and secure means of making calls even in quiet environments or where privacy is required. By utilizing past training data and conversation context, the system can generate text with higher accuracy and achieve natural-sounding voice calls.

[0663] Prompt Sentence Examples

[0664] Here are some example prompts to present to a generative AI model:

[0665] User: Use the "Silent Call" system to say silently, "Thank you for your hard work, regarding tomorrow's meeting..."

[0666] Device: Capture lip and tongue movements with the front camera and send the data to the server.

[0667] Server: Analyze the received data, perform syntax analysis and speech synthesis, and send it to the other party as audio.

[0668] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0669] Program processing flow

[0670] Step 1: Capture process

[0671] The user speaks silently.

[0672] The device uses a built-in camera to capture the user's lip and tongue movements in real time, and specifically, it uses a high-resolution camera and a dedicated real-time image processing chip to acquire visual data.

[0673] Input: User's lip and tongue movements

[0674] Output: Visual data (images or video frames)

[0675] Step 2: Send data

[0676] The device encrypts the captured visual data and sends it to a server using the HTTPS protocol.

[0677] The server receives the received visual data.

[0678] Input: Visual data (terminal input data)

[0679] Output: Visual data (received by the server)

[0680] Step 3: Data analysis and text generation

[0681] The server analyzes the received visual data using a high-speed image processing algorithm, specifically, a convolutional neural network (CNN) to extract lip and tongue movements as features.

[0682] The server inputs the features into a recurrent neural network (RNN) model and generates the corresponding text, taking into account the user's past learning data and conversational context.

[0683] Input: Visual data (received data from the server)

[0684] Output: The generated text

[0685] Step 4: Text-to-speech

[0686] The server converts the generated text into voice data using TTS (Text to Speech) technology, for example, by using a "voice synthesis engine."

[0687] The server transmits the generated voice data to the other terminal in real time.

[0688] Input: Generated text

[0689] Output: Audio data

[0690] Step 5: Processing the other party's response

[0691] The other terminal captures the voice data uttered by the other user and transmits it to the server.

[0692] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology, for example, by using a "voice recognition engine."

[0693] The server sends the converted text to the user terminal.

[0694] Input: Voice data (the other person's input data)

[0695] Output: The converted text

[0696] Step 6: Display to the user

[0697] The device displays the received text to the user or plays it as audio using TTS technology.

[0698] Input: Translated text

[0699] Output: Text or audio displayed to the user

[0700] This allows the overall processing flow to be specified and the operations and data flow at each step to be clearly explained.

[0701] (Application example 1)

[0702] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0703] In security work, it is necessary to be able to make calls and share information smoothly even in quiet environments and when privacy is required. Current voice communication systems have the risk of information leaking to those around them due to the sound they emit, and are difficult to use in quiet environments. A system that can solve these problems is needed.

[0704] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0705] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for capturing speech data from the other party and converting it into text. This enables smooth and safe silent communication and information sharing even in security work that requires quiet environments and privacy.

[0706] "User" means an individual who uses the System to make silent calls.

[0707] "Lip and tongue movements" refers to the movements inside and outside the mouth that a user makes when silently speaking.

[0708] "Capture" refers to the act of recording visual data using an in-camera or other device.

[0709] "Text generation" refers to the process of analyzing captured visual data and creating corresponding string data.

[0710] "Text-to-speech" refers to the technique of converting generated text into audio data.

[0711] "Other device" refers to the device used by the person with whom you are making a call.

[0712] "Server" refers to a central device that performs processing such as data analysis, text generation, and speech conversion.

[0713] "Silent security call application" refers to software that enables silent calls in security work where quiet environments and privacy are required.

[0714] This invention relates to a "Silent Security Call Application" that enables users to make silent calls in situations where quiet environments and privacy protection are required in security work. The following is a detailed explanation of how to specifically implement this system.

[0715] System program generation

[0716] The system achieves the following main functions:

[0717] 1. Hardware Configuration

[0718] User device: Use a device equipped with an in-camera, such as a smartphone or smart glasses.

[0719] Server: Use a high-performance computer server to perform data analysis, text generation, and speech conversion.

[0720] 2. Software Configuration

[0721] OpenCV: A library used to capture the user's lip and tongue movements.

[0722] gTTS (Google Text-to-Speech): A library for converting generated text to speech.

[0723] speech_recognition: A library for converting speech into text.

[0724] Program processing

[0725] 1. Data Capture

[0726] On the device: The front camera captures the user's lip and tongue movements in real time.

[0727] Data transmission: The captured visual data is transmitted from the device to the server.

[0728] 2. Text Generation

[0729] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[0730] 3. Audio conversion

[0731] Server: Converts the generated text into speech data using the gTTS library.

[0732] Data transmission: The generated voice data is transmitted to other devices in real time.

[0733] 4. Processing replies from the other party

[0734] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0735] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[0736] Specific examples

[0737] 1. User training

[0738] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0739] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[0740] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[0741] 2. Actual call

[0742] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0743] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[0744] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[0745] Server: Converts text into speech and sends the speech data to the other device.

[0746] The other device: Plays the received voice data and transmits the user's words to the other device.

[0747] 3. Response from the other person

[0748] Other party: Reply verbally with "Got it, I'll join."

[0749] Remote device: Captures audio data and sends it to the server.

[0750] Server: Converts voice data into text and sends it to the user's device.

[0751] On the device: The received text is displayed to the user or played as audio.

[0752] Prompt Sentence Examples

[0753] Your application will analyze the user's lip movements to generate text and convert it to speech. The following requirements must be met:

[0754] The user makes a call without speaking.

[0755] The capture device uses the in-camera.

[0756] The data is sent to a server and converted into text and audio data.

[0757] The software used is Python's OpenCV, gTTS, and speech_recognition libraries.

[0758] This makes it possible for the system to provide a means for smooth, safe, and silent communication and information sharing, even in quiet environments or places where privacy is required.

[0759] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0760] Step 1:

[0761] The user speaks silently. The user's lip and tongue movements are captured in real time by the in-camera. The input is the user's lip and tongue movements, and the output is visual data. The device records this visual data.

[0762] Step 2:

[0763] The terminal transmits the captured visual data to the server, where it is sent in the form of a data package that reaches the server over the network. The input is the visual data, and the output is the data package received at the server side.

[0764] Step 3:

[0765] The server analyzes the received visual data and generates corresponding text. To analyze the visual data, it uses machine learning models, which utilize the user's past learning data and conversational context. The input is the visual data, and the output is the generated text.

[0766] Step 4:

[0767] The server converts the generated text into speech data using speech synthesis technology and the gTTS (Google Text-to-Speech) library. The input is text data and the output is speech data.

[0768] Step 5:

[0769] The server then transmits the generated audio data to other devices in real time. This must be done quickly over the network, in the form of a data package. The input is the audio data, and the output is an audio file that can be played on other devices.

[0770] Step 6:

[0771] The other device plays back the received voice data and conveys what the other party said to the user. The input is voice data, and the output is the played-back voice.

[0772] Step 7:

[0773] The other party responds by voice. The other party's device captures the voice data and sends it to the server. The input is the other party's voice, and the output is the voice data sent to the server.

[0774] Step 8:

[0775] The server converts the received voice data into text using voice recognition technology. The input is voice data and the output is text data.

[0776] Step 9:

[0777] The terminal displays the text data sent from the server to the user or plays it aloud. The input is the text data, and the output is the displayed text or played audio.

[0778] This allows the entire system to smoothly communicate and share information in security operations without making any sound.

[0779] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0780] This invention relates to a "Silent Call" system that allows users to make calls without speaking. This system achieves more natural conversations by capturing the user's lip and tongue movements and combining them with an emotion engine. Below, we will provide an overview of the system's program and its specific processing in natural language.

[0781] System configuration

[0782] Capture device and emotion recognition engine

[0783] User: Speak silently.

[0784] On-device: Uses the front camera to capture the user's lip and tongue movements in real time, and uses an emotion recognition engine to analyze the user's emotions from their facial expressions.

[0785] Device: The captured movement and facial expression data is stored as visual data.

[0786] Data transmission and analysis

[0787] Terminal: Sends captured visual data to the server.

[0788] Server: Analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, utilizing the user's past learning data and conversational context.

[0789] Text-to-speech

[0790] Server: Based on the generated text and the recognized emotion data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[0791] Server: Sends the generated audio data to other devices in real time.

[0792] Response processing from the other party

[0793] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[0794] Server: Converts the received voice data into text using speech recognition technology and displays it to the user or plays it back as audio.

[0795] Specific examples

[0796] 1. User training

[0797] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0798] Device: The front camera captures lip and tongue movements and sends visual data to the server. It also captures facial expressions and sends them as emotional data.

[0799] Server: Analyzes the received data and saves the text "I am currently in Tokyo" and the corresponding visual and emotional data as training data.

[0800] 2. Actual call

[0801] User: Starts a call and silently says, "My meeting starts at 3 PM."

[0802] Device: The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[0803] Server: Analyzes visual and emotional data and generates text such as "The meeting starts at 3 p.m." based on the training data and context.

[0804] Server: Converts text into speech using emotion-based speech synthesis technology and sends the speech data to the other device.

[0805] The other device: Plays the received voice data and transmits the user's words to the other device.

[0806] 3. Response from the other person

[0807] Other party: Reply verbally with "Got it, I'll join."

[0808] Remote device: Captures audio data and sends it to the server.

[0809] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0810] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[0811] This system provides a means for smooth and secure communication even in quiet environments or places where privacy is required. By analyzing emotional data from the user's facial expressions and reflecting emotions in text generation and speech conversion, it is possible to achieve more natural and human-like communication.

[0812] The processing flow will be explained below.

[0813] Training Data Collection Processing Steps

[0814] Step 1: Training Initial Setup

[0815] User: Select "Silent Call" and start training.

[0816] Terminal: Presents training example sentences to the user.

[0817] Step 2: Silent reading of sample sentences

[0818] User: Silently read the provided sample sentences aloud.

[0819] On the device: The front camera captures the user's lip and tongue movements, as well as facial expressions to obtain emotional data.

[0820] Step 3: Sending visual data

[0821] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[0822] Step 4: Data analysis and training

[0823] Server: Analyzes the received visual and emotional data and saves the training data as data corresponding to the text.

[0824] Actual call processing steps

[0825] Step 1: Start the call

[0826] User: Select the person to call and start the call.

[0827] Device: Activate the front camera and emotion recognition engine and prepare.

[0828] Step 2: Silent speech

[0829] User: Speak silently.

[0830] On the device: The front camera captures the user's lip and tongue movements in real time, while the emotion engine simultaneously captures facial expression data.

[0831] Step 3: Sending visual data

[0832] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[0833] Step 4: Text Generation

[0834] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[0835] Step 5: Analyze the sentiment data

[0836] Server: Analyzes the user's emotions from facial expression data and assigns the recognized emotions to text data.

[0837] Step 6: Audio conversion

[0838] Server: The text generated reflecting emotions is converted into voice data using speech synthesis technology.

[0839] Step 7: Sending audio data

[0840] Server: Sends the generated voice data to the other device in real time.

[0841] Step 8: Playing Audio

[0842] The other device: Plays the received voice data and transmits the user's words to the other device.

[0843] Steps for processing a reply from the other party

[0844] Step 1: Capture audio data

[0845] Other party: Speak as if you were making a normal voice call.

[0846] Remote device: Captures the remote device's voice data and sends it to the server.

[0847] Step 2: Analyzing the audio data

[0848] Server: Converts the received voice data into text using voice recognition technology.

[0849] Step 3: Send a text

[0850] Server: Sends the generated text to the user's device and also analyzes emotion data if necessary.

[0851] Step 4: Display text or play audio

[0852] Terminal: Receives text and emotion data and displays it to the user or plays it back as audio.

[0853] The above is the flow of specific processing steps of the invention that combines an emotion engine. This system enables smooth and safe calls even in quiet environments or places where privacy is required, and also enables natural communication that reflects emotions.

[0854] Example 2

[0855] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0856] Conventional voice communication systems have the problem that it is difficult for users to communicate smoothly in quiet environments or places where privacy is required, and they also have the problem that it is difficult to express emotions, making it difficult to realize natural conversations.

[0857] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[0858] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for analyzing the user's facial expression to obtain emotional data, means for generating speech data reflecting the emotion based on the generated text and the obtained emotional data, and means for transmitting the generated speech data to another terminal in real time, thereby enabling smooth and natural conversation even in quiet environments or places where privacy is required.

[0859] "User" refers to an individual who uses the system to make silent speech.

[0860] "Means for capturing lip and tongue movements" refers to a device equipped with a camera and / or sensors for capturing the user's lip and tongue movements in real time.

[0861] "Means for generating text based on captured movements" refers to algorithms or software that analyze captured lip and tongue movements and generate corresponding text data.

[0862] "Generated text-to-speech means" refers to an algorithm or program that uses speech synthesis technology to convert text data into speech data.

[0863] "Means of analyzing facial expressions to obtain emotional data" refers to technology and software that captures a user's facial expressions and recognizes and analyzes emotions from those expressions.

[0864] "Means for generating voice data that reflects emotions based on acquired emotional data" refers to algorithms and programs for synthesizing voice that reflects emotional data.

[0865] "Means for transmitting generated voice data to other terminals in real time" refers to communication technologies and protocols for delivering voice data to other terminals without delay.

[0866] The present invention relates to a "Silent Call" system that enables users to have natural conversations without vocalizations in quiet environments or situations where privacy is required. This system acquires emotional data by capturing the movements of the user's lips and tongue and analyzing their facial expressions, and combines them to generate text and voice data for voice calls.

