System

Abstract

A system is provided.SOLUTION: A system comprising: means for obtaining speech input; means for pre-processing the obtained speech; means for transmitting the pre-processed speech data to a server; means for converting the transmitted speech data into another speech; means for receiving the converted speech data from the server; means for encoding the received speech data; and means for playing the encoded 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] Existing voice conversion systems require high-performance equipment and a high-speed Internet connection, resulting in delays in real-time voice conversion. Furthermore, their applications are limited, and there is a lack of easily accessible services. There is a need to solve these issues and provide low-cost, high-quality voice conversion. [Means for solving the problem]

[0005] This invention provides a system including a means for acquiring a voice input, a means for preprocessing the acquired voice, a means for transmitting the preprocessed voice data to a server, a means for converting the transmitted voice data into another voice, a means for receiving the converted voice data from the server, a means for encoding the received voice data, and a means for playing back the encoded voice. This enables high-speed voice conversion, enabling real-time voice conversion with almost no perceptible delay. Furthermore, since voice conversion can be easily performed without requiring special equipment, it is expected to be used in a wide range of applications.

[0006] "Audio input" refers to the collection of audio signals emitted by a user through a device such as a microphone.

[0007] "Preprocessing" is a process in which acquired audio data is processed, such as for noise removal and volume normalization, to enable the conversion model to analyze audio more accurately.

[0008] A "server" is a remote computing system that receives voice data and converts it into another voice based on a specified voice conversion model.

[0009] "Transmitting voice data" means transferring the preprocessed voice data to the server via data communication.

[0010] "Voice conversion" is the process of using AI models and algorithms to convert pre-processed audio data into a different sound.

[0011] "Encoding" is the process of converting received audio data into a format that can be played on the user's terminal.

[0012] "Playback" means outputting encoded audio data as auditory information through speakers or headphones. [Brief explanation of the drawings]

[0013] [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

[0014] 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.

[0015] First, the terms used in the following description will be explained.

[0016] 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).

[0017] 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.

[0018] 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.

[0019] 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.

[0020] 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."

[0021] [First embodiment]

[0022] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0023] 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.

[0024] 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).

[0025] 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.

[0026] 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.

[0027] 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.

[0028] 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.

[0029] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0030] 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.

[0031] 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.

[0032] 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.

[0033] 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."

[0034] The present invention provides a system for converting a user's voice into a voice of another person in real time. The system includes the following means.

[0035] First, we will explain how to obtain voice input: the user speaks into the device's microphone, and this voice input is captured by dedicated software within the device.

[0036] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0037] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0038] Next, we will explain how the server converts voice data. The server processes the received voice data based on an AI voice conversion model and converts it into another specified voice. This voice conversion uses a highly accurate deep learning model.

[0039] Next, we will explain how to receive the converted voice data from the server. The server re-encodes the converted voice data and sends it to the terminal. The terminal receives this encoded voice data.

[0040] Then, we explain how to encode the audio data. The terminal decodes the received audio data and converts it into a playable format. The encoded audio data is then processed so that it can be used by the user in a specific format.

[0041] Finally, we explain how to play the encoded audio data. The device plays the audio data encoded in a playable format through speakers or headphones. This allows the user to listen to the new audio converted from their own voice in real time.

[0042] Illustrative Usage Scenarios

[0043] For example, when user A says "hello," the audio is captured by the device's microphone. The device removes noise from the audio and sends it to the server. The server processes the received audio using an AI voice conversion model and converts it into user B's voice. This converted audio data is then sent back from the server to the device, where it is played back. This series of processes is performed in real time, so user A can hear "hello" in user B's voice without any delay.

[0044] As described above, the present invention effectively executes all processes from acquiring voice input to converting and playing, thereby providing users with an advanced real-time voice conversion experience.

[0045] The processing flow will be explained below.

[0046] Step 1:

[0047] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0048] Step 2:

[0049] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0050] Step 3:

[0051] The device performs preprocessing on the acquired audio data, such as noise reduction and volume normalization, which includes filtering and volume adjustment using an audio signal processing library.

[0052] Step 4:

[0053] The terminal packages the pre-processed audio data into a suitable format (e.g., binary format) for transmission to the server, which transmits the packaged data over the Internet to the server.

[0054] Step 5:

[0055] The server analyzes the received voice data and inputs it into an AI voice conversion model, which is based on a deep learning algorithm to extract voice features and convert them into a specified target voice.

[0056] Step 6:

[0057] The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0058] Step 7:

[0059] The terminal receives the encoded voice data sent from the server, and this reception process uses a common data communication protocol.

[0060] Step 8:

[0061] The device decodes the received audio data and converts it into a playable format. This decoding process is performed using an audio playback library.

[0062] Step 9:

[0063] The device then plays the decoded audio data over a speaker or headphones, allowing the user to hear the converted voice of another person in real time.

[0064] Example 1

[0065] 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."

[0066] Conventional speech conversion systems often suffer from degradation of speech quality and conversion delays during the process from speech capture to final speech playback. Furthermore, noise removal accuracy is low, which can result in a loss of naturalness in the converted speech. Furthermore, it is difficult to utilize highly accurate models in the conversion process. Therefore, converting speech in real time while providing high-quality speech remains a challenge.

[0067] 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.

[0068] In this invention, the server includes means for acquiring a user's voice, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server via a data communication network, means for converting the transmitted voice data into another voice using a deep learning model, means for receiving the converted voice data from the server, means for decoding the received voice data into a playable format, and means for playing the decoded voice on an audio output device, thereby enabling voice conversion in real time while maintaining the quality of the voice.

[0069] "User" refers to the person who uses the system and is responsible for inputting voice and listening to the final converted voice.

[0070] "Speech" refers to the sound produced by a person, which is picked up by a microphone and processed within the system.

[0071] "Preprocessing" refers to processing that removes noise and adjusts the volume of acquired audio data to improve the quality of the data.

[0072] "Data communications network" means a digital communications infrastructure, such as the Internet, used to transmit voice data from a terminal to a server or from a server to a terminal.

[0073] The "server" is a computer system that handles the conversion of voice data and performs voice conversion using a deep learning model.

[0074] A "deep learning model" is a type of artificial intelligence that uses neural networks to perform advanced conversion and analysis of voice data.

[0075] "Decoding" is the process of converting encoded data back into its original, playable format.

[0076] "Audio output device" refers to a device, such as a speaker or headphones, used to play decoded audio to a user.

[0077] "Noise reduction" is the process of removing unwanted noise from audio data, improving the clarity of the audio.

[0078] "Playable format" means that audio data has been converted into an appropriate format so that the user can listen to the audio.

[0079] This invention is a system that converts user-spoken speech into another speech in real time. This system effectively executes a series of processes from speech input to speech output, focusing on three elements: the user, the terminal, and the server. The specific configuration and operation are described below.

[0080] First, the user speaks into the device's microphone. This speech is captured in real time using dedicated software such as Audacity. The captured speech data is then pre-processed on the device. This pre-processing includes noise reduction and volume adjustment to improve the quality of the speech data. Audacity's "noise reduction" and "volume normalization" functions are used in this process.

[0081] The preprocessed audio data is then sent over a data communications network to a server, typically using a protocol such as an HTTP POST request. The server processes the received audio data using a deep learning model, such as Google Cloud Text-to-Speech or OpenAI's Whisper model, to convert it into a different sound.

[0082] The converted audio data is then encoded on the server and sent back to the device via the data communications network. For example, ffmpeg is used to encode the audio data into an audio format such as FLAC or AAC. The device then decodes the received audio data and converts it into a playable format. Once decoded, the audio data is played on the device's audio output device (speakers or headphones) using the device's VLC media player or the system's default audio playback function.

[0083] For example, when user A says "hello," the audio is captured through the device's microphone. Audacity software captures the audio in real time and performs noise reduction and volume normalization. The processed audio data is sent to the server using an HTTP POST request. The server uses the Google Cloud Text-to-Speech API to convert the audio into user B's voice and sends the converted audio back to the device. On the device, the FLAC format audio data is decoded into MP3 format using ffmpeg and played back using VLC media player. In this way, user A can hear "hello" in user B's voice, converted in real time.

[0084] As described above, the present invention executes the entire process from acquiring voice input to conversion and playback with high quality and in real time, providing users with an advanced voice conversion experience.

[0085] Prompt Sentence Examples

[0086] "Please explain the specific steps involved in converting a user's voice into a different person's voice in real time. Please describe in detail the process from voice capture to conversion and playback, and indicate the software and hardware used in the process."

[0087] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0088] Step 1: Getting voice input

[0089] The user speaks into the microphone of the terminal, for example, saying "hello."

[0090] The device captures this audio using specialized software (e.g., Audacity).

[0091] Input: User's live voice

[0092] Data processing: Audio is captured as digital data using dedicated software

[0093] Output: Unprocessed digital audio data

[0094] Step 2: Preprocessing

[0095] The terminal performs noise reduction and volume adjustment on the captured digital audio data.

[0096] Use Audacity's "Noise Reduction" and "Volume Normalization" features.

[0097] Input: Unprocessed digital audio data

[0098] Data processing: noise removal and volume adjustment

[0099] Output: Pre-processed digital audio data

[0100] Step 3: Sending audio data to the server

[0101] The terminal transmits the pre-processed voice data to a server via a data communication network (eg, the Internet).

[0102] Use an HTTP POST request.

[0103] Input: Preprocessed digital audio data

[0104] Data processing: Data communication via HTTP POST request

[0105] Output: Audio data sent to the server

[0106] Step 4: Audio conversion by server

[0107] The server converts the transmitted voice data into a different voice using a deep learning model (e.g., Google Cloud Text-to-Speech or OpenAI's Whisper model).

[0108] Input: Audio data sent to the server

[0109] Data Computing: Voice Conversion with Deep Learning Models

[0110] Output: Converted audio data

[0111] Step 5: Translated audio data sent from the server

[0112] The server encodes the converted voice data and transmits it to the terminal via a data communication network.

[0113] For encoding, ffmpeg is used to apply audio formats such as FLAC and AAC.

[0114] Input: Converted audio data

[0115] Data processing: Encoding into audio format

[0116] Output: Encoded audio data sent to the device

[0117] Step 6: Decode the audio data

[0118] The terminal decodes the audio data received from the server into a reproducible format.

[0119] Use ffmpeg to decode to MP3 or other formats.

[0120] Input: Encoded audio data received from the server

[0121] Data processing: Decoding into a playable format

[0122] Output: Decoded audio data

[0123] Step 7: Playing back audio data

[0124] The terminal plays the decoded audio data on an audio output device (e.g., a speaker or headphones).

[0125] Use VLC media player or the system's default audio playback function for playback.

[0126] Input: Decoded audio data

[0127] Data calculation: Audio data playback

[0128] Output: Playback of the converted audio from an audio output device

[0129] This allows the user to listen to the converted audio in real time.

[0130] (Application example 1)

[0131] 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."

[0132] Conventional voice conversion systems have problems when converting a user's voice into a different person's voice in real time, such as difficulty in achieving high-quality conversion, insufficient countermeasures against noise contamination, and difficulty in instantly adding a wide variety of voice effects depending on the distribution. In particular, for distribution services of entertainment and educational content, there is a demand for real-time conversion into a different voice, as well as support for live distribution and on-demand effects, in order to improve the user experience.

[0133] 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.

[0134] In this invention, the server includes means for acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for converting a user's voice into another voice in real time, means for live-streaming the converted voice data, and means for adding on-demand voice effects, thereby enabling real-time, high-quality voice conversion, the addition of a variety of effects, and the provision of attractive content in a distribution service.

[0135] The "means for acquiring voice input" is a function for acquiring the voice uttered by the user as digital data using a device such as a microphone.

[0136] The "means for preprocessing the acquired audio" is a function for performing noise removal and volume adjustment on the acquired audio data to improve the quality.

[0137] The "means for transmitting preprocessed voice data to a server" is a function for transmitting preprocessed voice data to a server via the Internet or other communication means.

[0138] "Means for converting transmitted voice data into another voice" refers to a function that converts voice data received on a server into the voice of another person using an AI model, etc.

[0139] The "means for receiving converted voice data from the server" is a function for transmitting the converted voice data from the server again to the client terminal and receiving it.

[0140] The "means for encoding received audio data" is a function that performs processing to convert audio data received from the server into a format that can be used by the user.

[0141] "Means for playing encoded audio" refers to a function for playing encoded audio data through speakers or headphones.

[0142] "Means for converting a user's voice into a different voice in real time" refers to technology and devices for converting a user's voice into a different voice as the user speaks.

[0143] The "means for live distribution of converted voice data" is a function for distributing converted voice data in real time via the Internet.

[0144] The "means for adding on-demand audio effects" is a function for adding new audio effects to existing audio data to improve its entertainment value and educational effectiveness.

[0145] The present invention is a system for converting a user's voice into another person's voice in real time. The implementation of the present invention includes the following components and procedures.

[0146] System Configuration

[0147] 1. Hardware:

[0148] Device: A smartphone or other audio input device that has a microphone and speaker.

[0149] Server: A high-performance computing device for processing voice data online.

[0150] 2. Software:

[0151] WebRTC: A library for capturing, transmitting, and receiving voice data in real time.

[0152] TensorFlow: A machine learning library for building and using AI voice models.

[0153] Flask: A web framework for processing audio data on the server side.

[0154] Processing flow

[0155] 1. Audio capture:

[0156] The user speaks into the smartphone's microphone, and the device captures the voice input and stores it as digital data using data capture software (e.g., WebRTC).

[0157] 2. Pretreatment:

[0158] The captured audio data is pre-processed with noise removal, volume adjustment, etc. This improves the quality of the audio data and increases the accuracy of the subsequent conversion process.

[0159] 3. Data transmission:

[0160] The pre-processed audio data is then transmitted from the device to a server over the internet in real time using WebRTC.

[0161] 4. Audio conversion:

[0162] The server converts the received voice data into another voice using an AI voice conversion model (using TensorFlow). Deep learning models are applied in this step to achieve high-precision voice conversion.

[0163] 5. Data Reception:

[0164] The converted voice data is sent from the server back to the terminal and received in real time.

[0165] 6. Encoding and Playback:

[0166] The device then encodes the received audio data into a playable format and plays it through speakers or headphones, allowing the user to hear the converted version of their own voice.

[0167] Specific examples

[0168] For example, a user may want to change their voice to sound like a historical figure during an online educational lecture. When the user says "hello," the voice is captured by the device's microphone and preprocessed to remove noise. The preprocessed voice data is then sent over the Internet to a server, where an AI model on the server converts the voice data into the voice of the specified person. The converted voice data is then sent back to the device and played back in real time on the device.

[0169] Prompt Sentence Examples

[0170] Please upload an audio file from the user:

[0171]

[0172] Choose an AI model to transform your voice:

[0173] In this way, the present invention provides users with a high quality real-time voice conversion experience, which is very useful in the fields of entertainment and education.

[0174] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0175] Step 1:

[0176] The user speaks into the smartphone's microphone. The device receives this voice input and captures it as digital data. Here, the input data is the user's voice data, and the output data is the digitized voice data. In this step, the microphone is used to convert the analog voice signal into digital data, which is then saved by the voice capture software.

[0177] Step 2:

[0178] The device performs preprocessing on the captured audio data. This involves processes to improve the quality of the audio data, such as noise removal and volume adjustment. Here, the input is the digital audio data captured in step 1, and the output is the preprocessed audio data. Specifically, a library such as webrtcvad is used to remove noise and apply a volume normalization algorithm.

[0179] Step 3:

[0180] The preprocessed voice data is transmitted to a server via the Internet. The terminal divides the digital data into packets and transmits them using a communication module, where the input is the preprocessed voice data and the output is the data packets transmitted via network communication.

[0181] Step 4:

[0182] The server converts the received voice data into a different voice using an AI voice conversion model (using a generative AI model). The server uses the received data as input, applies the generative AI model, and outputs voice data of a different person. Specifically, it executes a deep learning model using the TensorFlow library and performs voice conversion based on the specified prompt sentence.

[0183] Step 5:

[0184] The converted voice data is sent back to the terminal from the server, where it is re-encoded into packets and sent to the terminal via the Internet, where the input is the converted voice data and the output is the data packets received by the terminal via network communication.

[0185] Step 6:

[0186] The terminal decodes the received converted audio data into a playable format. The decoding process involves converting the audio data using signal processing algorithms and converting it into a file format for playback through speakers or headphones. Here, the input is the data packets received from the server, and the output is playable audio data.

[0187] Step 7:

[0188] The device plays the decoded audio data over speakers or headphones, allowing the user to listen to another person's voice in real time. Here, the input is the decoded audio data, and the output is the physical playback of audio. Specifically, audio playback software converts the digital audio signal to an analog signal and plays it on an audio output device.

[0189] 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.

[0190] The present invention provides a system for converting a user's voice into a different person's voice in real time, and further recognizing the user's emotions and reflecting them in the voice conversion. The system includes the following means.

[0191] First, we will explain how to obtain voice input. A user speaks into the microphone of the device. For example, the user says "hello." This voice input is captured by dedicated software in the device.

[0192] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0193] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0194] Next, we will explain how the server converts voice data. The server inputs the received voice data into an AI voice conversion model. This model is based on a deep learning algorithm, extracts voice features, and converts them into a specified target voice.

[0195] Next, we will explain the emotion engine. The server analyzes the voice data provided by the terminal and recognizes the user's emotional state. For example, if the user is angry or happy, the emotion engine determines that emotion.