[0867] System configuration

[0868] Capture device and emotion recognition engine

[0869] User: Speaks silently. Imagine a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[0870] Device:

[0871] An in-camera (e.g., a commercially available high-resolution camera) is used to capture the user's lip and tongue movements in real time, and an emotion recognition engine (e.g., emotion analysis software) is used to analyze the user's emotions from their facial expressions and obtain emotion data.

[0872] The captured movement and emotion data is stored as visual data.

[0873] Hardware used:

[0874] High-resolution in-camera, microprocessor

[0875] Software used:

[0876] Sentiment analysis software, data analysis algorithms

[0877] Text generation and speech synthesis

[0878] Device:

[0879] The captured visual and emotional data is sent to a server, using an internet connection for data communication.

[0880] server:

[0881] The received visual data is analyzed and the text "Hello, how are you?" is generated from lip and tongue movements and facial expressions using a generative AI model (e.g., a machine learning model, as the term is generally used).

[0882] Based on the generated text and the acquired emotional data, it is converted into voice data using emotion-reflecting voice synthesis technology (commonly known as voice synthesis software).

[0883] Hardware used:

[0884] Server computers, high-speed data processing units

[0885] Software used:

[0886] Machine learning models, speech synthesis software

[0887] Communication and playback

[0888] server:

[0889] The generated voice data is sent to other devices in real time using a real-time protocol.

[0890] Remote device:

[0891] The received voice data is played back and the user's silent speech is conveyed to the other party.

[0892] Hardware used:

[0893] Communication modem, speaker

[0894] Specific examples

[0895] 1. User training

[0896] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[0897] Device:

[0898] The front camera captures lip and tongue movements and sends visual data to the server, as well as facial expressions, which are then sent as emotional data.

[0899] server:

[0900] The received data is analyzed, and the visual and emotional data corresponding to the text "I am in Tokyo now" are saved as training data.

[0901] 2. Actual call

[0902] User: Starts a call and silently says "My meeting starts at 3 PM."

[0903] Device:

[0904] The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[0905] server:

[0906] It analyzes visual and emotional data and generates the text "The meeting starts at 3 p.m." based on the training data and context.

[0907] server:

[0908] Text is converted into speech using emotion-reflecting speech synthesis technology, and the speech data is sent to the other party's terminal.

[0909] Remote device:

[0910] The received voice data is played back and the user's words are conveyed to the other party.

[0911] 3. Response from the other person

[0912] Other party: Reply verbally with "Got it, I'll join."

[0913] Remote device: Captures audio data and sends it to the server.

[0914] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0915] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[0916] This system enables smooth and natural conversations even in quiet environments or places where privacy is required. It also analyzes emotional data from the user's facial expressions and reflects those emotions in text generation and speech conversion, enabling more human-like communication.

[0917] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0918] Step 1:

[0919] User: Speaks silently. For example, consider a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[0920] Input: The user's lip and tongue movements.

[0921] Output: Real-time lip and tongue movement data.

[0922] Specific actions: The user speaks silently in front of the camera.

[0923] Step 2:

[0924] On the device: The front camera captures the user's lip and tongue movements in real time, and an emotion recognition engine is used to analyze facial expressions and obtain emotional data.

[0925] Input: Real-time lip and tongue movement data, facial expression data.

[0926] Output: Captured movement and emotion data are saved as visual data.

[0927] What it does: The camera and emotion recognition engine work together to record the user's lips, tongue, and facial expressions.

[0928] Step 3:

[0929] Device: The saved visual and emotional data is sent to the server. Data communication is via an internet connection.

[0930] Input: visual and emotional data.

[0931] Output: Sending data to the server.

[0932] What happens: Your device uploads data to a server via your internet connection.

[0933] Step 4:

[0934] Server: Analyzes the received visual data and generates text from lip and tongue movements.

[0935] Input: Captured visual data, emotion data.

[0936] Output: The generated text.

[0937] What it does: Uses a generative AI model to generate the text "Hello, how are you?" from lip and tongue movements.

[0938] Step 5:

[0939] Server: Based on the generated text and the acquired emotional data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[0940] Input: Text data, emotion data.

[0941] Output: Emotionally-informed audio data.

[0942] Specific operation: Create emotionally-charged speech data using a text generation model.

[0943] Step 6:

[0944] Server: The generated voice data is sent to other devices in real time using a real-time protocol.

[0945] Input: Emotionally-charged speech data.

[0946] Output: Audio data sent to other devices.

[0947] Specific operation: Sends voice data to a remote terminal using a real-time protocol.

[0948] Step 7:

[0949] The other party's terminal: Plays back the received voice data and conveys the user's silent speech to the other party.

[0950] Input: Audio data sent from the server.

[0951] Output: The audio played.

[0952] Specific operation: Audio is output from the speaker and the receiver listens to the user's speech.

[0953] Step 8:

[0954] Recipient: "Yes, I'm fine" replies in a normal voice.

[0955] Input: The other person's speech.

[0956] Output: Audio data.

[0957] Specific operation: Convert analog voice into digital data and input it into the terminal.

[0958] Step 9:

[0959] Remote device: Captures the remote device's voice data and sends it to the server.

[0960] Input: The other party's voice data.

[0961] Output: Sends audio data to the server.

[0962] Specific operation: After capturing the audio, send the data to the server.

[0963] Step 10:

[0964] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[0965] Input: The other party's voice data.

[0966] Output: Text data, emotion data.

[0967] Specific operation: Converts speech into text using speech recognition technology and analyzes emotional data.

[0968] Step 11:

[0969] Device: Receives text and emotion data and displays it to the user, or plays back a synthesized voice that reflects the emotion.

[0970] Input: Text data, emotion data.

[0971] Output: The displayed text or played audio.

[0972] Specific Action: Display the text to the user or play it as emotional speech.

[0973] (Application example 2)

[0974] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0975] In modern society, there is a growing demand for calls and communication in situations where confidentiality and privacy are essential. However, conventional voice calls pose the risk of sound leakage and eavesdropping, making them unsuitable for use in quiet environments or sensitive situations. There is also a need for a method to convey text information to the other party more accurately and emotionally. Technology is needed to solve this problem and enable emotionally charged, natural communication even in silent environments.

[0976] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for capturing the movements of the user's lips and tongue, means for generating text based on the captured movements, means for recognizing the user's emotions and reflecting them in the generated text, and means for converting the generated text into speech. This makes it possible to accurately generate text from the movements of the lips and tongue even in a silent environment, enabling voice calls that reflect the user's emotions.

[0977] "Means for capturing the movement of a user's lips and tongue" refers to devices or technologies that use an in-camera or specific sensors to capture the movement and shape of the area around a user's mouth in real time and record it as digital data.

[0978] "Means for generating text from captured movements" means algorithms and software for analyzing captured lip and tongue movement data and converting it into corresponding characters, words, or sentences.

[0979] "Means for recognizing user emotions and reflecting them in text generation" refers to technology that uses image analysis technology and facial expression recognition engines to detect emotions from the user's facial expressions and incorporate that emotional information into the text generation process.

[0980] "Means for converting generated text into speech" refers to software and hardware that utilizes speech synthesis technology using artificial intelligence to convert text data into speech data and generate natural-sounding speech that reflects emotion.

[0981] "Means for improving the accuracy of text generation by utilizing conversational context and the user's past learning data" refers to algorithms and databases that utilize the user's past conversation history and contextual information from dialogue to improve the content and accuracy of the generated text.

[0982] "Means for transmitting the generated voice data and emotional data to another terminal in real time and conversing with the other party" refers to communication technologies and protocols that enable real-time communication by quickly transmitting the generated voice data and emotional information to another device via a network.

[0983] The "Secure Silent Call System" that realizes this application example captures the movement of the user's lips and tongue, generates text based on that movement, recognizes the user's emotions and reflects them in the generated text, converts the generated text into speech, and transmits it to other devices in real time, enabling communication.

[0984] System configuration

[0985] 1. Capture device and emotion recognition engine

[0986] user:

[0987] This feature is intended for use in quiet environments such as conference rooms and libraries.

[0988] Device:

[0989] The system uses a built-in camera and specific sensors to capture the user's lip and tongue movements in real time. It also detects facial expressions and analyzes the user's emotions using an emotion recognition engine. The hardware used includes a standard webcam and a highly accurate facial recognition sensor.

[0990] Device:

[0991] The captured movement and facial expression data are stored as visual data and sent to a server.

[0992] 2. Data submission and analysis

[0993] server:

[0994] The software analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, leveraging the user's past learning data and conversational context. The software includes a lip movement recognition engine (LipMovementRecognizer) and an emotion recognition engine (EmotionRecognizer).

[0995] 3. Text-to-speech

[0996] server:

[0997] The generated text and emotion data are then converted into emotion-reflecting voice data using speech synthesis technology. This process is carried out using TextToSpeech software.

[0998] server:

[0999] The generated voice data is transmitted to another terminal in real time.

[1000] 4. Processing replies from the other party

[1001] Remote device:

[1002] As with a normal voice call, the voice data from the other party is captured and sent to the server.

[1003] server:

[1004] The received voice data is converted into text using voice recognition technology, and the text and emotional data are sent to the user's device.

[1005] Device:

[1006] It receives text and emotion data and displays it to the user. It also synthesizes the text into speech and plays it back with the emotion reflected.

[1007] Specific examples

[1008] 1. User training:

[1009] The user silently reads the training sample sentence, "I am in Tokyo now." The device captures lip and tongue movements with the front camera and sends the visual data to the server. It also captures facial expressions and sends them as emotional data. The server analyzes the received data and saves the visual and emotional data corresponding to the text "I am in Tokyo now" as training data.

[1010] 2. Actual call:

[1011] The user silently utters, "The meeting will start at 3 PM." The device captures lip and tongue movements with the in-camera and sends the visual data to the server. It also sends facial expression data for emotion recognition. The server analyzes the visual and emotional data and generates the text "The meeting will start at 3 PM" based on the learning data and context. It then converts this into speech using emotion-reflecting speech synthesis technology and sends it to the other device. The other device receives the voice data and conveys the user's remarks to the other party.

[1012] 3. Response from the other person:

[1013] The other party responds by saying "Got it. I'll join." The other party's device captures the voice data and sends it to the server. The server converts the voice data into text and sends the text and emotion data to the user's device. The device receives the text and emotion data and displays it to the user or plays it back as voice.

[1014] Example prompt for a generative AI model:

[1015] "Analyze the user's lip and tongue movements and generate text. Use the following video stream data. Also, recognize emotions from facial expressions and generate audio data that reflects them. Video stream data example: [Image data] Emotion data example: [Emotion label based on image data]"

[1016] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1017] Step 1:

[1018] The user silently speaks. For example, in a quiet environment, the user silently utters, "The meeting will start at 3:00 PM." At this time, the user moves their mouth in the same way as they would during a normal phone call.

[1019] Step 2:

[1020] The device uses an in-camera and specific sensors to capture the user's lip and tongue movements in real time. The video data captured by the in-camera is input, and visual data for analyzing lip and tongue movements is output. The device also captures the user's facial expressions. This facial expression data is also output as visual data.

[1021] Step 3:

[1022] The terminal transmits the captured visual data to the server. The input is the captured visual data, and the output is the data transmission to the server. Network communication technology is used for data transmission.

[1023] Step 4:

[1024] The server analyzes the received visual data. Specifically, it analyzes lip and tongue movement data and facial expression data to generate text data. The input is the received visual data, and the output is the text data generated by the data analysis process. LipMovementRecognizer and EmotionRecognizer are used for the analysis.

[1025] Step 5:

[1026] The server uses the conversational context and the user's past learning data to improve the accuracy of the generated text. The input is the initial text data, the user's past learning data, and the conversational context, and the output is the final text data with high accuracy. Natural language processing technology is used for data processing.

[1027] Step 6:

[1028] The server converts the final text data and emotion data into speech. The input is the final text data and emotion data, and the output is speech data that reflects the emotion expression. TextToSpeech software is used for speech synthesis.

[1029] Step 7:

[1030] The server transmits the generated voice data to the remote terminal in real time. The input is voice data, and the output is voice data transmitted over the network. A communication protocol is used for transmission.

[1031] Step 8:

[1032] The other device plays back the received voice data and transmits the user's speech to the other device. The input is the received voice data, and the output is the played voice.

[1033] Step 9:

[1034] The other party responds with voice, just like in a normal voice call. For example, they might respond with, "Got it. I'll join." This voice is captured by the other party's device.

[1035] Step 10:

[1036] The remote terminal transmits the captured voice data to the server. The input is the voice data, and the output is the data transmission to the server.

[1037] Step 11:

[1038] The server analyzes the received voice data and converts it into text data. The input is voice data and the output is text data. The analysis is performed using voice recognition technology.

[1039] Step 12:

[1040] The server sends the generated text data and emotion data to the user terminal. The input is the text data and emotion data, and the output is the data transmission to the user terminal.

[1041] Step 13:

[1042] The terminal displays the received text data and emotion data to the user, and also plays the text as speech that reflects the emotion. The input is the text data and emotion data, and the output is the displayed text and the played speech.

[1043] 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 a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1044] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1045] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1046] [Third embodiment]

[1047] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1048] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1049] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1050] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1051] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1052] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1053] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1054] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1055] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1056] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1057] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1058] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. 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."