[0196] Furthermore, we describe a method for adjusting the voice conversion model based on the recognized emotions. The server uses feedback from the emotion engine to readjust the voice conversion model to match the user's emotions. This adjustment allows for voice conversion that is faithful to the user's emotions.

[0197] This section explains how to receive converted audio data from the server. The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0198] This section then describes how the terminal decodes and plays back the audio data. The terminal decodes the encoded audio data it receives and converts it into a playable format. This decoding process is performed using an audio playback library.

[0199] Finally, we will explain how to play back the encoded voice data. The terminal plays back the decoded voice data through a speaker or headphones. This allows the user to hear in real time the voice of another person that has been converted from their own voice. This voice reflects the user's emotions.

[0200] Illustrative Usage Scenarios

[0201] For example, if User A says "Hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model and uses an emotion engine to recognize User A's emotion of "anger." The voice conversion model is then adjusted based on this emotion and converted into User B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear User B's angry voice saying "Hello" in real time, without any delay.

[0202] As described above, the present invention effectively executes all processes from voice input acquisition to emotion recognition, conversion, and playback, thereby providing users with an advanced real-time voice conversion experience.

[0203] The processing flow will be explained below.

[0204] Step 1:

[0205] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0206] Step 2:

[0207] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0208] Step 3:

[0209] The device performs preprocessing on the acquired voice data, such as noise reduction and volume adjustment, to remove excess noise to improve the clarity of the voice and to normalize the voice volume to keep it constant.

[0210] Step 4:

[0211] The device packages the preprocessed audio data into an appropriate format (e.g., binary format) for transmission to the server. The packaged data is sent to the server via the Internet. HTTP or WebSocket is often used as the communication protocol.

[0212] Step 5:

[0213] The server analyzes the received voice data and inputs it into an emotion engine that recognizes the user's emotions. The emotion engine identifies the user's emotional state (e.g., joy, anger, sadness, etc.) based on the tone, rhythm, and intonation of the voice.

[0214] Step 6:

[0215] The server acquires the emotion information recognized by the emotion engine and adjusts the AI ​​voice conversion model based on that information. For example, if the user is angry, the AI ​​model will set the voice conversion parameters to reflect the angry emotion.

[0216] Step 7:

[0217] The server uses a trained AI voice conversion model to convert the pre-processed voice data into a different voice, which faithfully matches the target voice based on the user's voice characteristics and emotional information.

[0218] Step 8:

[0219] The server then encodes the converted audio data into a playable format (such as OGG or MP3) and sends it to the device. The encoded data is compressed for efficient transmission.

[0220] Step 9:

[0221] The terminal receives the encoded voice data sent from the server, and the receiving process follows a pre-established communication protocol.

[0222] Step 10:

[0223] The device then decodes the received encoded audio data and converts it into a playable format. This decoding process is performed using an audio playback library, which removes the compression applied during encoding and restores the original audio format.

[0224] Step 11:

[0225] The device then plays the decoded voice data over a speaker or headphones, allowing the user to listen to a voice that faithfully reflects the user's emotional information, providing a more natural communication experience.

[0226] Example 2

[0227] 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."

[0228] Conventional speech conversion systems simply convert speech into another speech without taking into account the user's emotions, resulting in unnatural speech and poor emotional expression. Real-time speech conversion also entails delays, which can negatively impact the user experience. It is necessary to solve these problems and achieve more natural and emotionally rich speech conversion.

[0229] 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.

[0230] In this invention, the server includes means for converting transmitted voice data into a specified target voice, means for analyzing the voice data to recognize the user's emotional state, and means for adjusting the conversion model based on the recognized emotional state, thereby enabling natural voice conversion that reflects the user's emotions.

[0231] "User" refers to a person who uses the system to provide voice input.

[0232] "Terminal" refers to a device that acquires audio, performs preprocessing, and then transmits and receives the audio data to a server.

[0233] "Voice input" refers to voice data uttered by a user through a microphone of a terminal.

[0234] "Preprocessing" refers to the act of performing processes such as noise removal and volume adjustment on audio data to improve data quality.

[0235] "Server" refers to a device that receives voice data sent from a terminal and converts it into the voice of a specified target.

[0236] A "voice conversion model" refers to an algorithm that uses AI technology to convert voice data into another voice.

[0237] "Emotion engine" refers to an algorithm that analyzes and recognizes the user's emotional state from voice data.

[0238] "Adjusting the conversion model" refers to the act of adjusting the voice conversion model based on the analysis results of the emotion engine, and performing conversion that reflects emotions.

[0239] "Encoding" refers to the act of converting audio data into a playable format.

[0240] "Decoding" refers to the act of restoring encoded audio data to a playable format.

[0241] "Playback" refers to the act of allowing a user to listen to audio data through the terminal's speakers or headphones.

[0242] This invention is a system that converts a user's voice into another person's voice in real time, recognizes the user's emotions, and reflects them in the voice conversion. The following hardware and software are used to implement this system.

[0243] First, a device needs a microphone as hardware to acquire voice input. A user inputs voice by speaking into the microphone. For example, a user might say "hello." This voice input is captured by a voice capture application in the device.

[0244] Next, the device preprocesses the captured audio data by noise reduction and volume adjustment using a noise reduction software library (e.g., Audacity). This preprocessing improves the quality of the audio data and increases conversion accuracy.

[0245] The pre-processed voice data is sent to a server over the Internet, usually using the HTTP protocol. The server then inputs the received voice data into an AI voice conversion model (e.g., WaveNet) to convert it into a specified target voice. This model is based on a deep learning algorithm and analyzes the voice characteristics to perform the conversion.

[0246] The server is also equipped with an emotion engine that identifies the user's emotions from the received voice data. Software such as IBM Tone Analyzer is used for emotion recognition. The recognized emotions (e.g., anger, joy, etc.) are fed back to the voice conversion model, which is then adjusted according to the user's emotions.

[0247] The adjusted audio data is then sent from the server to the device. The server encodes the audio data into a playable format (e.g., OGG or MP3) and sends it to the device. The device then decodes the encoded audio data and converts it into a playable format using a playback library (e.g., FFmpeg).

[0248] Finally, the device plays the decoded voice data over a speaker or headphones, allowing the user to hear their own voice transformed into a different person's voice in real time, with the voice reflecting the user's emotions.

[0249] Specific examples

[0250] For example, if user A says "hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model (WaveNet) and uses an emotion engine (IBM Tone Analyzer) to recognize user A's emotion of "anger." The voice conversion model is then adjusted based on this emotion, and the voice is converted into user B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear user B's angry voice saying "hello" in real time, without any delay.

[0251] In this way, the system of the present invention effectively executes a series of processes from voice input acquisition to emotion recognition, conversion, and playback, providing users with an advanced real-time voice conversion experience.

[0252] Example prompt sentences to use

[0253] Example prompt sentence:

[0254] User A says "Hello" in an angry voice. The audio is captured, noise is removed, and then input into the speech conversion model. During conversion, the emotion recognition engine detects the emotion "anger" and converts it into the angry voice of User B. Play the converted audio.

[0255] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0256] Step 1:

[0257] Acquiring voice input

[0258] Input: The user's voice speaking into the device's microphone (e.g., "Hello")

[0259] Specific operation: The user speaks into the microphone of the terminal.

[0260] Data processing: The device uses an audio capture application to capture this audio as digital audio data.

[0261] Output: Captured audio data (e.g., WAV file with a sampling rate of 44.1 kHz)

[0262] Step 2:

[0263] Audio data preprocessing

[0264] Input: Captured audio data

[0265] Specific operation: The device uses a noise reduction software library (e.g., Audacity) to remove noise from the audio data and adjust the volume.

[0266] Data processing: Apply noise filtering algorithms to improve audio clarity.

[0267] Output: Preprocessed audio data (e.g., a denoised WAV file)

[0268] Step 3:

[0269] Sending audio data to the server

[0270] Input: Preprocessed audio data

[0271] Specific operation: The terminal sends audio data to the server using the HTTP protocol.

[0272] Data processing: The audio data is sent in the body of the HTTP request.

[0273] Output: Audio data received by the server

[0274] Step 4:

[0275] Processing with voice conversion models

[0276] Input: Audio data sent to the server

[0277] Specific operation: The server inputs voice data into an AI voice conversion model (e.g., WaveNet).

[0278] Data Computing: Using deep learning algorithms to extract voice features and convert them into the target voice.

[0279] Output: Converted voice data (e.g., a WAV file converted into a different person's voice)

[0280] Step 5:

[0281] Emotion determination using an emotion recognition engine

[0282] Input: Audio data sent to the server

[0283] Specific operation: The server inputs voice data into an emotion recognition engine (e.g., IBM Tone Analyzer).

[0284] Data processing: Analyze the voice data and extract emotional features.

[0285] Output: The user's emotional state (e.g., "anger" or "joy")

[0286] Step 6:

[0287] Tuning the speech conversion model

[0288] Input: Feedback from the emotion recognition engine (user's emotional state)

[0289] Specific operation: The server adjusts the voice conversion model based on the results of the emotion recognition engine.

[0290] Data calculation: Taking into account emotion features, the parameters of the conversion model are changed.

[0291] Output: A voice conversion model that reflects the user's emotions

[0292] Step 7:

[0293] Encoding the converted audio data

[0294] Input: Speech data converted by the trained speech conversion model

[0295] What happens: The server encodes the audio data into a playable format (e.g., OGG or MP3).

[0296] Data processing: Converting file formats using audio encoding algorithms.

[0297] Output: Encoded audio data (e.g. MP3 file)

[0298] Step 8:

[0299] Sending encoded audio data to the device

[0300] Input: Encoded audio data

[0301] Specific operation: The server uses the HTTP protocol to send encoded audio data to the terminal.

[0302] Data processing: The audio data is sent in the body of the HTTP response.

[0303] Output: The encoded audio data received by the device.

[0304] Step 9:

[0305] Decoding audio data

[0306] Input: Encoded audio data

[0307] Specific behavior: The device uses an audio playback library (e.g., FFmpeg) to decode the encoded audio data.

[0308] Data processing: Converting from an encoded format to a playable format.

[0309] Output: Decoded audio data (e.g. WAV file)

[0310] Step 10:

[0311] Playing audio data

[0312] Input: Decoded audio data

[0313] Specific operation: The device plays audio data using a playback device (speaker or headphones).

[0314] Data calculation: The decoded audio is converted from digital to analog and output.

[0315] Output: The audio the user hears (a modified voice, e.g., "Hello" with the emotion "Anger")

[0316] (Application example 2)

[0317] 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."

[0318] In conventional brick-and-mortar stores, it has been difficult to immediately grasp a customer's emotions and respond appropriately. Furthermore, if a customer is angry or happy, it is not possible to communicate with them in a voice that reflects their emotions, which can lead to a decline in the quality of customer service. By solving these problems, it is necessary to provide more effective customer service and improve customer satisfaction.

[0319] 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 acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for recognizing emotions and reflecting them in voice conversion, and means for adjusting voice tone based on the customer's emotions using an application installed on the smart device. This enables real-time voice conversion that reflects customer emotions in physical stores.

[0320] The "means for acquiring voice input" refers to a device or method for acquiring the voice spoken by the user as digital data using an acoustic sensor such as a microphone.

[0321] "Means for preprocessing captured audio" refers to a device or method that performs preprocessing such as noise removal and volume adjustment on captured audio data to improve the quality of the data and increase the accuracy of subsequent processing.

[0322] The "means for transmitting preprocessed voice data to a server" refers to a device or method for transmitting preprocessed voice data to a server via a communication line such as the Internet.

[0323] "Means for converting transmitted voice data into another voice" refers to a device or method for converting received voice data into another voice using an AI voice model or other conversion algorithm.

[0324] The "means for receiving converted voice data from the server" refers to a device or method for the server to send the converted voice data back to the client device via a communication line such as the Internet.

[0325] "Means for encoding received audio data" refers to a device or method for converting received audio data into a playable format (e.g., OGG or MP3 format).

[0326] "Means for playing encoded audio" refers to an apparatus or method for playing encoded audio data using a playback device (speakers or headphones).

[0327] "Means for recognizing emotions and reflecting them in voice conversion" refers to a device or method that analyzes the emotions from the user's voice, adjusts the voice conversion model based on the recognition results, and converts the voice into a different voice that reflects the emotions.

[0328] "Means for adjusting voice tone based on customer emotions using an application installed on a smart device" refers to a device or method that uses software installed on a device such as smart glasses or a smartphone to adjust voice tone in real time during customer service based on the customer's emotions.

[0329] The present invention provides a system that converts a user's voice into a different person's voice in real time, and further recognizes the user's emotions and reflects them in the voice conversion. Specific embodiments of the system are described below.

[0330] 1. System Configuration

[0331] This system is primarily composed of smart devices, internet communications, and servers. It includes the following elements:

[0332] 1. A device with a microphone to capture the user's voice (e.g., smart glasses or a smartphone)

[0333] 2. Dedicated software for preprocessing audio data (e.g., noise reduction, volume adjustment)

[0334] 3. A data communication module for transmitting the preprocessed audio data to a server via the Internet.

[0335] 4. AI voice conversion model for converting voice data into another voice on the server

[0336] 5. Emotion engine for analyzing and recognizing user emotions on the server

[0337] 6. An audio playback library to encode the converted audio data (e.g., OGG or MP3 format)

[0338] 7. Speakers or headphones to play the received encoded audio on your device.

[0339] 2. Program Processing

[0340] When a user speaks into the microphone, the device captures the voice and preprocesses it using dedicated software. This preprocessing includes noise reduction and volume adjustment. A data communications module is used to send the processed voice data to a server. The sent voice data is then input into an AI voice conversion model on the server and converted into a different voice.

[0341] Next, the emotion engine on the server analyzes the voice data and recognizes the user's emotions. Based on this recognition result, the AI ​​voice conversion model is re-adjusted and converted into voice that reflects the emotion. The converted voice data is sent back from the server to the device, where it is encoded by the voice playback library and played back through speakers or headphones.

[0342] 3. Example Use Case Scenarios

[0343] Below is a usage scenario for a physical store.

[0344] Scenario 1: The customer is angry

[0345] A salesperson wears smart glasses and interacts with a customer. If the customer is angry, the salesperson responds by inputting a prompt sentence as follows:

[0346] Prompt statement:

[0347] "When a customer is angry, change the salesperson's voice to a calmer one and speak in a more soothing manner."

[0348] Scenario 2: Happy Customers

[0349] In the situation where the customer is happy, the salesperson enters the following prompt sentence:

[0350] Prompt statement:

[0351] "When a customer is happy, the salesperson should change their voice to a more cheerful and upbeat tone."

[0352] By implementing the present invention in this way, real-time voice conversion that reflects customer emotions becomes possible in physical stores, thereby improving customer satisfaction.

[0353] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0354] Step 1:

[0355] A user speaks into the microphone of a smart device (e.g., smart glasses or a smartphone). This voice input is captured as digital data. The input is the user's speech, and the output is digital voice data. Specifically, the microphone converts sound waves into electrical signals, which are then digitized and captured as voice data.

[0356] Step 2:

[0357] The device performs preprocessing on the acquired voice data. Preprocessing includes noise reduction (removing noise from the voice data) and volume adjustment (adjusting the volume of the voice to an appropriate level). The input is digital voice data, and the output is preprocessed voice data. Specifically, the noise reduction algorithm removes unnecessary components from the voice data, improving the sound quality.

[0358] Step 3:

[0359] The device sends the preprocessed audio data to the server via the Internet. This is done using a data communication module. The input is the preprocessed audio data, and the output is the audio data received by the server. Specifically, the audio data is sent to the server using a network protocol (e.g., HTTP or WebSocket).

[0360] Step 4:

[0361] The server inputs the received voice data into an AI voice conversion model and converts it into another voice. The input is the voice data received on the server side, and the output is the converted voice data. Specifically, the generative AI model extracts voice features and converts them into the specified target voice.

[0362] Step 5:

[0363] The server runs an emotion engine that analyzes the voice data and recognizes the user's emotions. The input is the converted voice data, and the output is the user's emotional state. Specifically, the emotion recognition algorithm analyzes changes in voice tone and pitch to determine emotions.

[0364] Step 6:

[0365] The server readjusts the AI ​​voice conversion model based on the results of the emotion engine. The input is the user's emotional state, and the output is voice data readjusted to reflect the emotion. Specifically, the AI ​​model processes the voice data again and adds appropriate voice tone and emphasis.

[0366] Step 7:

[0367] The server encodes the converted audio data and converts it into a playable format (e.g., OGG or MP3 format). The input is the reconditioned audio data, and the output is the encoded audio data. Specifically, the encoding library converts the audio data into a playable format.

[0368] Step 8:

[0369] The terminal receives encoded voice data from the server. The input is the encoded voice data, and the output is the voice data received on the terminal side. Specifically, the received data is obtained from the network and stored in memory.

[0370] Step 9:

[0371] The device decodes the received audio data and plays it on speakers or headphones. The input is the encoded audio data, and the output is the audio to be played. Specifically, the device uses a decoding library to convert the audio data into a digital audio signal and plays it on the audio output device.

[0372] By going through this series of steps, the user's voice is converted in real time into a different person's voice that reflects their emotions, and can be used in situations such as physical stores.

[0373] 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.

[0374] 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.

[0375] 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.

[0376] [Second embodiment]

[0377] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0378] 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.

[0379] 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).

[0380] 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.

[0381] 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.

[0382] 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).

[0383] 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.

[0384] 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.

[0385] 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.

[0386] 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.

[0387] 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.

[0388] 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."

[0389] The present invention provides a system for converting a user's voice into a voice of another person in real time. The system includes the following means.

[0390] First, we will explain how to obtain voice input: the user speaks into the device's microphone, and this voice input is captured by dedicated software within the device.

[0391] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0392] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0393] Next, we will explain how the server converts voice data. The server processes the received voice data based on an AI voice conversion model and converts it into another specified voice. This voice conversion uses a highly accurate deep learning model.

[0394] Next, we will explain how to receive the converted voice data from the server. The server re-encodes the converted voice data and sends it to the terminal. The terminal receives this encoded voice data.

[0395] Then, we explain how to encode the audio data. The terminal decodes the received audio data and converts it into a playable format. The encoded audio data is then processed so that it can be used by the user in a specific format.

[0396] Finally, we explain how to play the encoded audio data. The device plays the audio data encoded in a playable format through speakers or headphones. This allows the user to listen to the new audio converted from their own voice in real time.

[0397] Illustrative Usage Scenarios

[0398] For example, when user A says "hello," the audio is captured by the device's microphone. The device removes noise from the audio and sends it to the server. The server processes the received audio using an AI voice conversion model and converts it into user B's voice. This converted audio data is then sent back from the server to the device, where it is played back. This series of processes is performed in real time, so user A can hear "hello" in user B's voice without any delay.

[0399] As described above, the present invention effectively executes all processes from acquiring voice input to converting and playing, thereby providing users with an advanced real-time voice conversion experience.

[0400] The processing flow will be explained below.

[0401] Step 1:

[0402] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0403] Step 2:

[0404] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0405] Step 3:

[0406] The device performs preprocessing on the acquired audio data, such as noise reduction and volume normalization, which includes filtering and volume adjustment using an audio signal processing library.

[0407] Step 4:

[0408] The terminal packages the pre-processed audio data into a suitable format (e.g., binary format) for transmission to the server, which transmits the packaged data over the Internet to the server.

[0409] Step 5:

[0410] The server analyzes the received voice data and inputs it into an AI voice conversion model, which is based on a deep learning algorithm to extract voice features and convert them into a specified target voice.

[0411] Step 6:

[0412] The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0413] Step 7:

[0414] The terminal receives the encoded voice data sent from the server, and this reception process uses a common data communication protocol.

[0415] Step 8:

[0416] The device decodes the received audio data and converts it into a playable format. This decoding process is performed using an audio playback library.

[0417] Step 9:

[0418] The device then plays the decoded audio data over a speaker or headphones, allowing the user to hear the converted voice of another person in real time.

[0419] Example 1

[0420] 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."

[0421] Conventional speech conversion systems often suffer from degradation of speech quality and conversion delays during the process from speech capture to final speech playback. Furthermore, noise removal accuracy is low, which can result in a loss of naturalness in the converted speech. Furthermore, it is difficult to utilize highly accurate models in the conversion process. Therefore, converting speech in real time while providing high-quality speech remains a challenge.

[0422] 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.

[0423] In this invention, the server includes means for acquiring a user's voice, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server via a data communication network, means for converting the transmitted voice data into another voice using a deep learning model, means for receiving the converted voice data from the server, means for decoding the received voice data into a playable format, and means for playing the decoded voice on an audio output device, thereby enabling voice conversion in real time while maintaining the quality of the voice.

[0424] "User" refers to the person who uses the system and is responsible for inputting voice and listening to the final converted voice.

[0425] "Speech" refers to the sound produced by a person, which is picked up by a microphone and processed within the system.

[0426] "Preprocessing" refers to processing that removes noise and adjusts the volume of acquired audio data to improve the quality of the data.

[0427] "Data communications network" means a digital communications infrastructure, such as the Internet, used to transmit voice data from a terminal to a server or from a server to a terminal.

[0428] The "server" is a computer system that handles the conversion of voice data and performs voice conversion using a deep learning model.

[0429] A "deep learning model" is a type of artificial intelligence that uses neural networks to perform advanced conversion and analysis of voice data.

[0430] "Decoding" is the process of converting encoded data back into its original, playable format.

[0431] "Audio output device" refers to a device, such as a speaker or headphones, used to play decoded audio to a user.

[0432] "Noise reduction" is the process of removing unwanted noise from audio data, improving the clarity of the audio.

[0433] "Playable format" means that audio data has been converted into an appropriate format so that the user can listen to the audio.

[0434] This invention is a system that converts user-spoken speech into another speech in real time. This system effectively executes a series of processes from speech input to speech output, focusing on three elements: the user, the terminal, and the server. The specific configuration and operation are described below.

[0435] First, the user speaks into the device's microphone. This speech is captured in real time using dedicated software such as Audacity. The captured speech data is then pre-processed on the device. This pre-processing includes noise reduction and volume adjustment to improve the quality of the speech data. Audacity's "noise reduction" and "volume normalization" functions are used in this process.

[0436] The preprocessed audio data is then sent over a data communications network to a server, typically using a protocol such as an HTTP POST request. The server processes the received audio data using a deep learning model, such as Google Cloud Text-to-Speech or OpenAI's Whisper model, to convert it into a different sound.

[0437] The converted audio data is then encoded on the server and sent back to the device via the data communications network. For example, ffmpeg is used to encode the audio data into an audio format such as FLAC or AAC. The device then decodes the received audio data and converts it into a playable format. Once decoded, the audio data is played on the device's audio output device (speakers or headphones) using the device's VLC media player or the system's default audio playback function.

[0438] For example, when user A says "hello," the audio is captured through the device's microphone. Audacity software captures the audio in real time and performs noise reduction and volume normalization. The processed audio data is sent to the server using an HTTP POST request. The server uses the Google Cloud Text-to-Speech API to convert the audio into user B's voice and sends the converted audio back to the device. On the device, the FLAC format audio data is decoded into MP3 format using ffmpeg and played back using VLC media player. In this way, user A can hear "hello" in user B's voice, converted in real time.

[0439] As described above, the present invention executes the entire process from acquiring voice input to conversion and playback with high quality and in real time, providing users with an advanced voice conversion experience.

[0440] Prompt Sentence Examples

[0441] "Please explain the specific steps involved in converting a user's voice into a different person's voice in real time. Please describe in detail the process from voice capture to conversion and playback, and indicate the software and hardware used in the process."

[0442] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0443] Step 1: Getting voice input

[0444] The user speaks into the microphone of the terminal, for example, saying "hello."

[0445] The device captures this audio using specialized software (e.g., Audacity).

[0446] Input: User's live voice

[0447] Data processing: Audio is captured as digital data using dedicated software

[0448] Output: Unprocessed digital audio data

[0449] Step 2: Preprocessing

[0450] The terminal performs noise reduction and volume adjustment on the captured digital audio data.

[0451] Use Audacity's "Noise Reduction" and "Volume Normalization" features.

[0452] Input: Unprocessed digital audio data

[0453] Data processing: noise removal and volume adjustment

[0454] Output: Pre-processed digital audio data

[0455] Step 3: Sending audio data to the server

[0456] The terminal transmits the pre-processed voice data to a server via a data communication network (eg, the Internet).

[0457] Use an HTTP POST request.

[0458] Input: Preprocessed digital audio data

[0459] Data processing: Data communication via HTTP POST request

[0460] Output: Audio data sent to the server

[0461] Step 4: Audio conversion by server

[0462] The server converts the transmitted voice data into a different voice using a deep learning model (e.g., Google Cloud Text-to-Speech or OpenAI's Whisper model).

[0463] Input: Audio data sent to the server

[0464] Data Computing: Voice Conversion with Deep Learning Models

[0465] Output: Converted audio data

[0466] Step 5: Translated audio data sent from the server

[0467] The server encodes the converted voice data and transmits it to the terminal via a data communication network.

[0468] For encoding, ffmpeg is used to apply audio formats such as FLAC and AAC.

[0469] Input: Converted audio data

[0470] Data processing: Encoding into audio format

[0471] Output: Encoded audio data sent to the device

[0472] Step 6: Decode the audio data

[0473] The terminal decodes the audio data received from the server into a reproducible format.

[0474] Use ffmpeg to decode to MP3 or other formats.

[0475] Input: Encoded audio data received from the server

[0476] Data processing: Decoding into a playable format

[0477] Output: Decoded audio data

[0478] Step 7: Playing back audio data

[0479] The terminal plays the decoded audio data on an audio output device (e.g., a speaker or headphones).

[0480] Use VLC media player or the system's default audio playback function for playback.

[0481] Input: Decoded audio data

[0482] Data calculation: Audio data playback

[0483] Output: Playback of the converted audio from an audio output device

[0484] This allows the user to listen to the converted audio in real time.

[0485] (Application example 1)

[0486] 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."

[0487] Conventional voice conversion systems have problems when converting a user's voice into a different person's voice in real time, such as difficulty in achieving high-quality conversion, insufficient countermeasures against noise contamination, and difficulty in instantly adding a wide variety of voice effects depending on the distribution. In particular, for distribution services of entertainment and educational content, there is a demand for real-time conversion into a different voice, as well as support for live distribution and on-demand effects, in order to improve the user experience.

[0488] 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.

[0489] In this invention, the server includes means for acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for converting a user's voice into another voice in real time, means for live-streaming the converted voice data, and means for adding on-demand voice effects, thereby enabling real-time, high-quality voice conversion, the addition of a variety of effects, and the provision of attractive content in a distribution service.

[0490] The "means for acquiring voice input" is a function for acquiring the voice uttered by the user as digital data using a device such as a microphone.

[0491] The "means for preprocessing the acquired audio" is a function for performing noise removal and volume adjustment on the acquired audio data to improve the quality.

[0492] The "means for transmitting preprocessed voice data to a server" is a function for transmitting preprocessed voice data to a server via the Internet or other communication means.

[0493] "Means for converting transmitted voice data into another voice" refers to a function that converts voice data received on a server into the voice of another person using an AI model, etc.

[0494] The "means for receiving converted voice data from the server" is a function for transmitting the converted voice data from the server again to the client terminal and receiving it.

[0495] The "means for encoding received audio data" is a function that performs processing to convert audio data received from the server into a format that can be used by the user.

[0496] "Means for playing encoded audio" refers to a function for playing encoded audio data through speakers or headphones.

[0497] "Means for converting a user's voice into a different voice in real time" refers to technology and devices for converting a user's voice into a different voice as the user speaks.

[0498] The "means for live distribution of converted voice data" is a function for distributing converted voice data in real time via the Internet.

[0499] The "means for adding on-demand audio effects" is a function for adding new audio effects to existing audio data to improve its entertainment value and educational effectiveness.

[0500] The present invention is a system for converting a user's voice into another person's voice in real time. The implementation of the present invention includes the following components and procedures.

[0501] System Configuration

[0502] 1. Hardware:

[0503] Device: A smartphone or other audio input device that has a microphone and speaker.

[0504] Server: A high-performance computing device for processing voice data online.

[0505] 2. Software:

[0506] WebRTC: A library for capturing, transmitting, and receiving voice data in real time.

[0507] TensorFlow: A machine learning library for building and using AI voice models.

[0508] Flask: A web framework for processing audio data on the server side.

[0509] Processing flow

[0510] 1. Audio capture:

[0511] The user speaks into the smartphone's microphone, and the device captures the voice input and stores it as digital data using data capture software (e.g., WebRTC).

[0512] 2. Pretreatment:

[0513] The captured audio data is pre-processed with noise removal, volume adjustment, etc. This improves the quality of the audio data and increases the accuracy of the subsequent conversion process.

[0514] 3. Data transmission:

[0515] The pre-processed audio data is then transmitted from the device to a server over the internet in real time using WebRTC.

[0516] 4. Audio conversion:

[0517] The server converts the received voice data into another voice using an AI voice conversion model (using TensorFlow). Deep learning models are applied in this step to achieve high-precision voice conversion.

[0518] 5. Data Reception:

[0519] The converted voice data is sent from the server back to the terminal and received in real time.

[0520] 6. Encoding and Playback:

[0521] The device then encodes the received audio data into a playable format and plays it through speakers or headphones, allowing the user to hear the converted version of their own voice.

[0522] Specific examples

[0523] For example, a user may want to change their voice to sound like a historical figure during an online educational lecture. When the user says "hello," the voice is captured by the device's microphone and preprocessed to remove noise. The preprocessed voice data is then sent over the Internet to a server, where an AI model on the server converts the voice data into the voice of the specified person. The converted voice data is then sent back to the device and played back in real time on the device.

[0524] Prompt Sentence Examples

[0525] Please upload an audio file from the user:

[0526]

[0527] Choose an AI model to transform your voice:

[0528] In this way, the present invention provides users with a high quality real-time voice conversion experience, which is very useful in the fields of entertainment and education.

[0529] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0530] Step 1:

[0531] The user speaks into the smartphone's microphone. The device receives this voice input and captures it as digital data. Here, the input data is the user's voice data, and the output data is the digitized voice data. In this step, the microphone is used to convert the analog voice signal into digital data, which is then saved by the voice capture software.

[0532] Step 2:

[0533] The device performs preprocessing on the captured audio data. This involves processes to improve the quality of the audio data, such as noise removal and volume adjustment. Here, the input is the digital audio data captured in step 1, and the output is the preprocessed audio data. Specifically, a library such as webrtcvad is used to remove noise and apply a volume normalization algorithm.

[0534] Step 3:

[0535] The preprocessed voice data is transmitted to a server via the Internet. The terminal divides the digital data into packets and transmits them using a communication module, where the input is the preprocessed voice data and the output is the data packets transmitted via network communication.

[0536] Step 4:

[0537] The server converts the received voice data into a different voice using an AI voice conversion model (using a generative AI model). The server uses the received data as input, applies the generative AI model, and outputs voice data of a different person. Specifically, it executes a deep learning model using the TensorFlow library and performs voice conversion based on the specified prompt sentence.

[0538] Step 5:

[0539] The converted voice data is sent back to the terminal from the server, where it is re-encoded into packets and sent to the terminal via the Internet, where the input is the converted voice data and the output is the data packets received by the terminal via network communication.

[0540] Step 6:

[0541] The terminal decodes the received converted audio data into a playable format. The decoding process involves converting the audio data using signal processing algorithms and converting it into a file format for playback through speakers or headphones. Here, the input is the data packets received from the server, and the output is playable audio data.

[0542] Step 7:

[0543] The device plays the decoded audio data over speakers or headphones, allowing the user to listen to another person's voice in real time. Here, the input is the decoded audio data, and the output is the physical playback of audio. Specifically, audio playback software converts the digital audio signal to an analog signal and plays it on an audio output device.

[0544] 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.

[0545] The present invention provides a system for converting a user's voice into a different person's voice in real time, and further recognizing the user's emotions and reflecting them in the voice conversion. The system includes the following means.

[0546] First, we will explain how to obtain voice input. A user speaks into the microphone of the device. For example, the user says "hello." This voice input is captured by dedicated software in the device.

[0547] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0548] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0549] Next, we will explain how the server converts voice data. The server inputs the received voice data into an AI voice conversion model. This model is based on a deep learning algorithm, extracts voice features, and converts them into a specified target voice.

[0550] Next, we will explain the emotion engine. The server analyzes the voice data provided by the terminal and recognizes the user's emotional state. For example, if the user is angry or happy, the emotion engine determines that emotion.

[0551] Furthermore, we describe a method for adjusting the voice conversion model based on the recognized emotions. The server uses feedback from the emotion engine to readjust the voice conversion model to match the user's emotions. This adjustment allows for voice conversion that is faithful to the user's emotions.

[0552] This section explains how to receive converted audio data from the server. The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0553] This section then describes how the terminal decodes and plays back the audio data. The terminal decodes the encoded audio data it receives and converts it into a playable format. This decoding process is performed using an audio playback library.

[0554] Finally, we will explain how to play back the encoded voice data. The terminal plays back the decoded voice data through a speaker or headphones. This allows the user to hear in real time the voice of another person that has been converted from their own voice. This voice reflects the user's emotions.

[0555] Illustrative Usage Scenarios

[0556] For example, if User A says "Hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model and uses an emotion engine to recognize User A's emotion of "anger." The voice conversion model is then adjusted based on this emotion and converted into User B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear User B's angry voice saying "Hello" in real time, without any delay.

[0557] As described above, the present invention effectively executes all processes from voice input acquisition to emotion recognition, conversion, and playback, thereby providing users with an advanced real-time voice conversion experience.

[0558] The processing flow will be explained below.

[0559] Step 1:

[0560] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0561] Step 2:

[0562] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0563] Step 3:

[0564] The device performs preprocessing on the acquired voice data, such as noise reduction and volume adjustment, to remove excess noise to improve the clarity of the voice and to normalize the voice volume to keep it constant.

[0565] Step 4:

[0566] The device packages the preprocessed audio data into an appropriate format (e.g., binary format) for transmission to the server. The packaged data is sent to the server via the Internet. HTTP or WebSocket is often used as the communication protocol.

[0567] Step 5:

[0568] The server analyzes the received voice data and inputs it into an emotion engine that recognizes the user's emotions. The emotion engine identifies the user's emotional state (e.g., joy, anger, sadness, etc.) based on the tone, rhythm, and intonation of the voice.