[1059] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. The system captures the movements of the user's lips and tongue, generates text based on the captured movements, and converts the text into speech. Below, we will provide an overview of the system's program and its specific processing in natural language.

[1060] System configuration

[1061] Capture Device

[1062] User: Speak silently.

[1063] On the device: The front camera is used to capture the user's lip and tongue movements in real time, which are then stored as visual data.

[1064] Data transmission and analysis

[1065] Terminal: Sends captured visual data to the server.

[1066] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[1067] Text-to-speech

[1068] Server: The generated text is converted into voice data using speech synthesis technology.

[1069] Server: Sends the generated audio data to other devices in real time.

[1070] Response processing from the other party

[1071] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1072] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[1073] Specific examples

[1074] 1. User training

[1075] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1076] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[1077] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[1078] 2. Actual call

[1079] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1080] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[1081] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[1082] Server: Converts text into speech and sends the speech data to the other device.

[1083] The other device: Plays the received voice data and transmits the user's words to the other device.

[1084] 3. Response from the other person

[1085] Other party: Reply verbally with "Got it, I'll join."

[1086] Remote device: Captures audio data and sends it to the server.

[1087] Server: Converts voice data into text and sends it to the user's device.

[1088] On the device: The received text is displayed to the user or played as audio.

[1089] This allows the system to provide a smooth and secure means of making calls even in quiet environments or places where privacy is required. By utilizing the user's past learning data and conversation context, the system can generate text with higher accuracy and realize natural-sounding voice calls.

[1090] The processing flow will be explained below.

[1091] Training Data Collection Processing Steps

[1092] Step 1: Training Initial Setup

[1093] User: Select "Silent Calling."

[1094] Terminal: Presents training sample sentences to the user and prepares them to begin training.

[1095] Step 2: Silent reading of sample sentences

[1096] User: Silently read the provided sample sentences aloud.

[1097] On your device: Use the front camera to capture the movements of your lips and tongue.

[1098] Step 3: Sending visual data

[1099] Terminal: Sends captured visual data to the server.

[1100] Step 4: Data analysis and training

[1101] Server: Analyzes the received visual data and stores the lip and tongue movements and corresponding text as training data.

[1102] Actual call processing steps

[1103] Step 1: Start the call

[1104] User: Select the person to call and start the call.

[1105] Device: Prepare for the call and activate the front camera.

[1106] Step 2: Silent speech

[1107] User: Speak silently.

[1108] On the device: Uses the front camera to capture the user's lip and tongue movements in real time.

[1109] Step 3: Sending visual data

[1110] Terminal: Sends captured visual data to the server.

[1111] Step 4: Text Generation

[1112] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[1113] Step 5: Audio conversion

[1114] Server: The generated text is converted into voice data using speech synthesis technology.

[1115] Step 6: Sending audio data

[1116] Server: Sends the generated voice data to the other device.

[1117] Step 7: Playing Audio

[1118] The other device: Plays the received voice data and transmits the user's words to the other device.

[1119] Steps for processing a reply from the other party

[1120] Step 1: Capture audio data

[1121] Recipient: Respond by voice as usual.

[1122] Remote device: Captures the response audio data and sends it to the server.

[1123] Step 2: Analyzing the audio data

[1124] Server: Converts the received voice data into text using voice recognition technology.

[1125] Step 3: Send a text

[1126] Server: Sends the textual response to the user's device.

[1127] Step 4: Display text or play audio

[1128] Device: Displays the received text to the user, or synthesizes the text to speech and plays it through headphones.

[1129] This is a step-by-step explanation of the program's processing. Through this series of processes, the system enables smooth and safe calls even in quiet environments or places where privacy is required.

[1130] Example 1

[1131] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1132] Currently, voice communication relies on direct voice communication, which makes it difficult to use in quiet environments or places where privacy is required. Voice communication is also difficult for people with hearing impairments. Furthermore, conventional voice input systems have problems with misrecognition due to background noise and differences in speaking speed.

[1133] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1134] In this invention, the server includes means for capturing the user's lip and tongue movements, means for generating text based on the captured movements, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for converting the speech data received from the other terminal into text. This enables silent calls, allowing for smooth and safe calls even in quiet environments or places where privacy is required. It also provides an alternative means of voice calls for people with hearing impairments, enabling highly accurate calls unaffected by background noise or speech rate.

[1135] "Capturing means" refers to devices and software that detect the user's lip and tongue movements in real time and acquire and store them as visual data.

[1136] The "means for generating text" refers to a device and software for analyzing the captured visual data and generating text data based on the content of the user's silent utterances.

[1137] The "means for converting to speech" refers to devices and software for converting the generated text data into synthetic speech. Specifically, speech synthesis technology is used.

[1138] The "transmitting means" refers to a device and software for transmitting the generated voice data to another terminal via a network.

[1139] The "means for converting into text" refers to the device and software that analyzes the voice data received from the other party's terminal and converts it into text data. Specifically, it uses voice recognition technology.

[1140] "Conversational context" refers to information and processing methods for generating appropriate text based on the context of the user's statements and the flow of the dialogue.

[1141] "Past learning data" refers to data accumulated based on the user's past training and conversation content, and is the data and processing method used to improve the accuracy of text generation.

[1142] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. This system provides the function of capturing the movements of the user's lips and tongue, generating text based on the captured movements, and converting the text into speech. It is also possible to convert the speech of the other party into text and provide it to the user.

[1143] System configuration

[1144] Capture Device

[1145] The user speaks silently, and the device uses the front camera to capture the user's lip and tongue movements in real time. The captured movements are saved as visual data. The front camera uses a high-resolution camera and a dedicated chip for real-time image processing.

[1146] Data transmission and analysis

[1147] The device sends the captured visual data to a server, where it is encrypted and transmitted securely using the HTTPS protocol.

[1148] The server analyzes the received visual data and generates corresponding text using CNN (Convolutional Neural Network) and RNN (Recurrent Neural Network) technologies, generating highly accurate text based on the user's past learning data and conversational context.

[1149] Text-to-speech

[1150] The server converts the generated text into audio data using TTS (Text to Speech) technology, such as Google Cloud Text-to-Speech. This audio data is generated as a high-quality audio file (e.g., WAV format).

[1151] The server transmits the generated voice data to the other terminal in real time.

[1152] Response processing from the other party

[1153] The other device captures the voice data spoken by the other user and sends it to the server. The voice data is encrypted and sent using the HTTPS protocol.

[1154] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology. In this case, we use "Google Cloud Speech-to-Text."

[1155] The server sends the converted text to the user terminal.

[1156] The device will then display the received text to the user or play it as audio.

[1157] Specific examples

[1158] 1. User training

[1159] The user silently reads the training sample sentence, "I am currently in Tokyo."

[1160] The device's in-camera captures lip and tongue movements and sends the visual data to a server.

[1161] The server analyzes the received data and saves the text "I am in Tokyo now" along with the corresponding visual data as training data.

[1162] 2. Actual call

[1163] The user starts a call and silently says, "The meeting starts at 3:00 PM."

[1164] The device's in-camera captures the user's lip and tongue movements and sends the visual data to a server.

[1165] The server analyzes the visual data and generates the text "The meeting starts at 3pm" based on the training data and context.

[1166] The server converts the text into voice and transmits the voice data to the other terminal.

[1167] The other device plays the received voice data and transmits the user's words to the other device.

[1168] 3. Response from the other person

[1169] The other person will respond by voice, "Understood, I'll join."

[1170] The remote terminal captures the voice data and sends it to the server.

[1171] The server converts the voice data into text and sends it to the user's device.

[1172] The device will then display the received text to the user or play it as audio.

[1173] This allows the system to provide a smooth and secure means of making calls even in quiet environments or where privacy is required. By utilizing past training data and conversation context, the system can generate text with higher accuracy and achieve natural-sounding voice calls.

[1174] Prompt Sentence Examples

[1175] Here are some example prompts to present to a generative AI model:

[1176] User: Use the "Silent Call" system to say silently, "Thank you for your hard work, regarding tomorrow's meeting..."

[1177] Device: Capture lip and tongue movements with the front camera and send the data to the server.

[1178] Server: Analyze the received data, perform syntax analysis and speech synthesis, and send it to the other party as audio.

[1179] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1180] Program processing flow

[1181] Step 1: Capture process

[1182] The user speaks silently.

[1183] The device uses a built-in camera to capture the user's lip and tongue movements in real time, and specifically, it uses a high-resolution camera and a dedicated real-time image processing chip to acquire visual data.

[1184] Input: User's lip and tongue movements

[1185] Output: Visual data (images or video frames)

[1186] Step 2: Send data

[1187] The device encrypts the captured visual data and sends it to a server using the HTTPS protocol.

[1188] The server receives the received visual data.

[1189] Input: Visual data (terminal input data)

[1190] Output: Visual data (received by the server)

[1191] Step 3: Data analysis and text generation

[1192] The server analyzes the received visual data using a high-speed image processing algorithm, specifically, a convolutional neural network (CNN) to extract lip and tongue movements as features.

[1193] The server inputs the features into a recurrent neural network (RNN) model and generates the corresponding text, taking into account the user's past learning data and conversational context.

[1194] Input: Visual data (received data from the server)

[1195] Output: The generated text

[1196] Step 4: Text-to-speech

[1197] The server converts the generated text into voice data using TTS (Text to Speech) technology, for example, by using a "voice synthesis engine."

[1198] The server transmits the generated voice data to the other terminal in real time.

[1199] Input: Generated text

[1200] Output: Audio data

[1201] Step 5: Processing the other party's response

[1202] The other terminal captures the voice data uttered by the other user and transmits it to the server.

[1203] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology, for example, by using a "voice recognition engine."

[1204] The server sends the converted text to the user terminal.

[1205] Input: Voice data (the other person's input data)

[1206] Output: The converted text

[1207] Step 6: Display to the user

[1208] The device displays the received text to the user or plays it as audio using TTS technology.

[1209] Input: Translated text

[1210] Output: Text or audio displayed to the user

[1211] This allows the overall processing flow to be specified and the operations and data flow at each step to be clearly explained.

[1212] (Application example 1)

[1213] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1214] In security work, it is necessary to be able to make calls and share information smoothly even in quiet environments and when privacy is required. Current voice communication systems have the risk of information leaking to those around them due to the sound they emit, and are difficult to use in quiet environments. A system that can solve these problems is needed.

[1215] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1216] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for capturing speech data from the other party and converting it into text. This enables smooth and safe silent communication and information sharing even in security work that requires quiet environments and privacy.

[1217] "User" means an individual who uses the System to make silent calls.

[1218] "Lip and tongue movements" refers to the movements inside and outside the mouth that a user makes when silently speaking.

[1219] "Capture" refers to the act of recording visual data using an in-camera or other device.

[1220] "Text generation" refers to the process of analyzing captured visual data and creating corresponding string data.

[1221] "Text-to-speech" refers to the technique of converting generated text into audio data.

[1222] "Other device" refers to the device used by the person with whom you are making a call.

[1223] "Server" refers to a central device that performs processing such as data analysis, text generation, and speech conversion.

[1224] "Silent security call application" refers to software that enables silent calls in security work where quiet environments and privacy are required.

[1225] This invention relates to a "Silent Security Call Application" that enables users to make silent calls in situations where quiet environments and privacy protection are required in security work. The following is a detailed explanation of how to specifically implement this system.

[1226] System program generation

[1227] The system achieves the following main functions:

[1228] 1. Hardware Configuration

[1229] User device: Use a device equipped with an in-camera, such as a smartphone or smart glasses.

[1230] Server: Use a high-performance computer server to perform data analysis, text generation, and speech conversion.

[1231] 2. Software Configuration

[1232] OpenCV: A library used to capture the user's lip and tongue movements.

[1233] gTTS (Google Text-to-Speech): A library for converting generated text to speech.

[1234] speech_recognition: A library for converting speech into text.

[1235] Program processing

[1236] 1. Data Capture

[1237] On the device: The front camera captures the user's lip and tongue movements in real time.

[1238] Data transmission: The captured visual data is transmitted from the device to the server.

[1239] 2. Text Generation

[1240] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[1241] 3. Audio conversion

[1242] Server: Converts the generated text into speech data using the gTTS library.

[1243] Data transmission: The generated voice data is transmitted to other devices in real time.

[1244] 4. Processing replies from the other party

[1245] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1246] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[1247] Specific examples

[1248] 1. User training

[1249] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1250] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[1251] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[1252] 2. Actual call

[1253] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1254] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[1255] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[1256] Server: Converts text into speech and sends the speech data to the other device.

[1257] The other device: Plays the received voice data and transmits the user's words to the other device.

[1258] 3. Response from the other person

[1259] Other party: Reply verbally with "Got it, I'll join."

[1260] Remote device: Captures audio data and sends it to the server.

[1261] Server: Converts voice data into text and sends it to the user's device.

[1262] On the device: The received text is displayed to the user or played as audio.

[1263] Prompt Sentence Examples

[1264] Your application will analyze the user's lip movements to generate text and convert it to speech. The following requirements must be met:

[1265] The user makes a call without speaking.

[1266] The capture device uses the in-camera.

[1267] The data is sent to a server and converted into text and audio data.

[1268] The software used is Python's OpenCV, gTTS, and speech_recognition libraries.

[1269] This makes it possible for the system to provide a means for smooth, safe, and silent communication and information sharing, even in quiet environments or places where privacy is required.

[1270] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1271] Step 1:

[1272] The user speaks silently. The user's lip and tongue movements are captured in real time by the in-camera. The input is the user's lip and tongue movements, and the output is visual data. The device records this visual data.