[0569] Step 6:

[0570] The server acquires the emotion information recognized by the emotion engine and adjusts the AI ​​voice conversion model based on that information. For example, if the user is angry, the AI ​​model will set the voice conversion parameters to reflect the angry emotion.

[0571] Step 7:

[0572] The server uses a trained AI voice conversion model to convert the pre-processed voice data into a different voice, which faithfully matches the target voice based on the user's voice characteristics and emotional information.

[0573] Step 8:

[0574] The server then encodes the converted audio data into a playable format (such as OGG or MP3) and sends it to the device. The encoded data is compressed for efficient transmission.

[0575] Step 9:

[0576] The terminal receives the encoded voice data sent from the server, and the receiving process follows a pre-established communication protocol.

[0577] Step 10:

[0578] The device then decodes the received encoded audio data and converts it into a playable format. This decoding process is performed using an audio playback library, which removes the compression applied during encoding and restores the original audio format.

[0579] Step 11:

[0580] The device then plays the decoded voice data over a speaker or headphones, allowing the user to listen to a voice that faithfully reflects the user's emotional information, providing a more natural communication experience.

[0581] Example 2

[0582] 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."

[0583] Conventional speech conversion systems simply convert speech into another speech without taking into account the user's emotions, resulting in unnatural speech and poor emotional expression. Real-time speech conversion also entails delays, which can negatively impact the user experience. It is necessary to solve these problems and achieve more natural and emotionally rich speech conversion.

[0584] 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.

[0585] In this invention, the server includes means for converting transmitted voice data into a specified target voice, means for analyzing the voice data to recognize the user's emotional state, and means for adjusting the conversion model based on the recognized emotional state, thereby enabling natural voice conversion that reflects the user's emotions.

[0586] "User" refers to a person who uses the system to provide voice input.

[0587] "Terminal" refers to a device that acquires audio, performs preprocessing, and then transmits and receives the audio data to a server.

[0588] "Voice input" refers to voice data uttered by a user through a microphone of a terminal.

[0589] "Preprocessing" refers to the act of performing processes such as noise removal and volume adjustment on audio data to improve data quality.

[0590] "Server" refers to a device that receives voice data sent from a terminal and converts it into the voice of a specified target.

[0591] A "voice conversion model" refers to an algorithm that uses AI technology to convert voice data into another voice.

[0592] "Emotion engine" refers to an algorithm that analyzes and recognizes the user's emotional state from voice data.

[0593] "Adjusting the conversion model" refers to the act of adjusting the voice conversion model based on the analysis results of the emotion engine, and performing conversion that reflects emotions.

[0594] "Encoding" refers to the act of converting audio data into a playable format.

[0595] "Decoding" refers to the act of restoring encoded audio data to a playable format.

[0596] "Playback" refers to the act of allowing a user to listen to audio data through the terminal's speakers or headphones.

[0597] This invention is a system that converts a user's voice into another person's voice in real time, recognizes the user's emotions, and reflects them in the voice conversion. The following hardware and software are used to implement this system.

[0598] First, a device needs a microphone as hardware to acquire voice input. A user inputs voice by speaking into the microphone. For example, a user might say "hello." This voice input is captured by a voice capture application in the device.

[0599] Next, the device preprocesses the captured audio data by noise reduction and volume adjustment using a noise reduction software library (e.g., Audacity). This preprocessing improves the quality of the audio data and increases conversion accuracy.

[0600] The pre-processed voice data is sent to a server over the Internet, usually using the HTTP protocol. The server then inputs the received voice data into an AI voice conversion model (e.g., WaveNet) to convert it into a specified target voice. This model is based on a deep learning algorithm and analyzes the voice characteristics to perform the conversion.

[0601] The server is also equipped with an emotion engine that identifies the user's emotions from the received voice data. Software such as IBM Tone Analyzer is used for emotion recognition. The recognized emotions (e.g., anger, joy, etc.) are fed back to the voice conversion model, which is then adjusted according to the user's emotions.

[0602] The adjusted audio data is then sent from the server to the device. The server encodes the audio data into a playable format (e.g., OGG or MP3) and sends it to the device. The device then decodes the encoded audio data and converts it into a playable format using a playback library (e.g., FFmpeg).

[0603] Finally, the device plays the decoded voice data over a speaker or headphones, allowing the user to hear their own voice transformed into a different person's voice in real time, with the voice reflecting the user's emotions.

[0604] Specific examples

[0605] For example, if user A says "hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model (WaveNet) and uses an emotion engine (IBM Tone Analyzer) to recognize user A's emotion of "anger." The voice conversion model is then adjusted based on this emotion, and the voice is converted into user B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear user B's angry voice saying "hello" in real time, without any delay.

[0606] In this way, the system of the present invention effectively executes a series of processes from voice input acquisition to emotion recognition, conversion, and playback, providing users with an advanced real-time voice conversion experience.

[0607] Example prompt sentences to use

[0608] Example prompt sentence:

[0609] User A says "Hello" in an angry voice. The audio is captured, noise is removed, and then input into the speech conversion model. During conversion, the emotion recognition engine detects the emotion "anger" and converts it into the angry voice of User B. Play the converted audio.

[0610] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0611] Step 1:

[0612] Acquiring voice input

[0613] Input: The user's voice speaking into the device's microphone (e.g., "Hello")

[0614] Specific operation: The user speaks into the microphone of the terminal.

[0615] Data processing: The device uses an audio capture application to capture this audio as digital audio data.

[0616] Output: Captured audio data (e.g., WAV file with a sampling rate of 44.1 kHz)

[0617] Step 2:

[0618] Audio data preprocessing

[0619] Input: Captured audio data

[0620] Specific operation: The device uses a noise reduction software library (e.g., Audacity) to remove noise from the audio data and adjust the volume.

[0621] Data processing: Apply noise filtering algorithms to improve audio clarity.

[0622] Output: Preprocessed audio data (e.g., a denoised WAV file)

[0623] Step 3:

[0624] Sending audio data to the server

[0625] Input: Preprocessed audio data

[0626] Specific operation: The terminal sends audio data to the server using the HTTP protocol.

[0627] Data processing: The audio data is sent in the body of the HTTP request.

[0628] Output: Audio data received by the server

[0629] Step 4:

[0630] Processing with voice conversion models

[0631] Input: Audio data sent to the server

[0632] Specific operation: The server inputs voice data into an AI voice conversion model (e.g., WaveNet).

[0633] Data Computing: Using deep learning algorithms to extract voice features and convert them into the target voice.

[0634] Output: Converted voice data (e.g., a WAV file converted into a different person's voice)

[0635] Step 5:

[0636] Emotion determination using an emotion recognition engine

[0637] Input: Audio data sent to the server

[0638] Specific operation: The server inputs voice data into an emotion recognition engine (e.g., IBM Tone Analyzer).

[0639] Data processing: Analyze the voice data and extract emotional features.

[0640] Output: The user's emotional state (e.g., "anger" or "joy")

[0641] Step 6:

[0642] Tuning the speech conversion model

[0643] Input: Feedback from the emotion recognition engine (user's emotional state)

[0644] Specific operation: The server adjusts the voice conversion model based on the results of the emotion recognition engine.

[0645] Data calculation: Taking into account emotion features, the parameters of the conversion model are changed.

[0646] Output: A voice conversion model that reflects the user's emotions

[0647] Step 7:

[0648] Encoding the converted audio data

[0649] Input: Speech data converted by the trained speech conversion model

[0650] What happens: The server encodes the audio data into a playable format (e.g., OGG or MP3).

[0651] Data processing: Converting file formats using audio encoding algorithms.

[0652] Output: Encoded audio data (e.g. MP3 file)

[0653] Step 8:

[0654] Sending encoded audio data to the device

[0655] Input: Encoded audio data

[0656] Specific operation: The server uses the HTTP protocol to send encoded audio data to the terminal.

[0657] Data processing: The audio data is sent in the body of the HTTP response.

[0658] Output: The encoded audio data received by the device.

[0659] Step 9:

[0660] Decoding audio data

[0661] Input: Encoded audio data

[0662] Specific behavior: The device uses an audio playback library (e.g., FFmpeg) to decode the encoded audio data.

[0663] Data processing: Converting from an encoded format to a playable format.

[0664] Output: Decoded audio data (e.g. WAV file)

[0665] Step 10:

[0666] Playing audio data

[0667] Input: Decoded audio data

[0668] Specific operation: The device plays audio data using a playback device (speaker or headphones).

[0669] Data calculation: The decoded audio is converted from digital to analog and output.

[0670] Output: The audio the user hears (a modified voice, e.g., "Hello" with the emotion "Anger")

[0671] (Application example 2)

[0672] 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."

[0673] In conventional brick-and-mortar stores, it has been difficult to immediately grasp a customer's emotions and respond appropriately. Furthermore, if a customer is angry or happy, it is not possible to communicate with them in a voice that reflects their emotions, which can lead to a decline in the quality of customer service. By solving these problems, it is necessary to provide more effective customer service and improve customer satisfaction.

[0674] 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 acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for recognizing emotions and reflecting them in voice conversion, and means for adjusting voice tone based on the customer's emotions using an application installed on the smart device. This enables real-time voice conversion that reflects customer emotions in physical stores.

[0675] The "means for acquiring voice input" refers to a device or method for acquiring the voice spoken by the user as digital data using an acoustic sensor such as a microphone.

[0676] "Means for preprocessing captured audio" refers to a device or method that performs preprocessing such as noise removal and volume adjustment on captured audio data to improve the quality of the data and increase the accuracy of subsequent processing.

[0677] The "means for transmitting preprocessed voice data to a server" refers to a device or method for transmitting preprocessed voice data to a server via a communication line such as the Internet.

[0678] "Means for converting transmitted voice data into another voice" refers to a device or method for converting received voice data into another voice using an AI voice model or other conversion algorithm.

[0679] The "means for receiving converted voice data from the server" refers to a device or method for the server to send the converted voice data back to the client device via a communication line such as the Internet.

[0680] "Means for encoding received audio data" refers to a device or method for converting received audio data into a playable format (e.g., OGG or MP3 format).

[0681] "Means for playing encoded audio" refers to an apparatus or method for playing encoded audio data using a playback device (speakers or headphones).

[0682] "Means for recognizing emotions and reflecting them in voice conversion" refers to a device or method that analyzes the emotions from the user's voice, adjusts the voice conversion model based on the recognition results, and converts the voice into a different voice that reflects the emotions.

[0683] "Means for adjusting voice tone based on customer emotions using an application installed on a smart device" refers to a device or method that uses software installed on a device such as smart glasses or a smartphone to adjust voice tone in real time during customer service based on the customer's emotions.

[0684] The present invention provides a system that converts a user's voice into a different person's voice in real time, and further recognizes the user's emotions and reflects them in the voice conversion. Specific embodiments of the system are described below.

[0685] 1. System Configuration

[0686] This system is primarily composed of smart devices, internet communications, and servers. It includes the following elements:

[0687] 1. A device with a microphone to capture the user's voice (e.g., smart glasses or a smartphone)

[0688] 2. Dedicated software for preprocessing audio data (e.g., noise reduction, volume adjustment)

[0689] 3. A data communication module for transmitting the preprocessed audio data to a server via the Internet.

[0690] 4. AI voice conversion model for converting voice data into another voice on the server

[0691] 5. Emotion engine for analyzing and recognizing user emotions on the server

[0692] 6. An audio playback library to encode the converted audio data (e.g., OGG or MP3 format)

[0693] 7. Speakers or headphones to play the received encoded audio on your device.

[0694] 2. Program Processing

[0695] When a user speaks into the microphone, the device captures the voice and preprocesses it using dedicated software. This preprocessing includes noise reduction and volume adjustment. A data communications module is used to send the processed voice data to a server. The sent voice data is then input into an AI voice conversion model on the server and converted into a different voice.

[0696] Next, the emotion engine on the server analyzes the voice data and recognizes the user's emotions. Based on this recognition result, the AI ​​voice conversion model is re-adjusted and converted into voice that reflects the emotion. The converted voice data is sent back from the server to the device, where it is encoded by the voice playback library and played back through speakers or headphones.

[0697] 3. Example Use Case Scenarios

[0698] Below is a usage scenario for a physical store.

[0699] Scenario 1: The customer is angry

[0700] A salesperson wears smart glasses and interacts with a customer. If the customer is angry, the salesperson responds by inputting a prompt sentence as follows:

[0701] Prompt statement:

[0702] "When a customer is angry, change the salesperson's voice to a calmer one and speak in a more soothing manner."

[0703] Scenario 2: Happy Customers

[0704] In the situation where the customer is happy, the salesperson enters the following prompt sentence:

[0705] Prompt statement:

[0706] "When a customer is happy, the salesperson should change their voice to a more cheerful and upbeat tone."

[0707] By implementing the present invention in this way, real-time voice conversion that reflects customer emotions becomes possible in physical stores, thereby improving customer satisfaction.

[0708] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0709] Step 1:

[0710] A user speaks into the microphone of a smart device (e.g., smart glasses or a smartphone). This voice input is captured as digital data. The input is the user's speech, and the output is digital voice data. Specifically, the microphone converts sound waves into electrical signals, which are then digitized and captured as voice data.

[0711] Step 2:

[0712] The device performs preprocessing on the acquired voice data. Preprocessing includes noise reduction (removing noise from the voice data) and volume adjustment (adjusting the volume of the voice to an appropriate level). The input is digital voice data, and the output is preprocessed voice data. Specifically, the noise reduction algorithm removes unnecessary components from the voice data, improving the sound quality.

[0713] Step 3:

[0714] The device sends the preprocessed audio data to the server via the Internet. This is done using a data communication module. The input is the preprocessed audio data, and the output is the audio data received by the server. Specifically, the audio data is sent to the server using a network protocol (e.g., HTTP or WebSocket).

[0715] Step 4:

[0716] The server inputs the received voice data into an AI voice conversion model and converts it into another voice. The input is the voice data received on the server side, and the output is the converted voice data. Specifically, the generative AI model extracts voice features and converts them into the specified target voice.

[0717] Step 5:

[0718] The server runs an emotion engine that analyzes the voice data and recognizes the user's emotions. The input is the converted voice data, and the output is the user's emotional state. Specifically, the emotion recognition algorithm analyzes changes in voice tone and pitch to determine emotions.

[0719] Step 6:

[0720] The server readjusts the AI ​​voice conversion model based on the results of the emotion engine. The input is the user's emotional state, and the output is voice data readjusted to reflect the emotion. Specifically, the AI ​​model processes the voice data again and adds appropriate voice tone and emphasis.

[0721] Step 7:

[0722] The server encodes the converted audio data and converts it into a playable format (e.g., OGG or MP3 format). The input is the reconditioned audio data, and the output is the encoded audio data. Specifically, the encoding library converts the audio data into a playable format.

[0723] Step 8:

[0724] The terminal receives encoded voice data from the server. The input is the encoded voice data, and the output is the voice data received on the terminal side. Specifically, the received data is obtained from the network and stored in memory.

[0725] Step 9:

[0726] The device decodes the received audio data and plays it on speakers or headphones. The input is the encoded audio data, and the output is the audio to be played. Specifically, the device uses a decoding library to convert the audio data into a digital audio signal and plays it on the audio output device.

[0727] By going through this series of steps, the user's voice is converted in real time into a different person's voice that reflects their emotions, and can be used in situations such as physical stores.

[0728] 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.

[0729] 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.

[0730] 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.

[0731] [Third embodiment]

[0732] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0733] 5, the data processing system 310 includes the data processing device 12 and a headset terminal 314. An example of the data processing device 12 is a server.

[0734] 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).

[0735] 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.

[0736] 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.

[0737] 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).

[0738] 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.

[0739] 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.

[0740] 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.

[0741] 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.

[0742] 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.

[0743] 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."

[0744] The present invention provides a system for converting a user's voice into a voice of another person in real time. The system includes the following means.

[0745] First, we will explain how to obtain voice input: the user speaks into the device's microphone, and this voice input is captured by dedicated software within the device.

[0746] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0747] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0748] Next, we will explain how the server converts voice data. The server processes the received voice data based on an AI voice conversion model and converts it into another specified voice. This voice conversion uses a highly accurate deep learning model.

[0749] Next, we will explain how to receive the converted voice data from the server. The server re-encodes the converted voice data and sends it to the terminal. The terminal receives this encoded voice data.

[0750] Then, we explain how to encode the audio data. The terminal decodes the received audio data and converts it into a playable format. The encoded audio data is then processed so that it can be used by the user in a specific format.

[0751] Finally, we explain how to play the encoded audio data. The device plays the audio data encoded in a playable format through speakers or headphones. This allows the user to listen to the new audio converted from their own voice in real time.

[0752] Illustrative Usage Scenarios

[0753] For example, when user A says "hello," the audio is captured by the device's microphone. The device removes noise from the audio and sends it to the server. The server processes the received audio using an AI voice conversion model and converts it into user B's voice. This converted audio data is then sent back from the server to the device, where it is played back. This series of processes is performed in real time, so user A can hear "hello" in user B's voice without any delay.

[0754] As described above, the present invention effectively executes all processes from acquiring voice input to converting and playing, thereby providing users with an advanced real-time voice conversion experience.

[0755] The processing flow will be explained below.

[0756] Step 1:

[0757] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0758] Step 2:

[0759] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0760] Step 3:

[0761] The device performs preprocessing on the acquired audio data, such as noise reduction and volume normalization, which includes filtering and volume adjustment using an audio signal processing library.

[0762] Step 4:

[0763] The terminal packages the pre-processed audio data into a suitable format (e.g., binary format) for transmission to the server, which transmits the packaged data over the Internet to the server.

[0764] Step 5:

[0765] The server analyzes the received voice data and inputs it into an AI voice conversion model, which is based on a deep learning algorithm to extract voice features and convert them into a specified target voice.

[0766] Step 6:

[0767] The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0768] Step 7:

[0769] The terminal receives the encoded voice data sent from the server, and this reception process uses a common data communication protocol.

[0770] Step 8:

[0771] The device decodes the received audio data and converts it into a playable format. This decoding process is performed using an audio playback library.

[0772] Step 9:

[0773] The device then plays the decoded audio data over a speaker or headphones, allowing the user to hear the converted voice of another person in real time.

[0774] Example 1

[0775] 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."

[0776] Conventional speech conversion systems often suffer from degradation of speech quality and conversion delays during the process from speech capture to final speech playback. Furthermore, noise removal accuracy is low, which can result in a loss of naturalness in the converted speech. Furthermore, it is difficult to utilize highly accurate models in the conversion process. Therefore, converting speech in real time while providing high-quality speech remains a challenge.

[0777] 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.

[0778] In this invention, the server includes means for acquiring a user's voice, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server via a data communication network, means for converting the transmitted voice data into another voice using a deep learning model, means for receiving the converted voice data from the server, means for decoding the received voice data into a playable format, and means for playing the decoded voice on an audio output device, thereby enabling voice conversion in real time while maintaining the quality of the voice.

[0779] "User" refers to the person who uses the system and is responsible for inputting voice and listening to the final converted voice.

[0780] "Speech" refers to the sound produced by a person, which is picked up by a microphone and processed within the system.

[0781] "Preprocessing" refers to processing that removes noise and adjusts the volume of acquired audio data to improve the quality of the data.

[0782] "Data communications network" means a digital communications infrastructure, such as the Internet, used to transmit voice data from a terminal to a server or from a server to a terminal.

[0783] The "server" is a computer system that handles the conversion of voice data and performs voice conversion using a deep learning model.

[0784] A "deep learning model" is a type of artificial intelligence that uses neural networks to perform advanced conversion and analysis of voice data.

[0785] "Decoding" is the process of converting encoded data back into its original, playable format.

[0786] "Audio output device" refers to a device, such as a speaker or headphones, used to play decoded audio to a user.

[0787] "Noise reduction" is the process of removing unwanted noise from audio data, improving the clarity of the audio.

[0788] "Playable format" means that audio data has been converted into an appropriate format so that the user can listen to the audio.

[0789] This invention is a system that converts user-spoken speech into another speech in real time. This system effectively executes a series of processes from speech input to speech output, focusing on three elements: the user, the terminal, and the server. The specific configuration and operation are described below.

[0790] First, the user speaks into the device's microphone. This speech is captured in real time using dedicated software such as Audacity. The captured speech data is then pre-processed on the device. This pre-processing includes noise reduction and volume adjustment to improve the quality of the speech data. Audacity's "noise reduction" and "volume normalization" functions are used in this process.

[0791] The preprocessed audio data is then sent over a data communications network to a server, typically using a protocol such as an HTTP POST request. The server processes the received audio data using a deep learning model, such as Google Cloud Text-to-Speech or OpenAI's Whisper model, to convert it into a different sound.

[0792] The converted audio data is then encoded on the server and sent back to the device via the data communications network. For example, ffmpeg is used to encode the audio data into an audio format such as FLAC or AAC. The device then decodes the received audio data and converts it into a playable format. Once decoded, the audio data is played on the device's audio output device (speakers or headphones) using the device's VLC media player or the system's default audio playback function.

[0793] For example, when user A says "hello," the audio is captured through the device's microphone. Audacity software captures the audio in real time and performs noise reduction and volume normalization. The processed audio data is sent to the server using an HTTP POST request. The server uses the Google Cloud Text-to-Speech API to convert the audio into user B's voice and sends the converted audio back to the device. On the device, the FLAC format audio data is decoded into MP3 format using ffmpeg and played back using VLC media player. In this way, user A can hear "hello" in user B's voice, converted in real time.

[0794] As described above, the present invention executes the entire process from acquiring voice input to conversion and playback with high quality and in real time, providing users with an advanced voice conversion experience.

[0795] Prompt Sentence Examples

[0796] "Please explain the specific steps involved in converting a user's voice into a different person's voice in real time. Please describe in detail the process from voice capture to conversion and playback, and indicate the software and hardware used in the process."

[0797] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0798] Step 1: Getting voice input

[0799] The user speaks into the microphone of the terminal, for example, saying "hello."

[0800] The device captures this audio using specialized software (e.g., Audacity).

[0801] Input: User's live voice

[0802] Data processing: Audio is captured as digital data using dedicated software

[0803] Output: Unprocessed digital audio data

[0804] Step 2: Preprocessing

[0805] The terminal performs noise reduction and volume adjustment on the captured digital audio data.

[0806] Use Audacity's "Noise Reduction" and "Volume Normalization" features.

[0807] Input: Unprocessed digital audio data

[0808] Data processing: noise removal and volume adjustment

[0809] Output: Pre-processed digital audio data

[0810] Step 3: Sending audio data to the server

[0811] The terminal transmits the pre-processed voice data to a server via a data communication network (eg, the Internet).

[0812] Use an HTTP POST request.

[0813] Input: Preprocessed digital audio data

[0814] Data processing: Data communication via HTTP POST request

[0815] Output: Audio data sent to the server

[0816] Step 4: Audio conversion by server

[0817] The server converts the transmitted voice data into a different voice using a deep learning model (e.g., Google Cloud Text-to-Speech or OpenAI's Whisper model).

[0818] Input: Audio data sent to the server

[0819] Data Computing: Voice Conversion with Deep Learning Models

[0820] Output: Converted audio data

[0821] Step 5: Translated audio data sent from the server

[0822] The server encodes the converted voice data and transmits it to the terminal via a data communication network.

[0823] For encoding, ffmpeg is used to apply audio formats such as FLAC and AAC.

[0824] Input: Converted audio data

[0825] Data processing: Encoding into audio format

[0826] Output: Encoded audio data sent to the device

[0827] Step 6: Decode the audio data

[0828] The terminal decodes the audio data received from the server into a reproducible format.

[0829] Use ffmpeg to decode to MP3 or other formats.

[0830] Input: Encoded audio data received from the server

[0831] Data processing: Decoding into a playable format

[0832] Output: Decoded audio data

[0833] Step 7: Playing back audio data

[0834] The terminal plays the decoded audio data on an audio output device (e.g., a speaker or headphones).

[0835] Use VLC media player or the system's default audio playback function for playback.

[0836] Input: Decoded audio data

[0837] Data calculation: Audio data playback

[0838] Output: Playback of the converted audio from an audio output device

[0839] This allows the user to listen to the converted audio in real time.

[0840] (Application example 1)

[0841] 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."

[0842] Conventional voice conversion systems have problems when converting a user's voice into a different person's voice in real time, such as difficulty in achieving high-quality conversion, insufficient countermeasures against noise contamination, and difficulty in instantly adding a wide variety of voice effects depending on the distribution. In particular, for distribution services of entertainment and educational content, there is a demand for real-time conversion into a different voice, as well as support for live distribution and on-demand effects, in order to improve the user experience.

[0843] 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.

[0844] In this invention, the server includes means for acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for converting a user's voice into another voice in real time, means for live-streaming the converted voice data, and means for adding on-demand voice effects, thereby enabling real-time, high-quality voice conversion, the addition of a variety of effects, and the provision of attractive content in a distribution service.

[0845] The "means for acquiring voice input" is a function for acquiring the voice uttered by the user as digital data using a device such as a microphone.

[0846] The "means for preprocessing the acquired audio" is a function for performing noise removal and volume adjustment on the acquired audio data to improve the quality.

[0847] The "means for transmitting preprocessed voice data to a server" is a function for transmitting preprocessed voice data to a server via the Internet or other communication means.

[0848] "Means for converting transmitted voice data into another voice" refers to a function that converts voice data received on a server into the voice of another person using an AI model, etc.

[0849] The "means for receiving converted voice data from the server" is a function for transmitting the converted voice data from the server again to the client terminal and receiving it.

[0850] The "means for encoding received audio data" is a function that performs processing to convert audio data received from the server into a format that can be used by the user.

[0851] "Means for playing encoded audio" refers to a function for playing encoded audio data through speakers or headphones.

[0852] "Means for converting a user's voice into a different voice in real time" refers to technology and devices for converting a user's voice into a different voice as the user speaks.

[0853] The "means for live distribution of converted voice data" is a function for distributing converted voice data in real time via the Internet.

[0854] The "means for adding on-demand audio effects" is a function for adding new audio effects to existing audio data to improve its entertainment value and educational effectiveness.

[0855] The present invention is a system for converting a user's voice into another person's voice in real time. The implementation of the present invention includes the following components and procedures.

[0856] System Configuration

[0857] 1. Hardware:

[0858] Device: A smartphone or other audio input device that has a microphone and speaker.

[0859] Server: A high-performance computing device for processing voice data online.

[0860] 2. Software:

[0861] WebRTC: A library for capturing, transmitting, and receiving voice data in real time.

[0862] TensorFlow: A machine learning library for building and using AI voice models.

[0863] Flask: A web framework for processing audio data on the server side.

[0864] Processing flow

[0865] 1. Audio capture:

[0866] The user speaks into the smartphone's microphone, and the device captures the voice input and stores it as digital data using data capture software (e.g., WebRTC).

[0867] 2. Pretreatment:

[0868] The captured audio data is pre-processed with noise removal, volume adjustment, etc. This improves the quality of the audio data and increases the accuracy of the subsequent conversion process.

[0869] 3. Data transmission:

[0870] The pre-processed audio data is then transmitted from the device to a server over the internet in real time using WebRTC.

[0871] 4. Audio conversion:

[0872] The server converts the received voice data into another voice using an AI voice conversion model (using TensorFlow). Deep learning models are applied in this step to achieve high-precision voice conversion.

[0873] 5. Data Reception:

[0874] The converted voice data is sent from the server back to the terminal and received in real time.

[0875] 6. Encoding and Playback:

[0876] The device then encodes the received audio data into a playable format and plays it through speakers or headphones, allowing the user to hear the converted version of their own voice.

[0877] Specific examples

[0878] For example, a user may want to change their voice to sound like a historical figure during an online educational lecture. When the user says "hello," the voice is captured by the device's microphone and preprocessed to remove noise. The preprocessed voice data is then sent over the Internet to a server, where an AI model on the server converts the voice data into the voice of the specified person. The converted voice data is then sent back to the device and played back in real time on the device.

[0879] Prompt Sentence Examples

[0880] Please upload an audio file from the user:

[0881]

[0882] Choose an AI model to transform your voice:

[0883] In this way, the present invention provides users with a high quality real-time voice conversion experience, which is very useful in the fields of entertainment and education.

[0884] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0885] Step 1:

[0886] The user speaks into the smartphone's microphone. The device receives this voice input and captures it as digital data. Here, the input data is the user's voice data, and the output data is the digitized voice data. In this step, the microphone is used to convert the analog voice signal into digital data, which is then saved by the voice capture software.

[0887] Step 2:

[0888] The device performs preprocessing on the captured audio data. This involves processes to improve the quality of the audio data, such as noise removal and volume adjustment. Here, the input is the digital audio data captured in step 1, and the output is the preprocessed audio data. Specifically, a library such as webrtcvad is used to remove noise and apply a volume normalization algorithm.

[0889] Step 3:

[0890] The preprocessed voice data is transmitted to a server via the Internet. The terminal divides the digital data into packets and transmits them using a communication module, where the input is the preprocessed voice data and the output is the data packets transmitted via network communication.

[0891] Step 4:

[0892] The server converts the received voice data into a different voice using an AI voice conversion model (using a generative AI model). The server uses the received data as input, applies the generative AI model, and outputs voice data of a different person. Specifically, it executes a deep learning model using the TensorFlow library and performs voice conversion based on the specified prompt sentence.

[0893] Step 5:

[0894] The converted voice data is sent back to the terminal from the server, where it is re-encoded into packets and sent to the terminal via the Internet, where the input is the converted voice data and the output is the data packets received by the terminal via network communication.

[0895] Step 6:

[0896] The terminal decodes the received converted audio data into a playable format. The decoding process involves converting the audio data using signal processing algorithms and converting it into a file format for playback through speakers or headphones. Here, the input is the data packets received from the server, and the output is playable audio data.

[0897] Step 7:

[0898] The device plays the decoded audio data over speakers or headphones, allowing the user to listen to another person's voice in real time. Here, the input is the decoded audio data, and the output is the physical playback of audio. Specifically, audio playback software converts the digital audio signal to an analog signal and plays it on an audio output device.

[0899] 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.

[0900] The present invention provides a system for converting a user's voice into a different person's voice in real time, and further recognizing the user's emotions and reflecting them in the voice conversion. The system includes the following means.

[0901] First, we will explain how to obtain voice input. A user speaks into the microphone of the device. For example, the user says "hello." This voice input is captured by dedicated software in the device.

[0902] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[0903] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[0904] Next, we will explain how the server converts voice data. The server inputs the received voice data into an AI voice conversion model. This model is based on a deep learning algorithm, extracts voice features, and converts them into a specified target voice.

[0905] Next, we will explain the emotion engine. The server analyzes the voice data provided by the terminal and recognizes the user's emotional state. For example, if the user is angry or happy, the emotion engine determines that emotion.

[0906] Furthermore, we describe a method for adjusting the voice conversion model based on the recognized emotions. The server uses feedback from the emotion engine to readjust the voice conversion model to match the user's emotions. This adjustment allows for voice conversion that is faithful to the user's emotions.

[0907] This section explains how to receive converted audio data from the server. The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[0908] This section then describes how the terminal decodes and plays back the audio data. The terminal decodes the encoded audio data it receives and converts it into a playable format. This decoding process is performed using an audio playback library.

[0909] Finally, we will explain how to play back the encoded voice data. The terminal plays back the decoded voice data through a speaker or headphones. This allows the user to hear in real time the voice of another person that has been converted from their own voice. This voice reflects the user's emotions.

[0910] Illustrative Usage Scenarios

[0911] For example, if User A says "Hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model and uses an emotion engine to recognize User A's emotion of "anger." The voice conversion model is then adjusted based on this emotion and converted into User B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear User B's angry voice saying "Hello" in real time, without any delay.

[0912] As described above, the present invention effectively executes all processes from voice input acquisition to emotion recognition, conversion, and playback, thereby providing users with an advanced real-time voice conversion experience.

[0913] The processing flow will be explained below.

[0914] Step 1:

[0915] The user speaks into the microphone of the terminal, for example, the user says "hello."

[0916] Step 2:

[0917] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[0918] Step 3:

[0919] The device performs preprocessing on the acquired voice data, such as noise reduction and volume adjustment, to remove excess noise to improve the clarity of the voice and to normalize the voice volume to keep it constant.

[0920] Step 4:

[0921] The device packages the preprocessed audio data into an appropriate format (e.g., binary format) for transmission to the server. The packaged data is sent to the server via the Internet. HTTP or WebSocket is often used as the communication protocol.

[0922] Step 5:

[0923] The server analyzes the received voice data and inputs it into an emotion engine that recognizes the user's emotions. The emotion engine identifies the user's emotional state (e.g., joy, anger, sadness, etc.) based on the tone, rhythm, and intonation of the voice.

[0924] Step 6:

[0925] The server acquires the emotion information recognized by the emotion engine and adjusts the AI ​​voice conversion model based on that information. For example, if the user is angry, the AI ​​model will set the voice conversion parameters to reflect the angry emotion.

[0926] Step 7:

[0927] The server uses a trained AI voice conversion model to convert the pre-processed voice data into a different voice, which faithfully matches the target voice based on the user's voice characteristics and emotional information.

[0928] Step 8:

[0929] The server then encodes the converted audio data into a playable format (such as OGG or MP3) and sends it to the device. The encoded data is compressed for efficient transmission.

[0930] Step 9:

[0931] The terminal receives the encoded voice data sent from the server, and the receiving process follows a pre-established communication protocol.

[0932] Step 10:

[0933] The device then decodes the received encoded audio data and converts it into a playable format. This decoding process is performed using an audio playback library, which removes the compression applied during encoding and restores the original audio format.

[0934] Step 11:

[0935] The device then plays the decoded voice data over a speaker or headphones, allowing the user to listen to a voice that faithfully reflects the user's emotional information, providing a more natural communication experience.

[0936] Example 2

[0937] 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."

[0938] Conventional speech conversion systems simply convert speech into another speech without taking into account the user's emotions, resulting in unnatural speech and poor emotional expression. Real-time speech conversion also entails delays, which can negatively impact the user experience. It is necessary to solve these problems and achieve more natural and emotionally rich speech conversion.

[0939] 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.

[0940] In this invention, the server includes means for converting transmitted voice data into a specified target voice, means for analyzing the voice data to recognize the user's emotional state, and means for adjusting the conversion model based on the recognized emotional state, thereby enabling natural voice conversion that reflects the user's emotions.

[0941] "User" refers to a person who uses the system to provide voice input.