[1273] Step 2:

[1274] The terminal transmits the captured visual data to the server, where it is sent in the form of a data package that reaches the server over the network. The input is the visual data, and the output is the data package received at the server side.

[1275] Step 3:

[1276] The server analyzes the received visual data and generates corresponding text. To analyze the visual data, it uses machine learning models, which utilize the user's past learning data and conversational context. The input is the visual data, and the output is the generated text.

[1277] Step 4:

[1278] The server converts the generated text into speech data using speech synthesis technology and the gTTS (Google Text-to-Speech) library. The input is text data and the output is speech data.

[1279] Step 5:

[1280] The server then transmits the generated audio data to other devices in real time. This must be done quickly over the network, in the form of a data package. The input is the audio data, and the output is an audio file that can be played on other devices.

[1281] Step 6:

[1282] The other device plays back the received voice data and conveys what the other party said to the user. The input is voice data, and the output is the played-back voice.

[1283] Step 7:

[1284] The other party responds by voice. The other party's device captures the voice data and sends it to the server. The input is the other party's voice, and the output is the voice data sent to the server.

[1285] Step 8:

[1286] The server converts the received voice data into text using voice recognition technology. The input is voice data and the output is text data.

[1287] Step 9:

[1288] The terminal displays the text data sent from the server to the user or plays it aloud. The input is the text data, and the output is the displayed text or played audio.

[1289] This allows the entire system to smoothly communicate and share information in security operations without making any sound.

[1290] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1291] This invention relates to a "Silent Call" system that allows users to make calls without speaking. This system achieves more natural conversations by capturing the user's lip and tongue movements and combining them with an emotion engine. Below, we will provide an overview of the system's program and its specific processing in natural language.

[1292] System configuration

[1293] Capture device and emotion recognition engine

[1294] User: Speak silently.

[1295] On-device: Uses the front camera to capture the user's lip and tongue movements in real time, and uses an emotion recognition engine to analyze the user's emotions from their facial expressions.

[1296] Device: The captured movement and facial expression data is stored as visual data.

[1297] Data transmission and analysis

[1298] Terminal: Sends captured visual data to the server.

[1299] Server: Analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, utilizing the user's past learning data and conversational context.

[1300] Text-to-speech

[1301] Server: Based on the generated text and the recognized emotion data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[1302] Server: Sends the generated audio data to other devices in real time.

[1303] Response processing from the other party

[1304] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1305] Server: Converts the received voice data into text using speech recognition technology and displays it to the user or plays it back as audio.

[1306] Specific examples

[1307] 1. User training

[1308] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1309] Device: The front camera captures lip and tongue movements and sends visual data to the server. It also captures facial expressions and sends them as emotional data.

[1310] Server: Analyzes the received data and saves the text "I am currently in Tokyo" and the corresponding visual and emotional data as training data.

[1311] 2. Actual call

[1312] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1313] Device: The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[1314] Server: Analyzes visual and emotional data and generates text such as "The meeting starts at 3 p.m." based on the training data and context.

[1315] Server: Converts text into speech using emotion-based speech synthesis technology and sends the speech data to the other device.

[1316] The other device: Plays the received voice data and transmits the user's words to the other device.

[1317] 3. Response from the other person

[1318] Other party: Reply verbally with "Got it, I'll join."

[1319] Remote device: Captures audio data and sends it to the server.

[1320] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1321] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[1322] This system provides a means for smooth and secure communication even in quiet environments or places where privacy is required. By analyzing emotional data from the user's facial expressions and reflecting emotions in text generation and speech conversion, it is possible to achieve more natural and human-like communication.

[1323] The processing flow will be explained below.

[1324] Training Data Collection Processing Steps

[1325] Step 1: Training Initial Setup

[1326] User: Select "Silent Call" and start training.

[1327] Terminal: Presents training example sentences to the user.

[1328] Step 2: Silent reading of sample sentences

[1329] User: Silently read the provided sample sentences aloud.

[1330] On the device: The front camera captures the user's lip and tongue movements, as well as facial expressions to obtain emotional data.

[1331] Step 3: Sending visual data

[1332] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[1333] Step 4: Data analysis and training

[1334] Server: Analyzes the received visual and emotional data and saves the training data as data corresponding to the text.

[1335] Actual call processing steps

[1336] Step 1: Start the call

[1337] User: Select the person to call and start the call.

[1338] Device: Activate the front camera and emotion recognition engine and prepare.

[1339] Step 2: Silent speech

[1340] User: Speak silently.

[1341] On the device: The front camera captures the user's lip and tongue movements in real time, while the emotion engine simultaneously captures facial expression data.

[1342] Step 3: Sending visual data

[1343] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[1344] Step 4: Text Generation

[1345] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[1346] Step 5: Analyze the sentiment data

[1347] Server: Analyzes the user's emotions from facial expression data and assigns the recognized emotions to text data.

[1348] Step 6: Audio conversion

[1349] Server: The text generated reflecting emotions is converted into voice data using speech synthesis technology.

[1350] Step 7: Sending audio data

[1351] Server: Sends the generated voice data to the other device in real time.

[1352] Step 8: Playing Audio

[1353] The other device: Plays the received voice data and transmits the user's words to the other device.

[1354] Steps for processing a reply from the other party

[1355] Step 1: Capture audio data

[1356] Other party: Speak as if you were making a normal voice call.

[1357] Remote device: Captures the remote device's voice data and sends it to the server.

[1358] Step 2: Analyzing the audio data

[1359] Server: Converts the received voice data into text using voice recognition technology.

[1360] Step 3: Send a text

[1361] Server: Sends the generated text to the user's device and also analyzes emotion data if necessary.

[1362] Step 4: Display text or play audio

[1363] Terminal: Receives text and emotion data and displays it to the user or plays it back as audio.

[1364] The above is the flow of specific processing steps of the invention that combines an emotion engine. This system enables smooth and safe calls even in quiet environments or places where privacy is required, and also enables natural communication that reflects emotions.

[1365] Example 2

[1366] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1367] Conventional voice communication systems have the problem that it is difficult for users to communicate smoothly in quiet environments or places where privacy is required, and they also have the problem that it is difficult to express emotions, making it difficult to realize natural conversations.

[1368] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1369] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for analyzing the user's facial expression to obtain emotional data, means for generating speech data reflecting the emotion based on the generated text and the obtained emotional data, and means for transmitting the generated speech data to another terminal in real time, thereby enabling smooth and natural conversation even in quiet environments or places where privacy is required.

[1370] "User" refers to an individual who uses the system to make silent speech.

[1371] "Means for capturing lip and tongue movements" refers to a device equipped with a camera and / or sensors for capturing the user's lip and tongue movements in real time.

[1372] "Means for generating text based on captured movements" refers to algorithms or software that analyze captured lip and tongue movements and generate corresponding text data.

[1373] "Generated text-to-speech means" refers to an algorithm or program that uses speech synthesis technology to convert text data into speech data.

[1374] "Means of analyzing facial expressions to obtain emotional data" refers to technology and software that captures a user's facial expressions and recognizes and analyzes emotions from those expressions.

[1375] "Means for generating voice data that reflects emotions based on acquired emotional data" refers to algorithms and programs for synthesizing voice that reflects emotional data.

[1376] "Means for transmitting generated voice data to other terminals in real time" refers to communication technologies and protocols for delivering voice data to other terminals without delay.

[1377] The present invention relates to a "Silent Call" system that enables users to have natural conversations without vocalizations in quiet environments or situations where privacy is required. This system acquires emotional data by capturing the movements of the user's lips and tongue and analyzing their facial expressions, and combines them to generate text and voice data for voice calls.

[1378] System configuration

[1379] Capture device and emotion recognition engine

[1380] User: Speaks silently. Imagine a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[1381] Device:

[1382] An in-camera (e.g., a commercially available high-resolution camera) is used to capture the user's lip and tongue movements in real time, and an emotion recognition engine (e.g., emotion analysis software) is used to analyze the user's emotions from their facial expressions and obtain emotion data.

[1383] The captured movement and emotion data is stored as visual data.

[1384] Hardware used:

[1385] High-resolution in-camera, microprocessor

[1386] Software used:

[1387] Sentiment analysis software, data analysis algorithms

[1388] Text generation and speech synthesis

[1389] Device:

[1390] The captured visual and emotional data is sent to a server, using an internet connection for data communication.

[1391] server:

[1392] The received visual data is analyzed and the text "Hello, how are you?" is generated from lip and tongue movements and facial expressions using a generative AI model (e.g., a machine learning model, as the term is generally used).

[1393] Based on the generated text and the acquired emotional data, it is converted into voice data using emotion-reflecting voice synthesis technology (commonly known as voice synthesis software).

[1394] Hardware used:

[1395] Server computers, high-speed data processing units

[1396] Software used:

[1397] Machine learning models, speech synthesis software

[1398] Communication and playback

[1399] server:

[1400] The generated voice data is sent to other devices in real time using a real-time protocol.

[1401] Remote device:

[1402] The received voice data is played back and the user's silent speech is conveyed to the other party.

[1403] Hardware used:

[1404] Communication modem, speaker

[1405] Specific examples

[1406] 1. User training

[1407] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1408] Device:

[1409] The front camera captures lip and tongue movements and sends visual data to the server, as well as facial expressions, which are then sent as emotional data.

[1410] server:

[1411] The received data is analyzed, and the visual and emotional data corresponding to the text "I am in Tokyo now" are saved as training data.

[1412] 2. Actual call

[1413] User: Starts a call and silently says "My meeting starts at 3 PM."

[1414] Device:

[1415] The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[1416] server:

[1417] It analyzes visual and emotional data and generates the text "The meeting starts at 3 p.m." based on the training data and context.

[1418] server:

[1419] Text is converted into speech using emotion-reflecting speech synthesis technology, and the speech data is sent to the other party's terminal.

[1420] Remote device:

[1421] The received voice data is played back and the user's words are conveyed to the other party.

[1422] 3. Response from the other person

[1423] Other party: Reply verbally with "Got it, I'll join."

[1424] Remote device: Captures audio data and sends it to the server.

[1425] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1426] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[1427] This system enables smooth and natural conversations even in quiet environments or places where privacy is required. It also analyzes emotional data from the user's facial expressions and reflects those emotions in text generation and speech conversion, enabling more human-like communication.

[1428] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1429] Step 1:

[1430] User: Speaks silently. For example, consider a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[1431] Input: The user's lip and tongue movements.

[1432] Output: Real-time lip and tongue movement data.

[1433] Specific actions: The user speaks silently in front of the camera.

[1434] Step 2:

[1435] On the device: The front camera captures the user's lip and tongue movements in real time, and an emotion recognition engine is used to analyze facial expressions and obtain emotional data.

[1436] Input: Real-time lip and tongue movement data, facial expression data.

[1437] Output: Captured movement and emotion data are saved as visual data.

[1438] What it does: The camera and emotion recognition engine work together to record the user's lips, tongue, and facial expressions.

[1439] Step 3:

[1440] Device: The saved visual and emotional data is sent to the server. Data communication is via an internet connection.

[1441] Input: visual and emotional data.

[1442] Output: Sending data to the server.

[1443] What happens: Your device uploads data to a server via your internet connection.

[1444] Step 4:

[1445] Server: Analyzes the received visual data and generates text from lip and tongue movements.

[1446] Input: Captured visual data, emotion data.

[1447] Output: The generated text.

[1448] What it does: Uses a generative AI model to generate the text "Hello, how are you?" from lip and tongue movements.

[1449] Step 5:

[1450] Server: Based on the generated text and the acquired emotional data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[1451] Input: Text data, emotion data.

[1452] Output: Emotionally-informed audio data.

[1453] Specific operation: Create emotionally-charged speech data using a text generation model.

[1454] Step 6:

[1455] Server: The generated voice data is sent to other devices in real time using a real-time protocol.

[1456] Input: Emotionally-charged speech data.

[1457] Output: Audio data sent to other devices.

[1458] Specific operation: Sends voice data to a remote terminal using a real-time protocol.

[1459] Step 7:

[1460] The other party's terminal: Plays back the received voice data and conveys the user's silent speech to the other party.

[1461] Input: Audio data sent from the server.

[1462] Output: The audio played.

[1463] Specific operation: Audio is output from the speaker and the receiver listens to the user's speech.

[1464] Step 8:

[1465] Recipient: "Yes, I'm fine" replies in a normal voice.

[1466] Input: The other person's speech.

[1467] Output: Audio data.

[1468] Specific operation: Convert analog voice into digital data and input it into the terminal.

[1469] Step 9:

[1470] Remote device: Captures the remote device's voice data and sends it to the server.

[1471] Input: The other party's voice data.

[1472] Output: Sends audio data to the server.

[1473] Specific operation: After capturing the audio, send the data to the server.

[1474] Step 10:

[1475] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1476] Input: The other party's voice data.

[1477] Output: Text data, emotion data.

[1478] Specific operation: Converts speech into text using speech recognition technology and analyzes emotional data.

[1479] Step 11:

[1480] Device: Receives text and emotion data and displays it to the user, or plays back a synthesized voice that reflects the emotion.

[1481] Input: Text data, emotion data.

[1482] Output: The displayed text or played audio.

[1483] Specific Action: Display the text to the user or play it as emotional speech.

[1484] (Application example 2)

[1485] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1486] In modern society, there is a growing demand for calls and communication in situations where confidentiality and privacy are essential. However, conventional voice calls pose the risk of sound leakage and eavesdropping, making them unsuitable for use in quiet environments or sensitive situations. There is also a need for a method to convey text information to the other party more accurately and emotionally. Technology is needed to solve this problem and enable emotionally charged, natural communication even in silent environments.