[0942] "Terminal" refers to a device that acquires audio, performs preprocessing, and then transmits and receives the audio data to a server.

[0943] "Voice input" refers to voice data uttered by a user through a microphone of a terminal.

[0944] "Preprocessing" refers to the act of performing processes such as noise removal and volume adjustment on audio data to improve data quality.

[0945] "Server" refers to a device that receives voice data sent from a terminal and converts it into the voice of a specified target.

[0946] A "voice conversion model" refers to an algorithm that uses AI technology to convert voice data into another voice.

[0947] "Emotion engine" refers to an algorithm that analyzes and recognizes the user's emotional state from voice data.

[0948] "Adjusting the conversion model" refers to the act of adjusting the voice conversion model based on the analysis results of the emotion engine, and performing conversion that reflects emotions.

[0949] "Encoding" refers to the act of converting audio data into a playable format.

[0950] "Decoding" refers to the act of restoring encoded audio data to a playable format.

[0951] "Playback" refers to the act of allowing a user to listen to audio data through the terminal's speakers or headphones.

[0952] This invention is a system that converts a user's voice into another person's voice in real time, recognizes the user's emotions, and reflects them in the voice conversion. The following hardware and software are used to implement this system.

[0953] First, a device needs a microphone as hardware to acquire voice input. A user inputs voice by speaking into the microphone. For example, a user might say "hello." This voice input is captured by a voice capture application in the device.

[0954] Next, the device preprocesses the captured audio data by noise reduction and volume adjustment using a noise reduction software library (e.g., Audacity). This preprocessing improves the quality of the audio data and increases conversion accuracy.

[0955] The pre-processed voice data is sent to a server over the Internet, usually using the HTTP protocol. The server then inputs the received voice data into an AI voice conversion model (e.g., WaveNet) to convert it into a specified target voice. This model is based on a deep learning algorithm and analyzes the voice characteristics to perform the conversion.

[0956] The server is also equipped with an emotion engine that identifies the user's emotions from the received voice data. Software such as IBM Tone Analyzer is used for emotion recognition. The recognized emotions (e.g., anger, joy, etc.) are fed back to the voice conversion model, which is then adjusted according to the user's emotions.

[0957] The adjusted audio data is then sent from the server to the device. The server encodes the audio data into a playable format (e.g., OGG or MP3) and sends it to the device. The device then decodes the encoded audio data and converts it into a playable format using a playback library (e.g., FFmpeg).

[0958] Finally, the device plays the decoded voice data over a speaker or headphones, allowing the user to hear their own voice transformed into a different person's voice in real time, with the voice reflecting the user's emotions.

[0959] Specific examples

[0960] For example, if user A says "hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model (WaveNet) and uses an emotion engine (IBM Tone Analyzer) to recognize user A's emotion of "anger." The voice conversion model is then adjusted based on this emotion, and the voice is converted into user B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear user B's angry voice saying "hello" in real time, without any delay.

[0961] In this way, the system of the present invention effectively executes a series of processes from voice input acquisition to emotion recognition, conversion, and playback, providing users with an advanced real-time voice conversion experience.

[0962] Example prompt sentences to use

[0963] Example prompt sentence:

[0964] User A says "Hello" in an angry voice. The audio is captured, noise is removed, and then input into the speech conversion model. During conversion, the emotion recognition engine detects the emotion "anger" and converts it into the angry voice of User B. Play the converted audio.

[0965] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0966] Step 1:

[0967] Acquiring voice input

[0968] Input: The user's voice speaking into the device's microphone (e.g., "Hello")

[0969] Specific operation: The user speaks into the microphone of the terminal.

[0970] Data processing: The device uses an audio capture application to capture this audio as digital audio data.

[0971] Output: Captured audio data (e.g., WAV file with a sampling rate of 44.1 kHz)

[0972] Step 2:

[0973] Audio data preprocessing

[0974] Input: Captured audio data

[0975] Specific operation: The device uses a noise reduction software library (e.g., Audacity) to remove noise from the audio data and adjust the volume.

[0976] Data processing: Apply noise filtering algorithms to improve audio clarity.

[0977] Output: Preprocessed audio data (e.g., a denoised WAV file)

[0978] Step 3:

[0979] Sending audio data to the server

[0980] Input: Preprocessed audio data

[0981] Specific operation: The terminal sends audio data to the server using the HTTP protocol.

[0982] Data processing: The audio data is sent in the body of the HTTP request.

[0983] Output: Audio data received by the server

[0984] Step 4:

[0985] Processing with voice conversion models

[0986] Input: Audio data sent to the server

[0987] Specific operation: The server inputs voice data into an AI voice conversion model (e.g., WaveNet).

[0988] Data Computing: Using deep learning algorithms to extract voice features and convert them into the target voice.

[0989] Output: Converted voice data (e.g., a WAV file converted into a different person's voice)

[0990] Step 5:

[0991] Emotion determination using an emotion recognition engine

[0992] Input: Audio data sent to the server

[0993] Specific operation: The server inputs voice data into an emotion recognition engine (e.g., IBM Tone Analyzer).

[0994] Data processing: Analyze the voice data and extract emotional features.

[0995] Output: The user's emotional state (e.g., "anger" or "joy")

[0996] Step 6:

[0997] Tuning the speech conversion model

[0998] Input: Feedback from the emotion recognition engine (user's emotional state)

[0999] Specific operation: The server adjusts the voice conversion model based on the results of the emotion recognition engine.

[1000] Data calculation: Taking into account emotion features, the parameters of the conversion model are changed.

[1001] Output: A voice conversion model that reflects the user's emotions

[1002] Step 7:

[1003] Encoding the converted audio data

[1004] Input: Speech data converted by the trained speech conversion model

[1005] What happens: The server encodes the audio data into a playable format (e.g., OGG or MP3).

[1006] Data processing: Converting file formats using audio encoding algorithms.

[1007] Output: Encoded audio data (e.g. MP3 file)

[1008] Step 8:

[1009] Sending encoded audio data to the device

[1010] Input: Encoded audio data

[1011] Specific operation: The server uses the HTTP protocol to send encoded audio data to the terminal.

[1012] Data processing: The audio data is sent in the body of the HTTP response.

[1013] Output: The encoded audio data received by the device.

[1014] Step 9:

[1015] Decoding audio data

[1016] Input: Encoded audio data

[1017] Specific behavior: The device uses an audio playback library (e.g., FFmpeg) to decode the encoded audio data.

[1018] Data processing: Converting from an encoded format to a playable format.

[1019] Output: Decoded audio data (e.g. WAV file)

[1020] Step 10:

[1021] Playing audio data

[1022] Input: Decoded audio data

[1023] Specific operation: The device plays audio data using a playback device (speaker or headphones).

[1024] Data calculation: The decoded audio is converted from digital to analog and output.

[1025] Output: The audio the user hears (a modified voice, e.g., "Hello" with the emotion "Anger")

[1026] (Application example 2)

[1027] 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."

[1028] In conventional brick-and-mortar stores, it has been difficult to immediately grasp a customer's emotions and respond appropriately. Furthermore, if a customer is angry or happy, it is not possible to communicate with them in a voice that reflects their emotions, which can lead to a decline in the quality of customer service. By solving these problems, it is necessary to provide more effective customer service and improve customer satisfaction.

[1029] 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 acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for recognizing emotions and reflecting them in voice conversion, and means for adjusting voice tone based on the customer's emotions using an application installed on the smart device. This enables real-time voice conversion that reflects customer emotions in physical stores.

[1030] The "means for acquiring voice input" refers to a device or method for acquiring the voice spoken by the user as digital data using an acoustic sensor such as a microphone.

[1031] "Means for preprocessing captured audio" refers to a device or method that performs preprocessing such as noise removal and volume adjustment on captured audio data to improve the quality of the data and increase the accuracy of subsequent processing.

[1032] The "means for transmitting preprocessed voice data to a server" refers to a device or method for transmitting preprocessed voice data to a server via a communication line such as the Internet.

[1033] "Means for converting transmitted voice data into another voice" refers to a device or method for converting received voice data into another voice using an AI voice model or other conversion algorithm.

[1034] The "means for receiving converted voice data from the server" refers to a device or method for the server to send the converted voice data back to the client device via a communication line such as the Internet.

[1035] "Means for encoding received audio data" refers to a device or method for converting received audio data into a playable format (e.g., OGG or MP3 format).

[1036] "Means for playing encoded audio" refers to an apparatus or method for playing encoded audio data using a playback device (speakers or headphones).

[1037] "Means for recognizing emotions and reflecting them in voice conversion" refers to a device or method that analyzes the emotions from the user's voice, adjusts the voice conversion model based on the recognition results, and converts the voice into a different voice that reflects the emotions.

[1038] "Means for adjusting voice tone based on customer emotions using an application installed on a smart device" refers to a device or method that uses software installed on a device such as smart glasses or a smartphone to adjust voice tone in real time during customer service based on the customer's emotions.

[1039] The present invention provides a system that converts a user's voice into a different person's voice in real time, and further recognizes the user's emotions and reflects them in the voice conversion. Specific embodiments of the system are described below.

[1040] 1. System Configuration

[1041] This system is primarily composed of smart devices, internet communications, and servers. It includes the following elements:

[1042] 1. A device with a microphone to capture the user's voice (e.g., smart glasses or a smartphone)

[1043] 2. Dedicated software for preprocessing audio data (e.g., noise reduction, volume adjustment)

[1044] 3. A data communication module for transmitting the preprocessed audio data to a server via the Internet.

[1045] 4. AI voice conversion model for converting voice data into another voice on the server

[1046] 5. Emotion engine for analyzing and recognizing user emotions on the server

[1047] 6. An audio playback library to encode the converted audio data (e.g., OGG or MP3 format)

[1048] 7. Speakers or headphones to play the received encoded audio on your device.

[1049] 2. Program Processing

[1050] When a user speaks into the microphone, the device captures the voice and preprocesses it using dedicated software. This preprocessing includes noise reduction and volume adjustment. A data communications module is used to send the processed voice data to a server. The sent voice data is then input into an AI voice conversion model on the server and converted into a different voice.

[1051] Next, the emotion engine on the server analyzes the voice data and recognizes the user's emotions. Based on this recognition result, the AI ​​voice conversion model is re-adjusted and converted into voice that reflects the emotion. The converted voice data is sent back from the server to the device, where it is encoded by the voice playback library and played back through speakers or headphones.

[1052] 3. Example Use Case Scenarios

[1053] Below is a usage scenario for a physical store.

[1054] Scenario 1: The customer is angry

[1055] A salesperson wears smart glasses and interacts with a customer. If the customer is angry, the salesperson responds by inputting a prompt sentence as follows:

[1056] Prompt statement:

[1057] "When a customer is angry, change the salesperson's voice to a calmer one and speak in a more soothing manner."

[1058] Scenario 2: Happy Customers

[1059] In the situation where the customer is happy, the salesperson enters the following prompt sentence:

[1060] Prompt statement:

[1061] "When a customer is happy, the salesperson should change their voice to a more cheerful and upbeat tone."

[1062] By implementing the present invention in this way, real-time voice conversion that reflects customer emotions becomes possible in physical stores, thereby improving customer satisfaction.

[1063] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1064] Step 1:

[1065] A user speaks into the microphone of a smart device (e.g., smart glasses or a smartphone). This voice input is captured as digital data. The input is the user's speech, and the output is digital voice data. Specifically, the microphone converts sound waves into electrical signals, which are then digitized and captured as voice data.

[1066] Step 2:

[1067] The device performs preprocessing on the acquired voice data. Preprocessing includes noise reduction (removing noise from the voice data) and volume adjustment (adjusting the volume of the voice to an appropriate level). The input is digital voice data, and the output is preprocessed voice data. Specifically, the noise reduction algorithm removes unnecessary components from the voice data, improving the sound quality.

[1068] Step 3:

[1069] The device sends the preprocessed audio data to the server via the Internet. This is done using a data communication module. The input is the preprocessed audio data, and the output is the audio data received by the server. Specifically, the audio data is sent to the server using a network protocol (e.g., HTTP or WebSocket).

[1070] Step 4:

[1071] The server inputs the received voice data into an AI voice conversion model and converts it into another voice. The input is the voice data received on the server side, and the output is the converted voice data. Specifically, the generative AI model extracts voice features and converts them into the specified target voice.

[1072] Step 5:

[1073] The server runs an emotion engine that analyzes the voice data and recognizes the user's emotions. The input is the converted voice data, and the output is the user's emotional state. Specifically, the emotion recognition algorithm analyzes changes in voice tone and pitch to determine emotions.

[1074] Step 6:

[1075] The server readjusts the AI ​​voice conversion model based on the results of the emotion engine. The input is the user's emotional state, and the output is voice data readjusted to reflect the emotion. Specifically, the AI ​​model processes the voice data again and adds appropriate voice tone and emphasis.

[1076] Step 7:

[1077] The server encodes the converted audio data and converts it into a playable format (e.g., OGG or MP3 format). The input is the reconditioned audio data, and the output is the encoded audio data. Specifically, the encoding library converts the audio data into a playable format.

[1078] Step 8:

[1079] The terminal receives encoded voice data from the server. The input is the encoded voice data, and the output is the voice data received on the terminal side. Specifically, the received data is obtained from the network and stored in memory.

[1080] Step 9:

[1081] The device decodes the received audio data and plays it on speakers or headphones. The input is the encoded audio data, and the output is the audio to be played. Specifically, the device uses a decoding library to convert the audio data into a digital audio signal and plays it on the audio output device.

[1082] By going through this series of steps, the user's voice is converted in real time into a different person's voice that reflects their emotions, and can be used in situations such as physical stores.

[1083] 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.

[1084] 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.

[1085] 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.

[1086] [Fourth embodiment]

[1087] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1088] 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.

[1089] 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).

[1090] 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.

[1091] 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.

[1092] 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).

[1093] 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.

[1094] 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.

[1095] 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.

[1096] 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.

[1097] 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.

[1098] 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.

[1099] 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."

[1100] The present invention provides a system for converting a user's voice into a voice of another person in real time. The system includes the following means.

[1101] First, we will explain how to obtain voice input: the user speaks into the device's microphone, and this voice input is captured by dedicated software within the device.

[1102] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[1103] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[1104] Next, we will explain how the server converts voice data. The server processes the received voice data based on an AI voice conversion model and converts it into another specified voice. This voice conversion uses a highly accurate deep learning model.

[1105] Next, we will explain how to receive the converted voice data from the server. The server re-encodes the converted voice data and sends it to the terminal. The terminal receives this encoded voice data.

[1106] Then, we explain how to encode the audio data. The terminal decodes the received audio data and converts it into a playable format. The encoded audio data is then processed so that it can be used by the user in a specific format.

[1107] Finally, we explain how to play the encoded audio data. The device plays the audio data encoded in a playable format through speakers or headphones. This allows the user to listen to the new audio converted from their own voice in real time.

[1108] Illustrative Usage Scenarios

[1109] For example, when user A says "hello," the audio is captured by the device's microphone. The device removes noise from the audio and sends it to the server. The server processes the received audio using an AI voice conversion model and converts it into user B's voice. This converted audio data is then sent back from the server to the device, where it is played back. This series of processes is performed in real time, so user A can hear "hello" in user B's voice without any delay.

[1110] As described above, the present invention effectively executes all processes from acquiring voice input to converting and playing, thereby providing users with an advanced real-time voice conversion experience.

[1111] The processing flow will be explained below.

[1112] Step 1:

[1113] The user speaks into the microphone of the terminal, for example, the user says "hello."

[1114] Step 2:

[1115] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[1116] Step 3:

[1117] The device performs preprocessing on the acquired audio data, such as noise reduction and volume normalization, which includes filtering and volume adjustment using an audio signal processing library.

[1118] Step 4:

[1119] The terminal packages the pre-processed audio data into a suitable format (e.g., binary format) for transmission to the server, which transmits the packaged data over the Internet to the server.

[1120] Step 5:

[1121] The server analyzes the received voice data and inputs it into an AI voice conversion model, which is based on a deep learning algorithm to extract voice features and convert them into a specified target voice.

[1122] Step 6:

[1123] The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[1124] Step 7:

[1125] The terminal receives the encoded voice data sent from the server, and this reception process uses a common data communication protocol.

[1126] Step 8:

[1127] The device decodes the received audio data and converts it into a playable format. This decoding process is performed using an audio playback library.

[1128] Step 9:

[1129] The device then plays the decoded audio data over a speaker or headphones, allowing the user to hear the converted voice of another person in real time.

[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 robot 414 will be referred to as a "terminal."

[1132] Conventional speech conversion systems often suffer from degradation of speech quality and conversion delays during the process from speech capture to final speech playback. Furthermore, noise removal accuracy is low, which can result in a loss of naturalness in the converted speech. Furthermore, it is difficult to utilize highly accurate models in the conversion process. Therefore, converting speech in real time while providing high-quality speech remains a challenge.

[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 acquiring a user's voice, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server via a data communication network, means for converting the transmitted voice data into another voice using a deep learning model, means for receiving the converted voice data from the server, means for decoding the received voice data into a playable format, and means for playing the decoded voice on an audio output device, thereby enabling voice conversion in real time while maintaining the quality of the voice.

[1135] "User" refers to the person who uses the system and is responsible for inputting voice and listening to the final converted voice.

[1136] "Speech" refers to the sound produced by a person, which is picked up by a microphone and processed within the system.