[1487] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for capturing the movements of the user's lips and tongue, means for generating text based on the captured movements, means for recognizing the user's emotions and reflecting them in the generated text, and means for converting the generated text into speech. This makes it possible to accurately generate text from the movements of the lips and tongue even in a silent environment, enabling voice calls that reflect the user's emotions.

[1488] "Means for capturing the movement of a user's lips and tongue" refers to devices or technologies that use an in-camera or specific sensors to capture the movement and shape of the area around a user's mouth in real time and record it as digital data.

[1489] "Means for generating text from captured movements" means algorithms and software for analyzing captured lip and tongue movement data and converting it into corresponding characters, words, or sentences.

[1490] "Means for recognizing user emotions and reflecting them in text generation" refers to technology that uses image analysis technology and facial expression recognition engines to detect emotions from the user's facial expressions and incorporate that emotional information into the text generation process.

[1491] "Means for converting generated text into speech" refers to software and hardware that utilizes speech synthesis technology using artificial intelligence to convert text data into speech data and generate natural-sounding speech that reflects emotion.

[1492] "Means for improving the accuracy of text generation by utilizing conversational context and the user's past learning data" refers to algorithms and databases that utilize the user's past conversation history and contextual information from dialogue to improve the content and accuracy of the generated text.

[1493] "Means for transmitting the generated voice data and emotional data to another terminal in real time and conversing with the other party" refers to communication technologies and protocols that enable real-time communication by quickly transmitting the generated voice data and emotional information to another device via a network.

[1494] The "Secure Silent Call System" that realizes this application example captures the movement of the user's lips and tongue, generates text based on that movement, recognizes the user's emotions and reflects them in the generated text, converts the generated text into speech, and transmits it to other devices in real time, enabling communication.

[1495] System configuration

[1496] 1. Capture device and emotion recognition engine

[1497] user:

[1498] This feature is intended for use in quiet environments such as conference rooms and libraries.

[1499] Device:

[1500] The system uses a built-in camera and specific sensors to capture the user's lip and tongue movements in real time. It also detects facial expressions and analyzes the user's emotions using an emotion recognition engine. The hardware used includes a standard webcam and a highly accurate facial recognition sensor.

[1501] Device:

[1502] The captured movement and facial expression data are stored as visual data and sent to a server.

[1503] 2. Data submission and analysis

[1504] server:

[1505] The software analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, leveraging the user's past learning data and conversational context. The software includes a lip movement recognition engine (LipMovementRecognizer) and an emotion recognition engine (EmotionRecognizer).

[1506] 3. Text-to-speech

[1507] server:

[1508] The generated text and emotion data are then converted into emotion-reflecting voice data using speech synthesis technology. This process is carried out using TextToSpeech software.

[1509] server:

[1510] The generated voice data is transmitted to another terminal in real time.

[1511] 4. Processing replies from the other party

[1512] Remote device:

[1513] As with a normal voice call, the voice data from the other party is captured and sent to the server.

[1514] server:

[1515] The received voice data is converted into text using voice recognition technology, and the text and emotional data are sent to the user's device.

[1516] Device:

[1517] It receives text and emotion data and displays it to the user. It also synthesizes the text into speech and plays it back with the emotion reflected.

[1518] Specific examples

[1519] 1. User training:

[1520] The user silently reads the training sample sentence, "I am in Tokyo now." The device captures lip and tongue movements with the front camera and sends the visual data to the server. It also captures facial expressions and sends them as emotional data. The server analyzes the received data and saves the visual and emotional data corresponding to the text "I am in Tokyo now" as training data.

[1521] 2. Actual call:

[1522] The user silently utters, "The meeting will start at 3 PM." The device captures lip and tongue movements with the in-camera and sends the visual data to the server. It also sends facial expression data for emotion recognition. The server analyzes the visual and emotional data and generates the text "The meeting will start at 3 PM" based on the learning data and context. It then converts this into speech using emotion-reflecting speech synthesis technology and sends it to the other device. The other device receives the voice data and conveys the user's remarks to the other party.

[1523] 3. Response from the other person:

[1524] The other party responds by saying "Got it. I'll join." The other party's device captures the voice data and sends it to the server. The server converts the voice data into text and sends the text and emotion data to the user's device. The device receives the text and emotion data and displays it to the user or plays it back as voice.

[1525] Example prompt for a generative AI model:

[1526] "Analyze the user's lip and tongue movements and generate text. Use the following video stream data. Also, recognize emotions from facial expressions and generate audio data that reflects them. Video stream data example: [Image data] Emotion data example: [Emotion label based on image data]"

[1527] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1528] Step 1:

[1529] The user silently speaks. For example, in a quiet environment, the user silently utters, "The meeting will start at 3:00 PM." At this time, the user moves their mouth in the same way as they would during a normal phone call.

[1530] Step 2:

[1531] The device uses an in-camera and specific sensors to capture the user's lip and tongue movements in real time. The video data captured by the in-camera is input, and visual data for analyzing lip and tongue movements is output. The device also captures the user's facial expressions. This facial expression data is also output as visual data.

[1532] Step 3:

[1533] The terminal transmits the captured visual data to the server. The input is the captured visual data, and the output is the data transmission to the server. Network communication technology is used for data transmission.

[1534] Step 4:

[1535] The server analyzes the received visual data. Specifically, it analyzes lip and tongue movement data and facial expression data to generate text data. The input is the received visual data, and the output is the text data generated by the data analysis process. LipMovementRecognizer and EmotionRecognizer are used for the analysis.

[1536] Step 5:

[1537] The server uses the conversational context and the user's past learning data to improve the accuracy of the generated text. The input is the initial text data, the user's past learning data, and the conversational context, and the output is the final text data with high accuracy. Natural language processing technology is used for data processing.

[1538] Step 6:

[1539] The server converts the final text data and emotion data into speech. The input is the final text data and emotion data, and the output is speech data that reflects the emotion expression. TextToSpeech software is used for speech synthesis.

[1540] Step 7:

[1541] The server transmits the generated voice data to the remote terminal in real time. The input is voice data, and the output is voice data transmitted over the network. A communication protocol is used for transmission.

[1542] Step 8:

[1543] The other device plays back the received voice data and transmits the user's speech to the other device. The input is the received voice data, and the output is the played voice.

[1544] Step 9:

[1545] The other party responds with voice, just like in a normal voice call. For example, they might respond with, "Got it. I'll join." This voice is captured by the other party's device.

[1546] Step 10:

[1547] The remote terminal transmits the captured voice data to the server. The input is the voice data, and the output is the data transmission to the server.

[1548] Step 11:

[1549] The server analyzes the received voice data and converts it into text data. The input is voice data and the output is text data. The analysis is performed using voice recognition technology.

[1550] Step 12:

[1551] The server sends the generated text data and emotion data to the user terminal. The input is the text data and emotion data, and the output is the data transmission to the user terminal.

[1552] Step 13:

[1553] The terminal displays the received text data and emotion data to the user, and also plays the text as speech that reflects the emotion. The input is the text data and emotion data, and the output is the displayed text and the played speech.

[1554] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1555] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1556] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1557] [Fourth embodiment]

[1558] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1559] 7, a 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.

[1560] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1561] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1562] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1563] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1564] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 are responsible for the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1565] The control object 443 includes a display device, LEDs in the eyes, and motors for driving 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 emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1566] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1567] The specific processing program 56 is an example of a "program" according to the technology of the present 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.

[1568] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1569] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. 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 process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1570] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1571] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. The system captures the movements of the user's lips and tongue, generates text based on the captured movements, and converts the text into speech. Below, we will provide an overview of the system's program and its specific processing in natural language.

[1572] System configuration

[1573] Capture Device

[1574] User: Speak silently.

[1575] On the device: The front camera is used to capture the user's lip and tongue movements in real time, which are then stored as visual data.

[1576] Data transmission and analysis

[1577] Terminal: Sends captured visual data to the server.

[1578] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[1579] Text-to-speech

[1580] Server: The generated text is converted into voice data using speech synthesis technology.

[1581] Server: Sends the generated audio data to other devices in real time.

[1582] Response processing from the other party

[1583] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1584] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[1585] Specific examples

[1586] 1. User training

[1587] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1588] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[1589] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[1590] 2. Actual call

[1591] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1592] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[1593] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[1594] Server: Converts text into speech and sends the speech data to the other device.

[1595] The other device: Plays the received voice data and transmits the user's words to the other device.

[1596] 3. Response from the other person

[1597] Other party: Reply verbally with "Got it, I'll join."

[1598] Remote device: Captures audio data and sends it to the server.

[1599] Server: Converts voice data into text and sends it to the user's device.

[1600] On the device: The received text is displayed to the user or played as audio.

[1601] This allows the system to provide a smooth and secure means of making calls even in quiet environments or places where privacy is required. By utilizing the user's past learning data and conversation context, the system can generate text with higher accuracy and realize natural-sounding voice calls.

[1602] The processing flow will be explained below.

[1603] Training Data Collection Processing Steps

[1604] Step 1: Training Initial Setup

[1605] User: Select "Silent Calling."

[1606] Terminal: Presents training sample sentences to the user and prepares them to begin training.

[1607] Step 2: Silent reading of sample sentences

[1608] User: Silently read the provided sample sentences aloud.

[1609] On your device: Use the front camera to capture the movements of your lips and tongue.

[1610] Step 3: Sending visual data

[1611] Terminal: Sends captured visual data to the server.

[1612] Step 4: Data analysis and training

[1613] Server: Analyzes the received visual data and stores the lip and tongue movements and corresponding text as training data.

[1614] Actual call processing steps

[1615] Step 1: Start the call

[1616] User: Select the person to call and start the call.

[1617] Device: Prepare for the call and activate the front camera.

[1618] Step 2: Silent speech

[1619] User: Speak silently.

[1620] On the device: Uses the front camera to capture the user's lip and tongue movements in real time.

[1621] Step 3: Sending visual data

[1622] Terminal: Sends captured visual data to the server.

[1623] Step 4: Text Generation

[1624] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[1625] Step 5: Audio conversion

[1626] Server: The generated text is converted into voice data using speech synthesis technology.

[1627] Step 6: Sending audio data

[1628] Server: Sends the generated voice data to the other device.

[1629] Step 7: Playing Audio

[1630] The other device: Plays the received voice data and transmits the user's words to the other device.

[1631] Steps for processing a reply from the other party

[1632] Step 1: Capture audio data

[1633] Recipient: Respond by voice as usual.

[1634] Remote device: Captures the response audio data and sends it to the server.

[1635] Step 2: Analyzing the audio data

[1636] Server: Converts the received voice data into text using voice recognition technology.

[1637] Step 3: Send a text

[1638] Server: Sends the textual response to the user's device.

[1639] Step 4: Display text or play audio

[1640] Device: Displays the received text to the user, or synthesizes the text to speech and plays it through headphones.

[1641] This is a step-by-step explanation of the program's processing. Through this series of processes, the system enables smooth and safe calls even in quiet environments or places where privacy is required.

[1642] Example 1

[1643] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1644] Currently, voice communication relies on direct voice communication, which makes it difficult to use in quiet environments or places where privacy is required. Voice communication is also difficult for people with hearing impairments. Furthermore, conventional voice input systems have problems with misrecognition due to background noise and differences in speaking speed.

[1645] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1646] In this invention, the server includes means for capturing the user's lip and tongue movements, means for generating text based on the captured movements, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for converting the speech data received from the other terminal into text. This enables silent calls, allowing for smooth and safe calls even in quiet environments or places where privacy is required. It also provides an alternative means of voice calls for people with hearing impairments, enabling highly accurate calls unaffected by background noise or speech rate.

[1647] "Capturing means" refers to devices and software that detect the user's lip and tongue movements in real time and acquire and store them as visual data.

[1648] The "means for generating text" refers to a device and software for analyzing the captured visual data and generating text data based on the content of the user's silent utterances.

[1649] The "means for converting to speech" refers to devices and software for converting the generated text data into synthetic speech. Specifically, speech synthesis technology is used.

[1650] The "transmitting means" refers to a device and software for transmitting the generated voice data to another terminal via a network.

[1651] The "means for converting into text" refers to the device and software that analyzes the voice data received from the other party's terminal and converts it into text data. Specifically, it uses voice recognition technology.

[1652] "Conversational context" refers to information and processing methods for generating appropriate text based on the context of the user's statements and the flow of the dialogue.

[1653] "Past learning data" refers to data accumulated based on the user's past training and conversation content, and is the data and processing method used to improve the accuracy of text generation.

[1654] The present invention relates to a "Silent Call" system that allows users to make calls without making any sound. This system provides the function of capturing the movements of the user's lips and tongue, generating text based on the captured movements, and converting the text into speech. It is also possible to convert the speech of the other party into text and provide it to the user.

[1655] System configuration

[1656] Capture Device

[1657] The user speaks silently, and the device uses the front camera to capture the user's lip and tongue movements in real time. The captured movements are saved as visual data. The front camera uses a high-resolution camera and a dedicated chip for real-time image processing.

[1658] Data transmission and analysis

[1659] The device sends the captured visual data to a server, where it is encrypted and transmitted securely using the HTTPS protocol.

[1660] The server analyzes the received visual data and generates corresponding text using CNN (Convolutional Neural Network) and RNN (Recurrent Neural Network) technologies, generating highly accurate text based on the user's past learning data and conversational context.

[1661] Text-to-speech

[1662] The server converts the generated text into audio data using TTS (Text to Speech) technology, such as Google Cloud Text-to-Speech. This audio data is generated as a high-quality audio file (e.g., WAV format).

[1663] The server transmits the generated voice data to the other terminal in real time.

[1664] Response processing from the other party

[1665] The other device captures the voice data spoken by the other user and sends it to the server. The voice data is encrypted and sent using the HTTPS protocol.

[1666] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology. In this case, we use "Google Cloud Speech-to-Text."

[1667] The server sends the converted text to the user terminal.

[1668] The device will then display the received text to the user or play it as audio.

[1669] Specific examples

[1670] 1. User training

[1671] The user silently reads the training sample sentence, "I am currently in Tokyo."

[1672] The device's in-camera captures lip and tongue movements and sends the visual data to a server.

[1673] The server analyzes the received data and saves the text "I am in Tokyo now" along with the corresponding visual data as training data.

[1674] 2. Actual call

[1675] The user starts a call and silently says, "The meeting starts at 3:00 PM."

[1676] The device's in-camera captures the user's lip and tongue movements and sends the visual data to a server.

[1677] The server analyzes the visual data and generates the text "The meeting starts at 3pm" based on the training data and context.

[1678] The server converts the text into voice and transmits the voice data to the other terminal.

[1679] The other device plays the received voice data and transmits the user's words to the other device.

[1680] 3. Response from the other person

[1681] The other person will respond by voice, "Understood, I'll join."

[1682] The remote terminal captures the voice data and sends it to the server.

[1683] The server converts the voice data into text and sends it to the user's device.

[1684] The device will then display the received text to the user or play it as audio.

[1685] This allows the system to provide a smooth and secure means of making calls even in quiet environments or where privacy is required. By utilizing past training data and conversation context, the system can generate text with higher accuracy and achieve natural-sounding voice calls.

[1686] Prompt Sentence Examples

[1687] Here are some example prompts to present to a generative AI model:

[1688] User: Use the "Silent Call" system to say silently, "Thank you for your hard work, regarding tomorrow's meeting..."

[1689] Device: Capture lip and tongue movements with the front camera and send the data to the server.

[1690] Server: Analyze the received data, perform syntax analysis and speech synthesis, and send it to the other party as audio.

[1691] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1692] Program processing flow

[1693] Step 1: Capture process

[1694] The user speaks silently.

[1695] The device uses a built-in camera to capture the user's lip and tongue movements in real time, and specifically, it uses a high-resolution camera and a dedicated real-time image processing chip to acquire visual data.

[1696] Input: User's lip and tongue movements

[1697] Output: Visual data (images or video frames)

[1698] Step 2: Send data

[1699] The device encrypts the captured visual data and sends it to a server using the HTTPS protocol.

[1700] The server receives the received visual data.

[1701] Input: Visual data (terminal input data)

[1702] Output: Visual data (received by the server)

[1703] Step 3: Data analysis and text generation

[1704] The server analyzes the received visual data using a high-speed image processing algorithm, specifically, a convolutional neural network (CNN) to extract lip and tongue movements as features.

[1705] The server inputs the features into a recurrent neural network (RNN) model and generates the corresponding text, taking into account the user's past learning data and conversational context.

[1706] Input: Visual data (received data from the server)

[1707] Output: The generated text

[1708] Step 4: Text-to-speech

[1709] The server converts the generated text into voice data using TTS (Text to Speech) technology, for example, by using a "voice synthesis engine."

[1710] The server transmits the generated voice data to the other terminal in real time.

[1711] Input: Generated text

[1712] Output: Audio data

[1713] Step 5: Processing the other party's response

[1714] The other terminal captures the voice data uttered by the other user and transmits it to the server.

[1715] The server converts the received voice data into text using ASR (Automatic Speech Recognition) technology, for example, by using a "voice recognition engine."

[1716] The server sends the converted text to the user terminal.

[1717] Input: Voice data (the other person's input data)

[1718] Output: The converted text

[1719] Step 6: Display to the user

[1720] The device displays the received text to the user or plays it as audio using TTS technology.

[1721] Input: Translated text

[1722] Output: Text or audio displayed to the user

[1723] This allows the overall processing flow to be specified and the operations and data flow at each step to be clearly explained.

[1724] (Application example 1)

[1725] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1726] In security work, it is necessary to be able to make calls and share information smoothly even in quiet environments and when privacy is required. Current voice communication systems have the risk of information leaking to those around them due to the sound they emit, and are difficult to use in quiet environments. A system that can solve these problems is needed.

[1727] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1728] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for transmitting the generated speech data to another terminal, and means for capturing speech data from the other party and converting it into text. This enables smooth and safe silent communication and information sharing even in security work that requires quiet environments and privacy.

[1729] "User" means an individual who uses the System to make silent calls.

[1730] "Lip and tongue movements" refers to the movements inside and outside the mouth that a user makes when silently speaking.

[1731] "Capture" refers to the act of recording visual data using an in-camera or other device.

[1732] "Text generation" refers to the process of analyzing captured visual data and creating corresponding string data.

[1733] "Text-to-speech" refers to the technique of converting generated text into audio data.

[1734] "Other device" refers to the device used by the person with whom you are making a call.

[1735] "Server" refers to a central device that performs processing such as data analysis, text generation, and speech conversion.

[1736] "Silent security call application" refers to software that enables silent calls in security work where quiet environments and privacy are required.

[1737] This invention relates to a "Silent Security Call Application" that enables users to make silent calls in situations where quiet environments and privacy protection are required in security work. The following is a detailed explanation of how to specifically implement this system.

[1738] System program generation

[1739] The system achieves the following main functions:

[1740] 1. Hardware Configuration

[1741] User device: Use a device equipped with an in-camera, such as a smartphone or smart glasses.

[1742] Server: Use a high-performance computer server to perform data analysis, text generation, and speech conversion.

[1743] 2. Software Configuration

[1744] OpenCV: A library used to capture the user's lip and tongue movements.

[1745] gTTS (Google Text-to-Speech): A library for converting generated text to speech.

[1746] speech_recognition: A library for converting speech into text.

[1747] Program processing

[1748] 1. Data Capture

[1749] On the device: The front camera captures the user's lip and tongue movements in real time.

[1750] Data transmission: The captured visual data is transmitted from the device to the server.

[1751] 2. Text Generation

[1752] Server: Analyzes the received visual data and generates corresponding text, using the user's past learning data and conversational context.

[1753] 3. Audio conversion

[1754] Server: Converts the generated text into speech data using the gTTS library.

[1755] Data transmission: The generated voice data is transmitted to other devices in real time.

[1756] 4. Processing replies from the other party

[1757] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1758] Server: Converts the received audio data into text and displays it to the user or plays it back as audio.

[1759] Specific examples

[1760] 1. User training

[1761] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1762] Device: The front camera captures lip and tongue movements and sends visual data to the server.

[1763] Server: Analyzes the received data and saves the text "I am currently in Tokyo" along with the corresponding visual data as training data.

[1764] 2. Actual call

[1765] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1766] Device: The front camera captures the user's lip and tongue movements and sends the visual data to the server.

[1767] Server: Analyzes the visual data and generates text such as "The meeting starts at 3pm" based on the training data and context.

[1768] Server: Converts text into speech and sends the speech data to the other device.

[1769] The other device: Plays the received voice data and transmits the user's words to the other device.

[1770] 3. Response from the other person

[1771] Other party: Reply verbally with "Got it, I'll join."

[1772] Remote device: Captures audio data and sends it to the server.

[1773] Server: Converts voice data into text and sends it to the user's device.

[1774] On the device: The received text is displayed to the user or played as audio.

[1775] Prompt Sentence Examples

[1776] Your application will analyze the user's lip movements to generate text and convert it to speech. The following requirements must be met:

[1777] The user makes a call without speaking.

[1778] The capture device uses the in-camera.

[1779] The data is sent to a server and converted into text and audio data.

[1780] The software used is Python's OpenCV, gTTS, and speech_recognition libraries.

[1781] This makes it possible for the system to provide a means for smooth, safe, and silent communication and information sharing, even in quiet environments or places where privacy is required.

[1782] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1783] Step 1:

[1784] The user speaks silently. The user's lip and tongue movements are captured in real time by the in-camera. The input is the user's lip and tongue movements, and the output is visual data. The device records this visual data.

[1785] Step 2:

[1786] The terminal transmits the captured visual data to the server, where it is sent in the form of a data package that reaches the server over the network. The input is the visual data, and the output is the data package received at the server side.

[1787] Step 3:

[1788] The server analyzes the received visual data and generates corresponding text. To analyze the visual data, it uses machine learning models, which utilize the user's past learning data and conversational context. The input is the visual data, and the output is the generated text.

[1789] Step 4:

[1790] The server converts the generated text into speech data using speech synthesis technology and the gTTS (Google Text-to-Speech) library. The input is text data and the output is speech data.

[1791] Step 5:

[1792] The server then transmits the generated audio data to other devices in real time. This must be done quickly over the network, in the form of a data package. The input is the audio data, and the output is an audio file that can be played on other devices.

[1793] Step 6:

[1794] The other device plays back the received voice data and conveys what the other party said to the user. The input is voice data, and the output is the played-back voice.

[1795] Step 7:

[1796] The other party responds by voice. The other party's device captures the voice data and sends it to the server. The input is the other party's voice, and the output is the voice data sent to the server.

[1797] Step 8:

[1798] The server converts the received voice data into text using voice recognition technology. The input is voice data and the output is text data.

[1799] Step 9:

[1800] The terminal displays the text data sent from the server to the user or plays it aloud. The input is the text data, and the output is the displayed text or played audio.

[1801] This allows the entire system to smoothly communicate and share information in security operations without making any sound.

[1802] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1803] This invention relates to a "Silent Call" system that allows users to make calls without speaking. This system achieves more natural conversations by capturing the user's lip and tongue movements and combining them with an emotion engine. Below, we will provide an overview of the system's program and its specific processing in natural language.

[1804] System configuration

[1805] Capture device and emotion recognition engine

[1806] User: Speak silently.

[1807] On-device: Uses the front camera to capture the user's lip and tongue movements in real time, and uses an emotion recognition engine to analyze the user's emotions from their facial expressions.

[1808] Device: The captured movement and facial expression data is stored as visual data.

[1809] Data transmission and analysis

[1810] Terminal: Sends captured visual data to the server.

[1811] Server: Analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, utilizing the user's past learning data and conversational context.

[1812] Text-to-speech

[1813] Server: Based on the generated text and the recognized emotion data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[1814] Server: Sends the generated audio data to other devices in real time.

[1815] Response processing from the other party

[1816] Remote device: As with a normal voice call, the remote device captures the voice data from the remote device and sends it to the server.

[1817] Server: Converts the received voice data into text using speech recognition technology and displays it to the user or plays it back as audio.

[1818] Specific examples

[1819] 1. User training

[1820] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1821] Device: The front camera captures lip and tongue movements and sends visual data to the server. It also captures facial expressions and sends them as emotional data.

[1822] Server: Analyzes the received data and saves the text "I am currently in Tokyo" and the corresponding visual and emotional data as training data.

[1823] 2. Actual call

[1824] User: Starts a call and silently says, "My meeting starts at 3 PM."

[1825] Device: The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[1826] Server: Analyzes visual and emotional data and generates text such as "The meeting starts at 3 p.m." based on the training data and context.

[1827] Server: Converts text into speech using emotion-based speech synthesis technology and sends the speech data to the other device.

[1828] The other device: Plays the received voice data and transmits the user's words to the other device.

[1829] 3. Response from the other person

[1830] Other party: Reply verbally with "Got it, I'll join."

[1831] Remote device: Captures audio data and sends it to the server.

[1832] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1833] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[1834] This system provides a means for smooth and secure communication even in quiet environments or places where privacy is required. By analyzing emotional data from the user's facial expressions and reflecting emotions in text generation and speech conversion, it is possible to achieve more natural and human-like communication.

[1835] The processing flow will be explained below.

[1836] Training Data Collection Processing Steps

[1837] Step 1: Training Initial Setup

[1838] User: Select "Silent Call" and start training.

[1839] Terminal: Presents training example sentences to the user.

[1840] Step 2: Silent reading of sample sentences

[1841] User: Silently read the provided sample sentences aloud.

[1842] On the device: The front camera captures the user's lip and tongue movements, as well as facial expressions to obtain emotional data.

[1843] Step 3: Sending visual data

[1844] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[1845] Step 4: Data analysis and training

[1846] Server: Analyzes the received visual and emotional data and saves the training data as data corresponding to the text.

[1847] Actual call processing steps

[1848] Step 1: Start the call

[1849] User: Select the person to call and start the call.

[1850] Device: Activate the front camera and emotion recognition engine and prepare.

[1851] Step 2: Silent speech

[1852] User: Speak silently.

[1853] On the device: The front camera captures the user's lip and tongue movements in real time, while the emotion engine simultaneously captures facial expression data.

[1854] Step 3: Sending visual data

[1855] Terminal: Sends captured lip and tongue movements and facial expression data to the server.

[1856] Step 4: Text Generation

[1857] Server: Analyzes the received visual data and generates text based on the user's past learning data and conversational context.

[1858] Step 5: Analyze the sentiment data

[1859] Server: Analyzes the user's emotions from facial expression data and assigns the recognized emotions to text data.

[1860] Step 6: Audio conversion

[1861] Server: The text generated reflecting emotions is converted into voice data using speech synthesis technology.