[1137] "Preprocessing" refers to processing that removes noise and adjusts the volume of acquired audio data to improve the quality of the data.

[1138] "Data communications network" means a digital communications infrastructure, such as the Internet, used to transmit voice data from a terminal to a server or from a server to a terminal.

[1139] The "server" is a computer system that handles the conversion of voice data and performs voice conversion using a deep learning model.

[1140] A "deep learning model" is a type of artificial intelligence that uses neural networks to perform advanced conversion and analysis of voice data.

[1141] "Decoding" is the process of converting encoded data back into its original, playable format.

[1142] "Audio output device" refers to a device, such as a speaker or headphones, used to play decoded audio to a user.

[1143] "Noise reduction" is the process of removing unwanted noise from audio data, improving the clarity of the audio.

[1144] "Playable format" means that audio data has been converted into an appropriate format so that the user can listen to the audio.

[1145] This invention is a system that converts user-spoken speech into another speech in real time. This system effectively executes a series of processes from speech input to speech output, focusing on three elements: the user, the terminal, and the server. The specific configuration and operation are described below.

[1146] First, the user speaks into the device's microphone. This speech is captured in real time using dedicated software such as Audacity. The captured speech data is then pre-processed on the device. This pre-processing includes noise reduction and volume adjustment to improve the quality of the speech data. Audacity's "noise reduction" and "volume normalization" functions are used in this process.

[1147] The preprocessed audio data is then sent over a data communications network to a server, typically using a protocol such as an HTTP POST request. The server processes the received audio data using a deep learning model, such as Google Cloud Text-to-Speech or OpenAI's Whisper model, to convert it into a different sound.

[1148] The converted audio data is then encoded on the server and sent back to the device via the data communications network. For example, ffmpeg is used to encode the audio data into an audio format such as FLAC or AAC. The device then decodes the received audio data and converts it into a playable format. Once decoded, the audio data is played on the device's audio output device (speakers or headphones) using the device's VLC media player or the system's default audio playback function.

[1149] For example, when user A says "hello," the audio is captured through the device's microphone. Audacity software captures the audio in real time and performs noise reduction and volume normalization. The processed audio data is sent to the server using an HTTP POST request. The server uses the Google Cloud Text-to-Speech API to convert the audio into user B's voice and sends the converted audio back to the device. On the device, the FLAC format audio data is decoded into MP3 format using ffmpeg and played back using VLC media player. In this way, user A can hear "hello" in user B's voice, converted in real time.

[1150] As described above, the present invention executes the entire process from acquiring voice input to conversion and playback with high quality and in real time, providing users with an advanced voice conversion experience.

[1151] Prompt Sentence Examples

[1152] "Please explain the specific steps involved in converting a user's voice into a different person's voice in real time. Please describe in detail the process from voice capture to conversion and playback, and indicate the software and hardware used in the process."

[1153] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1154] Step 1: Getting voice input

[1155] The user speaks into the microphone of the terminal, for example, saying "hello."

[1156] The device captures this audio using specialized software (e.g., Audacity).

[1157] Input: User's live voice

[1158] Data processing: Audio is captured as digital data using dedicated software

[1159] Output: Unprocessed digital audio data

[1160] Step 2: Preprocessing

[1161] The terminal performs noise reduction and volume adjustment on the captured digital audio data.

[1162] Use Audacity's "Noise Reduction" and "Volume Normalization" features.

[1163] Input: Unprocessed digital audio data

[1164] Data processing: noise removal and volume adjustment

[1165] Output: Pre-processed digital audio data

[1166] Step 3: Sending audio data to the server

[1167] The terminal transmits the pre-processed voice data to a server via a data communication network (eg, the Internet).

[1168] Use an HTTP POST request.

[1169] Input: Preprocessed digital audio data

[1170] Data processing: Data communication via HTTP POST request

[1171] Output: Audio data sent to the server

[1172] Step 4: Audio conversion by server

[1173] The server converts the transmitted voice data into a different voice using a deep learning model (e.g., Google Cloud Text-to-Speech or OpenAI's Whisper model).

[1174] Input: Audio data sent to the server

[1175] Data Computing: Voice Conversion with Deep Learning Models

[1176] Output: Converted audio data

[1177] Step 5: Translated audio data sent from the server

[1178] The server encodes the converted voice data and transmits it to the terminal via a data communication network.

[1179] For encoding, ffmpeg is used to apply audio formats such as FLAC and AAC.

[1180] Input: Converted audio data

[1181] Data processing: Encoding into audio format

[1182] Output: Encoded audio data sent to the device

[1183] Step 6: Decode the audio data

[1184] The terminal decodes the audio data received from the server into a reproducible format.

[1185] Use ffmpeg to decode to MP3 or other formats.

[1186] Input: Encoded audio data received from the server

[1187] Data processing: Decoding into a playable format

[1188] Output: Decoded audio data

[1189] Step 7: Playing back audio data

[1190] The terminal plays the decoded audio data on an audio output device (e.g., a speaker or headphones).

[1191] Use VLC media player or the system's default audio playback function for playback.

[1192] Input: Decoded audio data

[1193] Data calculation: Audio data playback

[1194] Output: Playback of the converted audio from an audio output device

[1195] This allows the user to listen to the converted audio in real time.

[1196] (Application example 1)

[1197] 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."

[1198] Conventional voice conversion systems have problems when converting a user's voice into a different person's voice in real time, such as difficulty in achieving high-quality conversion, insufficient countermeasures against noise contamination, and difficulty in instantly adding a wide variety of voice effects depending on the distribution. In particular, for distribution services of entertainment and educational content, there is a demand for real-time conversion into a different voice, as well as support for live distribution and on-demand effects, in order to improve the user experience.

[1199] 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.

[1200] In this invention, the server includes means for acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for converting a user's voice into another voice in real time, means for live-streaming the converted voice data, and means for adding on-demand voice effects, thereby enabling real-time, high-quality voice conversion, the addition of a variety of effects, and the provision of attractive content in a distribution service.

[1201] The "means for acquiring voice input" is a function for acquiring the voice uttered by the user as digital data using a device such as a microphone.

[1202] The "means for preprocessing the acquired audio" is a function for performing noise removal and volume adjustment on the acquired audio data to improve the quality.

[1203] The "means for transmitting preprocessed voice data to a server" is a function for transmitting preprocessed voice data to a server via the Internet or other communication means.

[1204] "Means for converting transmitted voice data into another voice" refers to a function that converts voice data received on a server into the voice of another person using an AI model, etc.

[1205] The "means for receiving converted voice data from the server" is a function for transmitting the converted voice data from the server again to the client terminal and receiving it.

[1206] The "means for encoding received audio data" is a function that performs processing to convert audio data received from the server into a format that can be used by the user.

[1207] "Means for playing encoded audio" refers to a function for playing encoded audio data through speakers or headphones.

[1208] "Means for converting a user's voice into a different voice in real time" refers to technology and devices for converting a user's voice into a different voice as the user speaks.

[1209] The "means for live distribution of converted voice data" is a function for distributing converted voice data in real time via the Internet.

[1210] The "means for adding on-demand audio effects" is a function for adding new audio effects to existing audio data to improve its entertainment value and educational effectiveness.

[1211] The present invention is a system for converting a user's voice into another person's voice in real time. The implementation of the present invention includes the following components and procedures.

[1212] System Configuration

[1213] 1. Hardware:

[1214] Device: A smartphone or other audio input device that has a microphone and speaker.

[1215] Server: A high-performance computing device for processing voice data online.

[1216] 2. Software:

[1217] WebRTC: A library for capturing, transmitting, and receiving voice data in real time.

[1218] TensorFlow: A machine learning library for building and using AI voice models.

[1219] Flask: A web framework for processing audio data on the server side.

[1220] Processing flow

[1221] 1. Audio capture:

[1222] The user speaks into the smartphone's microphone, and the device captures the voice input and stores it as digital data using data capture software (e.g., WebRTC).

[1223] 2. Pretreatment:

[1224] The captured audio data is pre-processed with noise removal, volume adjustment, etc. This improves the quality of the audio data and increases the accuracy of the subsequent conversion process.

[1225] 3. Data transmission:

[1226] The pre-processed audio data is then transmitted from the device to a server over the internet in real time using WebRTC.

[1227] 4. Audio conversion:

[1228] The server converts the received voice data into another voice using an AI voice conversion model (using TensorFlow). Deep learning models are applied in this step to achieve high-precision voice conversion.

[1229] 5. Data Reception:

[1230] The converted voice data is sent from the server back to the terminal and received in real time.

[1231] 6. Encoding and Playback:

[1232] The device then encodes the received audio data into a playable format and plays it through speakers or headphones, allowing the user to hear the converted version of their own voice.

[1233] Specific examples

[1234] For example, a user may want to change their voice to sound like a historical figure during an online educational lecture. When the user says "hello," the voice is captured by the device's microphone and preprocessed to remove noise. The preprocessed voice data is then sent over the Internet to a server, where an AI model on the server converts the voice data into the voice of the specified person. The converted voice data is then sent back to the device and played back in real time on the device.

[1235] Prompt Sentence Examples

[1236] Please upload an audio file from the user:

[1237]

[1238] Choose an AI model to transform your voice:

[1239] In this way, the present invention provides users with a high quality real-time voice conversion experience, which is very useful in the fields of entertainment and education.

[1240] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1241] Step 1:

[1242] The user speaks into the smartphone's microphone. The device receives this voice input and captures it as digital data. Here, the input data is the user's voice data, and the output data is the digitized voice data. In this step, the microphone is used to convert the analog voice signal into digital data, which is then saved by the voice capture software.

[1243] Step 2:

[1244] The device performs preprocessing on the captured audio data. This involves processes to improve the quality of the audio data, such as noise removal and volume adjustment. Here, the input is the digital audio data captured in step 1, and the output is the preprocessed audio data. Specifically, a library such as webrtcvad is used to remove noise and apply a volume normalization algorithm.

[1245] Step 3:

[1246] The preprocessed voice data is transmitted to a server via the Internet. The terminal divides the digital data into packets and transmits them using a communication module, where the input is the preprocessed voice data and the output is the data packets transmitted via network communication.

[1247] Step 4:

[1248] The server converts the received voice data into a different voice using an AI voice conversion model (using a generative AI model). The server uses the received data as input, applies the generative AI model, and outputs voice data of a different person. Specifically, it executes a deep learning model using the TensorFlow library and performs voice conversion based on the specified prompt sentence.

[1249] Step 5:

[1250] The converted voice data is sent back to the terminal from the server, where it is re-encoded into packets and sent to the terminal via the Internet, where the input is the converted voice data and the output is the data packets received by the terminal via network communication.

[1251] Step 6:

[1252] The terminal decodes the received converted audio data into a playable format. The decoding process involves converting the audio data using signal processing algorithms and converting it into a file format for playback through speakers or headphones. Here, the input is the data packets received from the server, and the output is playable audio data.

[1253] Step 7:

[1254] The device plays the decoded audio data over speakers or headphones, allowing the user to listen to another person's voice in real time. Here, the input is the decoded audio data, and the output is the physical playback of audio. Specifically, audio playback software converts the digital audio signal to an analog signal and plays it on an audio output device.

[1255] 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.

[1256] The present invention provides a system for converting a user's voice into a different person's voice in real time, and further recognizing the user's emotions and reflecting them in the voice conversion. The system includes the following means.

[1257] First, we will explain how to obtain voice input. A user speaks into the microphone of the device. For example, the user says "hello." This voice input is captured by dedicated software in the device.

[1258] Next, we will explain how to perform preprocessing. The terminal performs preprocessing on the acquired voice data, such as noise removal and volume adjustment. This preprocessing improves the quality of the voice data and increases the conversion accuracy.

[1259] After that, we will explain how to send the voice data to the server. The terminal sends the pre-processed voice data to the server via data communication, which is usually done via the Internet.

[1260] Next, we will explain how the server converts voice data. The server inputs the received voice data into an AI voice conversion model. This model is based on a deep learning algorithm, extracts voice features, and converts them into a specified target voice.

[1261] Next, we will explain the emotion engine. The server analyzes the voice data provided by the terminal and recognizes the user's emotional state. For example, if the user is angry or happy, the emotion engine determines that emotion.

[1262] Furthermore, we describe a method for adjusting the voice conversion model based on the recognized emotions. The server uses feedback from the emotion engine to readjust the voice conversion model to match the user's emotions. This adjustment allows for voice conversion that is faithful to the user's emotions.

[1263] This section explains how to receive converted audio data from the server. The server encodes the converted audio data and sends it to the device. This encoding process converts the audio data into a playable format (e.g., OGG or MP3).

[1264] This section then describes how the terminal decodes and plays back the audio data. The terminal decodes the encoded audio data it receives and converts it into a playable format. This decoding process is performed using an audio playback library.

[1265] Finally, we will explain how to play back the encoded voice data. The terminal plays back the decoded voice data through a speaker or headphones. This allows the user to hear in real time the voice of another person that has been converted from their own voice. This voice reflects the user's emotions.

[1266] Illustrative Usage Scenarios

[1267] For example, if User A says "Hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model and uses an emotion engine to recognize User A's emotion of "anger." The voice conversion model is then adjusted based on this emotion and converted into User B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear User B's angry voice saying "Hello" in real time, without any delay.

[1268] As described above, the present invention effectively executes all processes from voice input acquisition to emotion recognition, conversion, and playback, thereby providing users with an advanced real-time voice conversion experience.

[1269] The processing flow will be explained below.

[1270] Step 1:

[1271] The user speaks into the microphone of the terminal, for example, the user says "hello."

[1272] Step 2:

[1273] The device picks up spoken voice through a microphone and captures it as digital audio data using a dedicated voice recording library.

[1274] Step 3:

[1275] The device performs preprocessing on the acquired voice data, such as noise reduction and volume adjustment, to remove excess noise to improve the clarity of the voice and to normalize the voice volume to keep it constant.

[1276] Step 4:

[1277] The device packages the preprocessed audio data into an appropriate format (e.g., binary format) for transmission to the server. The packaged data is sent to the server via the Internet. HTTP or WebSocket is often used as the communication protocol.

[1278] Step 5:

[1279] The server analyzes the received voice data and inputs it into an emotion engine that recognizes the user's emotions. The emotion engine identifies the user's emotional state (e.g., joy, anger, sadness, etc.) based on the tone, rhythm, and intonation of the voice.

[1280] Step 6:

[1281] The server acquires the emotion information recognized by the emotion engine and adjusts the AI ​​voice conversion model based on that information. For example, if the user is angry, the AI ​​model will set the voice conversion parameters to reflect the angry emotion.

[1282] Step 7:

[1283] The server uses a trained AI voice conversion model to convert the pre-processed voice data into a different voice, which faithfully matches the target voice based on the user's voice characteristics and emotional information.

[1284] Step 8:

[1285] The server then encodes the converted audio data into a playable format (such as OGG or MP3) and sends it to the device. The encoded data is compressed for efficient transmission.

[1286] Step 9:

[1287] The terminal receives the encoded voice data sent from the server, and the receiving process follows a pre-established communication protocol.

[1288] Step 10:

[1289] The device then decodes the received encoded audio data and converts it into a playable format. This decoding process is performed using an audio playback library, which removes the compression applied during encoding and restores the original audio format.

[1290] Step 11:

[1291] The device then plays the decoded voice data over a speaker or headphones, allowing the user to listen to a voice that faithfully reflects the user's emotional information, providing a more natural communication experience.

[1292] Example 2

[1293] 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."

[1294] Conventional speech conversion systems simply convert speech into another speech without taking into account the user's emotions, resulting in unnatural speech and poor emotional expression. Real-time speech conversion also entails delays, which can negatively impact the user experience. It is necessary to solve these problems and achieve more natural and emotionally rich speech conversion.

[1295] 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.

[1296] In this invention, the server includes means for converting transmitted voice data into a specified target voice, means for analyzing the voice data to recognize the user's emotional state, and means for adjusting the conversion model based on the recognized emotional state, thereby enabling natural voice conversion that reflects the user's emotions.

[1297] "User" refers to a person who uses the system to provide voice input.

[1298] "Terminal" refers to a device that acquires audio, performs preprocessing, and then transmits and receives the audio data to a server.

[1299] "Voice input" refers to voice data uttered by a user through a microphone of a terminal.

[1300] "Preprocessing" refers to the act of performing processes such as noise removal and volume adjustment on audio data to improve data quality.

[1301] "Server" refers to a device that receives voice data sent from a terminal and converts it into the voice of a specified target.

[1302] A "voice conversion model" refers to an algorithm that uses AI technology to convert voice data into another voice.

[1303] "Emotion engine" refers to an algorithm that analyzes and recognizes the user's emotional state from voice data.

[1304] "Adjusting the conversion model" refers to the act of adjusting the voice conversion model based on the analysis results of the emotion engine, and performing conversion that reflects emotions.

[1305] "Encoding" refers to the act of converting audio data into a playable format.

[1306] "Decoding" refers to the act of restoring encoded audio data to a playable format.

[1307] "Playback" refers to the act of allowing a user to listen to audio data through the terminal's speakers or headphones.

[1308] This invention is a system that converts a user's voice into another person's voice in real time, recognizes the user's emotions, and reflects them in the voice conversion. The following hardware and software are used to implement this system.

[1309] First, a device needs a microphone as hardware to acquire voice input. A user inputs voice by speaking into the microphone. For example, a user might say "hello." This voice input is captured by a voice capture application in the device.