[1862] Step 7: Sending audio data

[1863] Server: Sends the generated voice data to the other device in real time.

[1864] Step 8: Playing Audio

[1865] The other device: Plays the received voice data and transmits the user's words to the other device.

[1866] Steps for processing a reply from the other party

[1867] Step 1: Capture audio data

[1868] Other party: Speak as if you were making a normal voice call.

[1869] Remote device: Captures the remote device's voice data and sends it to the server.

[1870] Step 2: Analyzing the audio data

[1871] Server: Converts the received voice data into text using voice recognition technology.

[1872] Step 3: Send a text

[1873] Server: Sends the generated text to the user's device and also analyzes emotion data if necessary.

[1874] Step 4: Display text or play audio

[1875] Terminal: Receives text and emotion data and displays it to the user or plays it back as audio.

[1876] The above is the flow of specific processing steps of the invention that combines an emotion engine. This system enables smooth and safe calls even in quiet environments or places where privacy is required, and also enables natural communication that reflects emotions.

[1877] Example 2

[1878] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1879] Conventional voice communication systems have the problem that it is difficult for users to communicate smoothly in quiet environments or places where privacy is required, and they also have the problem that it is difficult to express emotions, making it difficult to realize natural conversations.

[1880] The specific processing by the specific processing unit 290 of the data processing device 12 in the second embodiment is realized by the following means.

[1881] In this invention, the server includes means for capturing the movement of the user's lips and tongue, means for generating text based on the captured movement, means for converting the generated text into speech, means for analyzing the user's facial expression to obtain emotional data, means for generating speech data reflecting the emotion based on the generated text and the obtained emotional data, and means for transmitting the generated speech data to another terminal in real time, thereby enabling smooth and natural conversation even in quiet environments or places where privacy is required.

[1882] "User" refers to an individual who uses the system to make silent speech.

[1883] "Means for capturing lip and tongue movements" refers to a device equipped with a camera and / or sensors for capturing the user's lip and tongue movements in real time.

[1884] "Means for generating text based on captured movements" refers to algorithms or software that analyze captured lip and tongue movements and generate corresponding text data.

[1885] "Generated text-to-speech means" refers to an algorithm or program that uses speech synthesis technology to convert text data into speech data.

[1886] "Means of analyzing facial expressions to obtain emotional data" refers to technology and software that captures a user's facial expressions and recognizes and analyzes emotions from those expressions.

[1887] "Means for generating voice data that reflects emotions based on acquired emotional data" refers to algorithms and programs for synthesizing voice that reflects emotional data.

[1888] "Means for transmitting generated voice data to other terminals in real time" refers to communication technologies and protocols for delivering voice data to other terminals without delay.

[1889] The present invention relates to a "Silent Call" system that enables users to have natural conversations without vocalizations in quiet environments or situations where privacy is required. This system acquires emotional data by capturing the movements of the user's lips and tongue and analyzing their facial expressions, and combines them to generate text and voice data for voice calls.

[1890] System configuration

[1891] Capture device and emotion recognition engine

[1892] User: Speaks silently. Imagine a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[1893] Device:

[1894] An in-camera (e.g., a commercially available high-resolution camera) is used to capture the user's lip and tongue movements in real time, and an emotion recognition engine (e.g., emotion analysis software) is used to analyze the user's emotions from their facial expressions and obtain emotion data.

[1895] The captured movement and emotion data is stored as visual data.

[1896] Hardware used:

[1897] High-resolution in-camera, microprocessor

[1898] Software used:

[1899] Sentiment analysis software, data analysis algorithms

[1900] Text generation and speech synthesis

[1901] Device:

[1902] The captured visual and emotional data is sent to a server, using an internet connection for data communication.

[1903] server:

[1904] The received visual data is analyzed and the text "Hello, how are you?" is generated from lip and tongue movements and facial expressions using a generative AI model (e.g., a machine learning model, as the term is generally used).

[1905] Based on the generated text and the acquired emotional data, it is converted into voice data using emotion-reflecting voice synthesis technology (commonly known as voice synthesis software).

[1906] Hardware used:

[1907] Server computers, high-speed data processing units

[1908] Software used:

[1909] Machine learning models, speech synthesis software

[1910] Communication and playback

[1911] server:

[1912] The generated voice data is sent to other devices in real time using a real-time protocol.

[1913] Remote device:

[1914] The received voice data is played back and the user's silent speech is conveyed to the other party.

[1915] Hardware used:

[1916] Communication modem, speaker

[1917] Specific examples

[1918] 1. User training

[1919] User: Silently reads the training sample sentence, "I am currently in Tokyo."

[1920] Device:

[1921] The front camera captures lip and tongue movements and sends visual data to the server, as well as facial expressions, which are then sent as emotional data.

[1922] server:

[1923] The received data is analyzed, and the visual and emotional data corresponding to the text "I am in Tokyo now" are saved as training data.

[1924] 2. Actual call

[1925] User: Starts a call and silently says "My meeting starts at 3 PM."

[1926] Device:

[1927] The front camera captures the user's lip and tongue movements and sends visual data to the server, as well as emotional data from facial expressions.

[1928] server:

[1929] It analyzes visual and emotional data and generates the text "The meeting starts at 3 p.m." based on the training data and context.

[1930] server:

[1931] Text is converted into speech using emotion-reflecting speech synthesis technology, and the speech data is sent to the other party's terminal.

[1932] Remote device:

[1933] The received voice data is played back and the user's words are conveyed to the other party.

[1934] 3. Response from the other person

[1935] Other party: Reply verbally with "Got it, I'll join."

[1936] Remote device: Captures audio data and sends it to the server.

[1937] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1938] Device: Receives text and emotion data and displays it to the user, or synthesizes the text into speech and plays it back with the emotion reflected.

[1939] This system enables smooth and natural conversations even in quiet environments or places where privacy is required. It also analyzes emotional data from the user's facial expressions and reflects those emotions in text generation and speech conversion, enabling more human-like communication.

[1940] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1941] Step 1:

[1942] User: Speaks silently. For example, consider a user silently uttering "Hello, how are you?" by moving their lips and tongue.

[1943] Input: The user's lip and tongue movements.

[1944] Output: Real-time lip and tongue movement data.

[1945] Specific actions: The user speaks silently in front of the camera.

[1946] Step 2:

[1947] On the device: The front camera captures the user's lip and tongue movements in real time, and an emotion recognition engine is used to analyze facial expressions and obtain emotional data.

[1948] Input: Real-time lip and tongue movement data, facial expression data.

[1949] Output: Captured movement and emotion data are saved as visual data.

[1950] What it does: The camera and emotion recognition engine work together to record the user's lips, tongue, and facial expressions.

[1951] Step 3:

[1952] Device: The saved visual and emotional data is sent to the server. Data communication is via an internet connection.

[1953] Input: visual and emotional data.

[1954] Output: Sending data to the server.

[1955] What happens: Your device uploads data to a server via your internet connection.

[1956] Step 4:

[1957] Server: Analyzes the received visual data and generates text from lip and tongue movements.

[1958] Input: Captured visual data, emotion data.

[1959] Output: The generated text.

[1960] What it does: Uses a generative AI model to generate the text "Hello, how are you?" from lip and tongue movements.

[1961] Step 5:

[1962] Server: Based on the generated text and the acquired emotional data, it converts it into voice data using emotion-reflecting speech synthesis technology.

[1963] Input: Text data, emotion data.

[1964] Output: Emotionally-informed audio data.

[1965] Specific operation: Create emotionally-charged speech data using a text generation model.

[1966] Step 6:

[1967] Server: The generated voice data is sent to other devices in real time using a real-time protocol.

[1968] Input: Emotionally-charged speech data.

[1969] Output: Audio data sent to other devices.

[1970] Specific operation: Sends voice data to a remote terminal using a real-time protocol.

[1971] Step 7:

[1972] The other party's terminal: Plays back the received voice data and conveys the user's silent speech to the other party.

[1973] Input: Audio data sent from the server.

[1974] Output: The audio played.

[1975] Specific operation: Audio is output from the speaker and the receiver listens to the user's speech.

[1976] Step 8:

[1977] Recipient: "Yes, I'm fine" replies in a normal voice.

[1978] Input: The other person's speech.

[1979] Output: Audio data.

[1980] Specific operation: Convert analog voice into digital data and input it into the terminal.

[1981] Step 9:

[1982] Remote device: Captures the remote device's voice data and sends it to the server.

[1983] Input: The other party's voice data.

[1984] Output: Sends audio data to the server.

[1985] Specific operation: After capturing the audio, send the data to the server.

[1986] Step 10:

[1987] Server: Converts voice data into text and sends it to the user's device. It also analyzes emotional data using an emotion engine.

[1988] Input: The other party's voice data.

[1989] Output: Text data, emotion data.

[1990] Specific operation: Converts speech into text using speech recognition technology and analyzes emotional data.

[1991] Step 11:

[1992] Device: Receives text and emotion data and displays it to the user, or plays back a synthesized voice that reflects the emotion.

[1993] Input: Text data, emotion data.

[1994] Output: The displayed text or played audio.

[1995] Specific Action: Display the text to the user or play it as emotional speech.

[1996] (Application example 2)

[1997] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1998] In modern society, there is a growing demand for calls and communication in situations where confidentiality and privacy are essential. However, conventional voice calls pose the risk of sound leakage and eavesdropping, making them unsuitable for use in quiet environments or sensitive situations. There is also a need for a method to convey text information to the other party more accurately and emotionally. Technology is needed to solve this problem and enable emotionally charged, natural communication even in silent environments.

[1999] The specific processing by the specific processing unit 290 of the data processing device 12 in Application Example 2 is realized by the following means. In this invention, the server includes means for capturing the movements of the user's lips and tongue, means for generating text based on the captured movements, means for recognizing the user's emotions and reflecting them in the generated text, and means for converting the generated text into speech. This makes it possible to accurately generate text from the movements of the lips and tongue even in a silent environment, enabling voice calls that reflect the user's emotions.

[2000] "Means for capturing the movement of a user's lips and tongue" refers to devices or technologies that use an in-camera or specific sensors to capture the movement and shape of the area around a user's mouth in real time and record it as digital data.

[2001] "Means for generating text from captured movements" means algorithms and software for analyzing captured lip and tongue movement data and converting it into corresponding characters, words, or sentences.

[2002] "Means for recognizing user emotions and reflecting them in text generation" refers to technology that uses image analysis technology and facial expression recognition engines to detect emotions from the user's facial expressions and incorporate that emotional information into the text generation process.

[2003] "Means for converting generated text into speech" refers to software and hardware that utilizes speech synthesis technology using artificial intelligence to convert text data into speech data and generate natural-sounding speech that reflects emotion.

[2004] "Means for improving the accuracy of text generation by utilizing conversational context and the user's past learning data" refers to algorithms and databases that utilize the user's past conversation history and contextual information from dialogue to improve the content and accuracy of the generated text.

[2005] "Means for transmitting the generated voice data and emotional data to another terminal in real time and conversing with the other party" refers to communication technologies and protocols that enable real-time communication by quickly transmitting the generated voice data and emotional information to another device via a network.

[2006] The "Secure Silent Call System" that realizes this application example captures the movement of the user's lips and tongue, generates text based on that movement, recognizes the user's emotions and reflects them in the generated text, converts the generated text into speech, and transmits it to other devices in real time, enabling communication.

[2007] System configuration

[2008] 1. Capture device and emotion recognition engine

[2009] user:

[2010] This feature is intended for use in quiet environments such as conference rooms and libraries.

[2011] Device:

[2012] The system uses a built-in camera and specific sensors to capture the user's lip and tongue movements in real time. It also detects facial expressions and analyzes the user's emotions using an emotion recognition engine. The hardware used includes a standard webcam and a highly accurate facial recognition sensor.

[2013] Device:

[2014] The captured movement and facial expression data are stored as visual data and sent to a server.

[2015] 2. Data submission and analysis

[2016] server:

[2017] The software analyzes the received visual data and generates text based on lip and tongue movements and facial expressions, leveraging the user's past learning data and conversational context. The software includes a lip movement recognition engine (LipMovementRecognizer) and an emotion recognition engine (EmotionRecognizer).

[2018] 3. Text-to-speech

[2019] server:

[2020] The generated text and emotion data are then converted into emotion-reflecting voice data using speech synthesis technology. This process is carried out using TextToSpeech software.

[2021] server:

[2022] The generated voice data is transmitted to another terminal in real time.

[2023] 4. Processing replies from the other party

[2024] Remote device:

[2025] As with a normal voice call, the voice data from the other party is captured and sent to the server.

[2026] server:

[2027] The received voice data is converted into text using voice recognition technology, and the text and emotional data are sent to the user's device.

[2028] Device:

[2029] It receives text and emotion data and displays it to the user. It also synthesizes the text into speech and plays it back with the emotion reflected.

[2030] Specific examples

[2031] 1. User training:

[2032] The user silently reads the training sample sentence, "I am in Tokyo now." The device captures lip and tongue movements with the front camera and sends the visual data to the server. It also captures facial expressions and sends them as emotional data. The server analyzes the received data and saves the visual and emotional data corresponding to the text "I am in Tokyo now" as training data.

[2033] 2. Actual call:

[2034] The user silently utters, "The meeting will start at 3 PM." The device captures lip and tongue movements with the in-camera and sends the visual data to the server. It also sends facial expression data for emotion recognition. The server analyzes the visual and emotional data and generates the text "The meeting will start at 3 PM" based on the learning data and context. It then converts this into speech using emotion-reflecting speech synthesis technology and sends it to the other device. The other device receives the voice data and conveys the user's remarks to the other party.

[2035] 3. Response from the other person:

[2036] The other party responds by saying "Got it. I'll join." The other party's device captures the voice data and sends it to the server. The server converts the voice data into text and sends the text and emotion data to the user's device. The device receives the text and emotion data and displays it to the user or plays it back as voice.

[2037] Example prompt for a generative AI model:

[2038] "Analyze the user's lip and tongue movements and generate text. Use the following video stream data. Also, recognize emotions from facial expressions and generate audio data that reflects them. Video stream data example: [Image data] Emotion data example: [Emotion label based on image data]"

[2039] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[2040] Step 1:

[2041] The user silently speaks. For example, in a quiet environment, the user silently utters, "The meeting will start at 3:00 PM." At this time, the user moves their mouth in the same way as they would during a normal phone call.

[2042] Step 2:

[2043] The device uses an in-camera and specific sensors to capture the user's lip and tongue movements in real time. The video data captured by the in-camera is input, and visual data for analyzing lip and tongue movements is output. The device also captures the user's facial expressions. This facial expression data is also output as visual data.

[2044] Step 3:

[2045] The terminal transmits the captured visual data to the server. The input is the captured visual data, and the output is the data transmission to the server. Network communication technology is used for data transmission.

[2046] Step 4:

[2047] The server analyzes the received visual data. Specifically, it analyzes lip and tongue movement data and facial expression data to generate text data. The input is the received visual data, and the output is the text data generated by the data analysis process. LipMovementRecognizer and EmotionRecognizer are used for the analysis.

[2048] Step 5:

[2049] The server uses the conversational context and the user's past learning data to improve the accuracy of the generated text. The input is the initial text data, the user's past learning data, and the conversational context, and the output is the final text data with high accuracy. Natural language processing technology is used for data processing.

[2050] Step 6:

[2051] The server converts the final text data and emotion data into speech. The input is the final text data and emotion data, and the output is speech data that reflects the emotion expression. TextToSpeech software is used for speech synthesis.

[2052] Step 7:

[2053] The server transmits the generated voice data to the remote terminal in real time. The input is voice data, and the output is voice data transmitted over the network. A communication protocol is used for transmission.

[2054] Step 8:

[2055] The other device plays back the received voice data and transmits the user's speech to the other device. The input is the received voice data, and the output is the played voice.

[2056] Step 9:

[2057] The other party responds with voice, just like in a normal voice call. For example, they might respond with, "Got it. I'll join." This voice is captured by the other party's device.

[2058] Step 10:

[2059] The remote terminal transmits the captured voice data to the server. The input is the voice data, and the output is the data transmission to the server.

[2060] Step 11:

[2061] The server analyzes the received voice data and converts it into text data. The input is voice data and the output is text data. The analysis is performed using voice recognition technology.

[2062] Step 12:

[2063] The server sends the generated text data and emotion data to the user terminal. The input is the text data and emotion data, and the output is the data transmission to the user terminal.

[2064] Step 13:

[2065] The terminal displays the received text data and emotion data to the user, and also plays the text as speech that reflects the emotion. The input is the text data and emotion data, and the output is the displayed text and the played speech.

[2066] 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 control target 443 to output the result of the specific processing. The microphone 238 acquires voice indicating a user input regarding the result of the specific processing. The control unit 46A transmits voice data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the voice data.

[2067] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[2068] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the robot 414.

[2069] The emotion identification model 59 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 an emotion map (see FIG. 9), which is a specific mapping. Similarly, the emotion identification model 59 may determine the robot's emotion, and the identification processing unit 290 may perform identification processing using the robot's emotion.

[2070] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[2071] These emotions are distributed in the 3 o'clock direction on emotion map 400, and typically fluctuate between relief and anxiety. In the right half of emotion map 400, situational awareness dominates over internal sensations, resulting in a sense of calm.

[2072] The inside of emotion map 400 represents what is going on in the mind, and the outside of emotion map 400 represents behavior, so the further you go outside emotion map 400, the more visible the emotions become (the more they are expressed in behavior).

[2073] Human emotions are based on various balances, such as posture and blood sugar levels. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. Emotions can also be created for robots, automobiles, and motorcycles, based on various balances, such as posture and remaining battery life. When these balances deviate from the ideal, a state of discomfort is indicated, and when they approach the ideal, a state of pleasure is indicated. An emotion map can be generated, for example, based on Dr. Mitsuyoshi's emotion map (Research on Voice Emotion Recognition and Emotional Brain Physiological Signal Analysis Systems, Tokushima University, Doctoral Dissertation: https: / / ci.nii.ac.jp / naid / 500000375379). The left half of the emotion map lists emotions belonging to the "reaction" domain, where sensation is dominant. The right half of the emotion map lists emotions belonging to the "situation" domain, where situational awareness is dominant.

[2074] The emotion map defines two emotions that promote learning. One is a negative emotion on the situation side, around the middle of "repentance" or "reflection." In other words, this occurs 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 a positive emotion on the response side, around "desire." In other words, this occurs when the robot experiences positive feelings such as "I want more" or "I want to know more."

[2075] The emotion identification model 59 inputs user input into a pre-trained neural network, obtains emotion values ​​indicating each emotion shown in the emotion map 400, and determines the user's emotion. This neural network is pre-trained based on multiple pieces of training data that are combinations of user input and emotion values ​​indicating each emotion shown in the emotion map 400. Furthermore, this neural network is trained so that emotions that are located close to each other have similar values, as in the emotion map 900 shown in FIG. 10. FIG. 10 shows an example in which multiple emotions, "relieved," "calm," and "reassuring," have similar emotion values.

[2076] The system according to the present disclosure has been described above mainly with respect to the functions of the data processing device 12, but the system according to the present disclosure is not necessarily implemented on a server. The system according to the present disclosure may be implemented as a general information processing system. The present disclosure may be implemented, for example, as a software program running on a personal computer or an application running on a smartphone, etc. The method according to the present disclosure may be provided to users in the form of SaaS (Software as a Service).

[2077] In the above embodiment, an example was given in which the specific processing is performed by one computer 22, but the technology of the present disclosure is not limited to this, and the specific processing may be distributed and performed by a plurality of computers including the computer 22. For example, the data generation model 58 may be provided in an external device of the data processing device 12, and data may be generated in the external device in accordance with input data.

[2078] In the above embodiment, an example in which the specific processing program 56 is stored in the storage 32 has been described, but the technology of the present disclosure is not limited to this. For example, the specific processing program 56 may be stored in a portable, computer-readable, non-transitory storage medium such as a USB (Universal Serial Bus) memory. The specific processing program 56 stored in the non-transitory storage medium is installed in the computer 22 of the data processing device 12. The processor 28 executes the specific processing in accordance with the specific processing program 56.

[2079] 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.

[2080] It is not necessary to store all 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 all of the specific processing program 56 in the storage 32; only a portion of the specific processing program 56 may be stored.

[2081] The hardware resource for executing a specific process can be any of the following processors: An example of a processor is a CPU, which is a general-purpose processor that functions as a hardware resource for executing a specific process by executing software, i.e., a program. Another example of a processor is a dedicated electrical circuit, such as an FPGA (Field-Programmable Gate Array), a PLD (Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit), which is a processor with a circuit configuration designed specifically for executing a specific process. Each processor has built-in or connected memory, and each processor uses the memory to execute the specific process.

[2082] The hardware resource that executes the specific processing may be configured with one of these various processors, or may be configured with 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). Also, the hardware resource that executes the specific processing may be a single processor.

[2083] As an example of a system configured with a single processor, first, one processor is configured by combining one or more CPUs and software, and this processor functions as a hardware resource that executes a specific process. Second, there is a system that uses a processor that realizes the functions of an entire system including multiple hardware resources that execute a specific process on a single IC chip, as typified by SoC (System-on-a-chip). In this way, a specific process is realized using one or more of the above-mentioned various processors as hardware resources.

[2084] Furthermore, the hardware structure of these various processors can be, more specifically, an electric circuit that combines circuit elements such as semiconductor devices. The specific processing described above is merely an example. Therefore, it goes without saying that unnecessary steps may be deleted, new steps may be added, or the processing order may be rearranged, without departing from the spirit of the invention.

[2085] The above-described description and illustrations are a detailed explanation of the parts related to the technology of the present disclosure and are merely an example of the technology of the present disclosure. For example, the above description of the configuration, functions, actions, and effects is an explanation of an example of the configuration, functions, actions, and effects of the parts related to the technology of the present disclosure. Therefore, it goes without saying that unnecessary parts may be deleted, new elements may be added, or replacements may be made to the above-described description and illustrations within the scope of the gist of the technology of the present disclosure. Furthermore, to avoid confusion and facilitate understanding of the parts related to the technology of the present disclosure, the above-described description and illustrations omit explanations of common technical knowledge that do not require particular explanation to enable the implementation of the technology of the present disclosure.

[2086] All publications, patent applications, and technical standards mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent application, or technical standard was specifically and individually indicated to be incorporated by reference.

[2087] The following is further disclosed regarding the above embodiment.

[2088] (Claim 1)

[2089] means for capturing lip and tongue movements of a user;

[2090] a means for generating text based on the captured motion;

[2091] means for converting the generated text into speech;

[2092] A system including:

[2093] (Claim 2)

[2094] 10. The system of claim 1, further comprising means for utilizing conversational context and user past learning data to improve accuracy of text generation.

[2095] (Claim 3)

[2096] 2. The system according to claim 1, further comprising means for transmitting the generated voice data to another terminal in real time and for carrying out a conversation with the other party.

[2097] (Claim 4)

[2098] 2. The system according to claim 1, further comprising means for converting voice received from a communication partner into text and displaying it to the user and / or reproducing it as voice.

[2099] (Claim 5)

[2100] 2. The system of claim 1, further comprising means for displaying training sample sentences that a user reads silently and for saving the read data.

[2101] The above is the draft of the patent claim. It accurately indicates the distinctive features of this system and emphasizes its novelty and inventive step.

[2102] "Example 1"

[2103] (Claim 1)

[2104] means for capturing lip and tongue movements of a user;

[2105] a means for generating text based on the captured motion;

[2106] means for converting the generated text into speech;

[2107] means for transmitting the generated voice data to another terminal;

[2108] A means for converting voice data received from a remote terminal into text;

[2109] A system including:

[2110] (Claim 2)

[2111] 10. The system of claim 1, wherein the system utilizes conversational context and user past learning data to improve accuracy of text generation.

[2112] (Claim 3)

[2113] 2. The system according to claim 1, wherein the generated voice data is transmitted to another terminal in real time, and a conversation is carried out with the other party.

[2114] "Application Example 1"

[2115] (Claim 1)

[2116] means for capturing lip and tongue movements of a user;

[2117] a means for generating text based on the captured motion;

[2118] means for converting the generated text into speech;

[2119] means for transmitting the generated voice data to another terminal;

[2120] A means of capturing voice data from the other party and converting it into text;

[2121] A system including:

[2122] (Claim 2)

[2123] 10. The system of claim 1, further comprising means for utilizing conversational context and user past learning data to improve accuracy of text generation.

[2124] (Claim 3)

[2125] 10. The system of claim 1, further comprising a silent secure calling application that enables silent calling in a secure environment.

[2126] "Example 2: Combining Emotion Engines"

[2127] (Claim 1)

[2128] means for capturing lip and tongue movements of a user;

[2129] a means for generating text based on the captured motion;

[2130] means for converting the generated text into speech;

[2131] A means for analyzing a user's facial expressions to obtain emotional data;

[2132] a means for generating voice data that reflects emotions based on the generated text and the acquired emotion data;

[2133] means for transmitting the generated voice data to another terminal in real time;

[2134] A system including:

[2135] (Claim 2)

[2136] 10. The system of claim 1, further comprising means for utilizing conversational context and user past learning data to improve accuracy of text generation.

[2137] (Claim 3)

[2138] 2. The system according to claim 1, further comprising means for transmitting the generated voice data to another terminal in real time and for carrying out a conversation with the other party.

[2139] "Application example 2 when combining emotion engines"

[2140] (Claim 1)

[2141] means for capturing lip and tongue movements of a user;

[2142] a means for generating text based on the captured motion;

[2143] A means of recognizing user emotions and reflecting them in text generation;

[2144] means for converting the generated text into speech;

[2145] A system including:

[2146] (Claim 2)

[2147] 10. The system of claim 1, further comprising means for utilizing conversational context and user past learning data to improve accuracy of text generation.

[2148] (Claim 3)

[2149] 2. The system according to claim 1, further comprising means for transmitting the generated voice data and emotion data to another terminal in real time and for carrying out a conversation with the other party. [Explanation of symbols]

[2150] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for capturing lip and tongue movements of a user; a means for generating text based on the captured motion; means for converting the generated text into speech; A system including:

2. 10. The system of claim 1, further comprising means for utilizing conversational context and user past learning data to improve accuracy of text generation.

3. 2. The system according to claim 1, further comprising means for transmitting the generated voice data to another terminal in real time to carry out a conversation with the other party.

4. 2. The system according to claim 1, further comprising means for converting voice received from a communication partner into text and displaying it to the user and / or reproducing it as voice.

5. 2. The system of claim 1, further comprising means for displaying training sample sentences that a user reads silently and for storing the read data. The above is the draft of the patent claim. It accurately indicates the distinctive features of this system and emphasizes its novelty and inventive step.

Images