[1310] Next, the device preprocesses the captured audio data by noise reduction and volume adjustment using a noise reduction software library (e.g., Audacity). This preprocessing improves the quality of the audio data and increases conversion accuracy.

[1311] The pre-processed voice data is sent to a server over the Internet, usually using the HTTP protocol. The server then inputs the received voice data into an AI voice conversion model (e.g., WaveNet) to convert it into a specified target voice. This model is based on a deep learning algorithm and analyzes the voice characteristics to perform the conversion.

[1312] The server is also equipped with an emotion engine that identifies the user's emotions from the received voice data. Software such as IBM Tone Analyzer is used for emotion recognition. The recognized emotions (e.g., anger, joy, etc.) are fed back to the voice conversion model, which is then adjusted according to the user's emotions.

[1313] The adjusted audio data is then sent from the server to the device. The server encodes the audio data into a playable format (e.g., OGG or MP3) and sends it to the device. The device then decodes the encoded audio data and converts it into a playable format using a playback library (e.g., FFmpeg).

[1314] Finally, the device plays the decoded voice data over a speaker or headphones, allowing the user to hear their own voice transformed into a different person's voice in real time, with the voice reflecting the user's emotions.

[1315] Specific examples

[1316] For example, if user A says "hello" in an angry voice, the voice is captured by the device's microphone. The device removes noise from the voice data and sends it to the server. The server processes the received voice with an AI voice conversion model (WaveNet) and uses an emotion engine (IBM Tone Analyzer) to recognize user A's emotion of "anger." The voice conversion model is then adjusted based on this emotion, and the voice is converted into user B's angry voice. This converted voice data is again sent from the server to the device, where it is played back. User A can hear user B's angry voice saying "hello" in real time, without any delay.

[1317] In this way, the system of the present invention effectively executes a series of processes from voice input acquisition to emotion recognition, conversion, and playback, providing users with an advanced real-time voice conversion experience.

[1318] Example prompt sentences to use

[1319] Example prompt sentence:

[1320] User A says "Hello" in an angry voice. The audio is captured, noise is removed, and then input into the speech conversion model. During conversion, the emotion recognition engine detects the emotion "anger" and converts it into the angry voice of User B. Play the converted audio.

[1321] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1322] Step 1:

[1323] Acquiring voice input

[1324] Input: The user's voice speaking into the device's microphone (e.g., "Hello")

[1325] Specific operation: The user speaks into the microphone of the terminal.

[1326] Data processing: The device uses an audio capture application to capture this audio as digital audio data.

[1327] Output: Captured audio data (e.g., WAV file with a sampling rate of 44.1 kHz)

[1328] Step 2:

[1329] Audio data preprocessing

[1330] Input: Captured audio data

[1331] Specific operation: The device uses a noise reduction software library (e.g., Audacity) to remove noise from the audio data and adjust the volume.

[1332] Data processing: Apply noise filtering algorithms to improve audio clarity.

[1333] Output: Preprocessed audio data (e.g., a denoised WAV file)

[1334] Step 3:

[1335] Sending audio data to the server

[1336] Input: Preprocessed audio data

[1337] Specific operation: The terminal sends audio data to the server using the HTTP protocol.

[1338] Data processing: The audio data is sent in the body of the HTTP request.

[1339] Output: Audio data received by the server

[1340] Step 4:

[1341] Processing with voice conversion models

[1342] Input: Audio data sent to the server

[1343] Specific operation: The server inputs voice data into an AI voice conversion model (e.g., WaveNet).

[1344] Data Computing: Using deep learning algorithms to extract voice features and convert them into the target voice.

[1345] Output: Converted voice data (e.g., a WAV file converted into a different person's voice)

[1346] Step 5:

[1347] Emotion determination using an emotion recognition engine

[1348] Input: Audio data sent to the server

[1349] Specific operation: The server inputs voice data into an emotion recognition engine (e.g., IBM Tone Analyzer).

[1350] Data processing: Analyze the voice data and extract emotional features.

[1351] Output: The user's emotional state (e.g., "anger" or "joy")

[1352] Step 6:

[1353] Tuning the speech conversion model

[1354] Input: Feedback from the emotion recognition engine (user's emotional state)

[1355] Specific operation: The server adjusts the voice conversion model based on the results of the emotion recognition engine.

[1356] Data calculation: Taking into account emotion features, the parameters of the conversion model are changed.

[1357] Output: A voice conversion model that reflects the user's emotions

[1358] Step 7:

[1359] Encoding the converted audio data

[1360] Input: Speech data converted by the trained speech conversion model

[1361] What happens: The server encodes the audio data into a playable format (e.g., OGG or MP3).

[1362] Data processing: Converting file formats using audio encoding algorithms.

[1363] Output: Encoded audio data (e.g. MP3 file)

[1364] Step 8:

[1365] Sending encoded audio data to the device

[1366] Input: Encoded audio data

[1367] Specific operation: The server uses the HTTP protocol to send encoded audio data to the terminal.

[1368] Data processing: The audio data is sent in the body of the HTTP response.

[1369] Output: The encoded audio data received by the device.

[1370] Step 9:

[1371] Decoding audio data

[1372] Input: Encoded audio data

[1373] Specific behavior: The device uses an audio playback library (e.g., FFmpeg) to decode the encoded audio data.

[1374] Data processing: Converting from an encoded format to a playable format.

[1375] Output: Decoded audio data (e.g. WAV file)

[1376] Step 10:

[1377] Playing audio data

[1378] Input: Decoded audio data

[1379] Specific operation: The device plays audio data using a playback device (speaker or headphones).

[1380] Data calculation: The decoded audio is converted from digital to analog and output.

[1381] Output: The audio the user hears (a modified voice, e.g., "Hello" with the emotion "Anger")

[1382] (Application example 2)

[1383] 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."

[1384] In conventional brick-and-mortar stores, it has been difficult to immediately grasp a customer's emotions and respond appropriately. Furthermore, if a customer is angry or happy, it is not possible to communicate with them in a voice that reflects their emotions, which can lead to a decline in the quality of customer service. By solving these problems, it is necessary to provide more effective customer service and improve customer satisfaction.

[1385] 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 acquiring voice input, means for preprocessing the acquired voice, means for transmitting the preprocessed voice data to the server, means for converting the transmitted voice data into another voice, means for receiving the converted voice data from the server, means for encoding the received voice data, means for playing the encoded voice, means for recognizing emotions and reflecting them in voice conversion, and means for adjusting voice tone based on the customer's emotions using an application installed on the smart device. This enables real-time voice conversion that reflects customer emotions in physical stores.

[1386] The "means for acquiring voice input" refers to a device or method for acquiring the voice spoken by the user as digital data using an acoustic sensor such as a microphone.

[1387] "Means for preprocessing captured audio" refers to a device or method that performs preprocessing such as noise removal and volume adjustment on captured audio data to improve the quality of the data and increase the accuracy of subsequent processing.

[1388] The "means for transmitting preprocessed voice data to a server" refers to a device or method for transmitting preprocessed voice data to a server via a communication line such as the Internet.

[1389] "Means for converting transmitted voice data into another voice" refers to a device or method for converting received voice data into another voice using an AI voice model or other conversion algorithm.

[1390] The "means for receiving converted voice data from the server" refers to a device or method for the server to send the converted voice data back to the client device via a communication line such as the Internet.

[1391] "Means for encoding received audio data" refers to a device or method for converting received audio data into a playable format (e.g., OGG or MP3 format).

[1392] "Means for playing encoded audio" refers to an apparatus or method for playing encoded audio data using a playback device (speakers or headphones).

[1393] "Means for recognizing emotions and reflecting them in voice conversion" refers to a device or method that analyzes the emotions from the user's voice, adjusts the voice conversion model based on the recognition results, and converts the voice into a different voice that reflects the emotions.

[1394] "Means for adjusting voice tone based on customer emotions using an application installed on a smart device" refers to a device or method that uses software installed on a device such as smart glasses or a smartphone to adjust voice tone in real time during customer service based on the customer's emotions.

[1395] The present invention provides a system that converts a user's voice into a different person's voice in real time, and further recognizes the user's emotions and reflects them in the voice conversion. Specific embodiments of the system are described below.

[1396] 1. System Configuration

[1397] This system is primarily composed of smart devices, internet communications, and servers. It includes the following elements:

[1398] 1. A device with a microphone to capture the user's voice (e.g., smart glasses or a smartphone)

[1399] 2. Dedicated software for preprocessing audio data (e.g., noise reduction, volume adjustment)

[1400] 3. A data communication module for transmitting the preprocessed audio data to a server via the Internet.

[1401] 4. AI voice conversion model for converting voice data into another voice on the server

[1402] 5. Emotion engine for analyzing and recognizing user emotions on the server

[1403] 6. An audio playback library to encode the converted audio data (e.g., OGG or MP3 format)

[1404] 7. Speakers or headphones to play the received encoded audio on your device.

[1405] 2. Program Processing

[1406] When a user speaks into the microphone, the device captures the voice and preprocesses it using dedicated software. This preprocessing includes noise reduction and volume adjustment. A data communications module is used to send the processed voice data to a server. The sent voice data is then input into an AI voice conversion model on the server and converted into a different voice.

[1407] Next, the emotion engine on the server analyzes the voice data and recognizes the user's emotions. Based on this recognition result, the AI ​​voice conversion model is re-adjusted and converted into voice that reflects the emotion. The converted voice data is sent back from the server to the device, where it is encoded by the voice playback library and played back through speakers or headphones.

[1408] 3. Example Use Case Scenarios

[1409] Below is a usage scenario for a physical store.

[1410] Scenario 1: The customer is angry

[1411] A salesperson wears smart glasses and interacts with a customer. If the customer is angry, the salesperson responds by inputting a prompt sentence as follows:

[1412] Prompt statement:

[1413] "When a customer is angry, change the salesperson's voice to a calmer one and speak in a more soothing manner."

[1414] Scenario 2: Happy Customers

[1415] In the situation where the customer is happy, the salesperson enters the following prompt sentence:

[1416] Prompt statement:

[1417] "When a customer is happy, the salesperson should change their voice to a more cheerful and upbeat tone."

[1418] By implementing the present invention in this way, real-time voice conversion that reflects customer emotions becomes possible in physical stores, thereby improving customer satisfaction.

[1419] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1420] Step 1:

[1421] A user speaks into the microphone of a smart device (e.g., smart glasses or a smartphone). This voice input is captured as digital data. The input is the user's speech, and the output is digital voice data. Specifically, the microphone converts sound waves into electrical signals, which are then digitized and captured as voice data.

[1422] Step 2:

[1423] The device performs preprocessing on the acquired voice data. Preprocessing includes noise reduction (removing noise from the voice data) and volume adjustment (adjusting the volume of the voice to an appropriate level). The input is digital voice data, and the output is preprocessed voice data. Specifically, the noise reduction algorithm removes unnecessary components from the voice data, improving the sound quality.

[1424] Step 3:

[1425] The device sends the preprocessed audio data to the server via the Internet. This is done using a data communication module. The input is the preprocessed audio data, and the output is the audio data received by the server. Specifically, the audio data is sent to the server using a network protocol (e.g., HTTP or WebSocket).

[1426] Step 4:

[1427] The server inputs the received voice data into an AI voice conversion model and converts it into another voice. The input is the voice data received on the server side, and the output is the converted voice data. Specifically, the generative AI model extracts voice features and converts them into the specified target voice.

[1428] Step 5:

[1429] The server runs an emotion engine that analyzes the voice data and recognizes the user's emotions. The input is the converted voice data, and the output is the user's emotional state. Specifically, the emotion recognition algorithm analyzes changes in voice tone and pitch to determine emotions.

[1430] Step 6:

[1431] The server readjusts the AI ​​voice conversion model based on the results of the emotion engine. The input is the user's emotional state, and the output is voice data readjusted to reflect the emotion. Specifically, the AI ​​model processes the voice data again and adds appropriate voice tone and emphasis.

[1432] Step 7:

[1433] The server encodes the converted audio data and converts it into a playable format (e.g., OGG or MP3 format). The input is the reconditioned audio data, and the output is the encoded audio data. Specifically, the encoding library converts the audio data into a playable format.

[1434] Step 8:

[1435] The terminal receives encoded voice data from the server. The input is the encoded voice data, and the output is the voice data received on the terminal side. Specifically, the received data is obtained from the network and stored in memory.

[1436] Step 9:

[1437] The device decodes the received audio data and plays it on speakers or headphones. The input is the encoded audio data, and the output is the audio to be played. Specifically, the device uses a decoding library to convert the audio data into a digital audio signal and plays it on the audio output device.

[1438] By going through this series of steps, the user's voice is converted in real time into a different person's voice that reflects their emotions, and can be used in situations such as physical stores.

[1439] 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.

[1440] 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.

[1441] 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.

[1442] 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.

[1443] FIG. 9 illustrates 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 behaviors 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.

[1444] 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.

[1445] 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).

[1446] 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.

[1447] 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."

[1448] 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.

[1449] 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).

[1450] 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.

[1451] 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.

[1452] 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.

[1453] 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.

[1454] 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.

[1455] 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.

[1456] 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.

[1457] 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.

[1458] 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.

[1459] 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.

[1460] The following is further disclosed regarding the above embodiment.

[1461] (Claim 1)

[1462] a means for obtaining a voice input;

[1463] means for preprocessing the captured audio;

[1464] means for transmitting the preprocessed audio data to a server;

[1465] means for converting the transmitted voice data into another voice;

[1466] means for receiving the converted voice data from the server;

[1467] means for encoding received audio data;

[1468] a means for playing the encoded audio;

[1469] A system including:

[1470] (Claim 2)

[1471] 10. The system of claim 1, wherein preprocessing the audio data includes noise removal.

[1472] (Claim 3)

[1473] The system of claim 1, which uses an AI voice model to convert voice data.

[1474] "Example 1"

[1475] (Claim 1)

[1476] means for acquiring a user's voice;

[1477] means for preprocessing the captured audio;

[1478] means for transmitting the preprocessed audio data to a server via a data communications network;

[1479] A means for converting the transmitted voice data into another voice using a deep learning model;

[1480] means for receiving the converted voice data from the server;

[1481] means for decoding the received audio data into a playable format;

[1482] means for playing the decoded audio on an audio output device;

[1483] A system including:

[1484] (Claim 2)

[1485] 10. The system of claim 1, wherein preprocessing the audio data includes noise removal.

[1486] (Claim 3)

[1487] 10. The system of claim 1, wherein a generative AI model is used to convert audio data.

[1488] "Application Example 1"

[1489] (Claim 1)

[1490] a means for obtaining a voice input;

[1491] means for preprocessing the captured audio;

[1492] means for transmitting the preprocessed audio data to a server;

[1493] means for converting the transmitted voice data into another voice;

[1494] means for receiving the converted voice data from the server;

[1495] means for encoding received audio data;

[1496] a means for playing the encoded audio;

[1497] a means for converting the user's voice into a different voice in real time;

[1498] A means for live streaming the converted audio data;

[1499] A means to add on-demand audio effects;

[1500] A system including:

[1501] (Claim 2)

[1502] 10. The system of claim 1, wherein preprocessing the audio data includes noise removal.

[1503] (Claim 3)

[1504] 10. The system of claim 1, wherein a generative AI model is used to convert audio data.

[1505] "Example 2: Combining Emotion Engines"

[1506] (Claim 1)

[1507] a means for obtaining voice input from a user;

[1508] means for preprocessing the captured audio by the terminal;

[1509] means for transmitting the preprocessed audio data to a server;

[1510] A means for the server to convert the transmitted voice data into a designated target voice;

[1511] a means for the server to analyze the voice data and recognize the emotional state of the user;

[1512] means for the server to adjust the transformation model based on the recognized emotional state;

[1513] means for receiving the converted voice data from the server;

[1514] means for decoding the received encoded audio data;

[1515] a means for playing the encoded audio;

[1516] A system including:

[1517] (Claim 2)

[1518] 10. The system of claim 1, wherein preprocessing the audio data includes noise removal.

[1519] (Claim 3)

[1520] 10. The system of claim 1, wherein a generative AI model is used to convert audio data.

[1521] "Application example 2 when combining emotion engines"

[1522] (Claim 1)

[1523] a means for obtaining a voice input;

[1524] means for preprocessing the captured audio;

[1525] means for transmitting the preprocessed audio data to a server;

[1526] means for converting the transmitted voice data into another voice;

[1527] means for receiving the converted voice data from the server;

[1528] means for encoding received audio data;

[1529] a means for playing the encoded audio;

[1530] A means for recognizing emotions and reflecting them in speech conversion;

[1531] A means for adjusting the tone of voice based on the customer's emotions using an application installed on a smart device;

[1532] A system including:

[1533] (Claim 2)

[1534] 10. The system of claim 1, wherein preprocessing the audio data includes noise removal.

[1535] (Claim 3)

[1536] The system of claim 1, which uses an AI voice model to convert voice data. [Explanation of symbols]

[1537] 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. a means for obtaining a voice input; means for preprocessing the captured audio; means for transmitting the preprocessed audio data to a server; means for converting the transmitted voice data into another voice; means for receiving the converted voice data from the server; means for encoding received audio data; a means for playing the encoded audio; A system including:

2. 2. The system of claim 1, wherein pre-processing the audio data includes noise removal.

3. 2. The system of claim 1, wherein an AI voice model is used to convert voice data.

Images