System

Abstract

A system is provided.SOLUTION: A system comprising: means for capturing input speech to generate speech data; means for transmitting the speech data to a server; means for analyzing the speech data at the server to perform speech recognition and extract speech characteristics; means for comparing the speech characteristics with reference speech characteristics to identify pronunciation errors; means for generating feedback for pronunciation improvement based on the pronunciation errors; and means for transmitting the feedback to a terminal for display to a user.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] Accurate pronunciation is important, especially in foreign language learning. However, many learners have difficulty mastering correct pronunciation because they lack a way to instantly determine whether their pronunciation is correct. Traditional pronunciation training methods, which require textbooks, tapes, or even individual instruction, can be costly, time-consuming, and impractical. Therefore, there is a need for a system that can analyze pronunciation and provide instant feedback. [Means for solving the problem]

[0005] To solve this problem, the present invention provides the following means. First, the system includes means for capturing speech input by a user and generating speech data. Next, the system includes means for transmitting this speech data to a server. The server includes means for analyzing the received speech data, performing speech recognition, and extracting speech characteristics. The server then includes means for comparing the extracted speech characteristics with reference speech characteristics and identifying pronunciation errors. The system further includes means for generating feedback for pronunciation improvement based on the identified pronunciation errors. Finally, the system provides a system including means for transmitting the generated feedback to a terminal and displaying it to the user. This system allows users to immediately identify problems with their pronunciation and make effective improvements.

[0006] "Voice data" refers to a digital recording of a user's pronunciation.

[0007] A "server" is a computing device that analyzes voice data, performs voice recognition, comparisons, and generates feedback.

[0008] A "terminal" is a device that has a microphone for inputting voice, communicates with a server, and displays feedback to the user.

[0009] "Speech recognition" is a technology that converts input voice data into text.

[0010] "Audio characteristics" refer to characteristics such as pitch, intensity, and tempo of audio data.

[0011] "Reference speech characteristics" refer to speech characteristics that are set as correct pronunciations.

[0012] A "mispronunciation" is a portion of the input speech data that does not match the reference speech characteristics.

[0013] "Feedback" refers to instructions or advice, including information for correcting pronunciation errors.

[0014] "Visual feedback" refers to visual information such as diagrams or videos to help correct pronunciation.

[0015] "Text feedback" is written information to help correct pronunciation.

[0016] "Audio feedback" refers to audio samples that reproduce the correct pronunciation.

[0017] "Noise reduction" is a technique for removing unnecessary background sounds from audio data.

[0018] "Normalization" is the process of adjusting the volume level of audio data to a consistent standard.

[0019] "Volume adjustment" refers to the process of changing the volume of audio data to an appropriate level. [Brief explanation of the drawings]

[0020] [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

[0021] 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.

[0022] First, the terms used in the following description will be explained.

[0023] 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).

[0024] 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.

[0025] 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.

[0026] 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.

[0027] 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."

[0028] [First embodiment]

[0029] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0030] 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.

[0031] 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).

[0032] 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.

[0033] 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.

[0034] 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.

[0035] 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.

[0036] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0037] 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.

[0038] 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.

[0039] 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.

[0040] 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."

[0041] This system analyzes the user's input speech in real time and provides feedback to help the user master correct pronunciation. A specific embodiment of this system will be described below.

[0042] Audio input and capture

[0043] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[0044] Generate audio data and send it to the server

[0045] The device captures the user's speech, stores it locally as digital audio data, converts it into an appropriate format (e.g., WAV, MP3), and then transmits it to the server.

[0046] Voice analysis and speech recognition

[0047] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[0048] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[0049] Compare pronunciations and identify errors

[0050] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0051] Generate feedback

[0052] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0053] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0054] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0055] Audio feedback: Includes audio samples of correct pronunciation.

[0056] Sending and Viewing Feedback

[0057] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0058] This system allows people to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention will be a means to significantly improve the efficiency of phonetic learning.

[0059] The processing flow will be explained below.

[0060] Step 1:

[0061] User: Speak into the device's microphone. For example, say "Hello."

[0062] Step 2:

[0063] Device: Captures the user's speech and stores it locally as digital audio data.

[0064] Step 3:

[0065] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[0066] Step 4:

[0067] Terminal: Sends the converted audio data to the server.

[0068] Step 5:

[0069] Server: Receives the voice data sent from the terminal.

[0070] Step 6:

[0071] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[0072] Step 7:

[0073] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[0074] Step 8:

[0075] Server: Compares the recognized text with reference pronunciation data.

[0076] Step 9:

[0077] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[0078] Step 10:

[0079] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors. The types of feedback include visual feedback, text feedback, and audio feedback.

[0080] Step 11:

[0081] Server: Sends the generated feedback to the device.

[0082] Step 12:

[0083] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[0084] Step 13:

[0085] User: Based on the feedback provided, they correct and try again, which returns them to step 1 and repeats the process to improve their pronunciation.

[0086] Example 1

[0087] 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."

[0088] Conventional speech recognition systems lack the ability to provide specific feedback to help users master correct pronunciation. Furthermore, they lack the means to suggest specific correction methods for specific pronunciation errors, making it difficult for users to effectively correct their pronunciation.

[0089] 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.

[0090] In this invention, the server includes means for capturing input speech and generating speech data, means for transmitting the speech data to the server, means for analyzing the speech data, performing speech recognition, and extracting speech characteristics in the server, means for comparing the speech characteristics with reference speech characteristics and identifying pronunciation errors, means for generating visual feedback, text feedback, and speech feedback for pronunciation improvement based on the pronunciation errors, and means for transmitting the feedback to a terminal and displaying it to the user, thereby enabling the user to efficiently correct their pronunciation.

[0091] "Input voice" refers to the voice spoken by the user into the microphone.

[0092] "Audio Data" refers to data for storing or transmitting captured audio in digital form.

[0093] "Server" refers to a computer system that analyzes voice data, recognizes voice, and performs post-voice recognition processing.

[0094] "Speech recognition" refers to the process of converting input voice data into character string information.

[0095] "Audio characteristics" refers to characteristic information such as pitch, intensity, and tempo contained in audio data.

[0096] "Reference speech characteristics" refers to speech characteristics information collected and recorded based on correct pronunciation.

[0097] "Mispronunciation" refers to a mispronunciation portion identified by comparing the characteristics of the input speech with the reference speech characteristics.

[0098] "Visual feedback" refers to feedback in the form of diagrams or videos that show lip and tongue shape, mouth opening, etc. to help users improve their pronunciation.

[0099] "Textual feedback" refers to written feedback that includes specific instructions to improve a user's pronunciation.

[0100] "Audio feedback" refers to feedback that provides an audio sample of correct pronunciation.

[0101] "Terminal" refers to a device (smartphone, PC, etc.) through which a user inputs voice.

[0102] "Preprocessing" refers to performing processes such as noise removal, normalization, and volume adjustment on audio data.

[0103] This invention is a system that analyzes user-input speech in real time and provides feedback to help users learn correct pronunciation.

[0104] A specific embodiment of this system is described below: A user speaks a target word or sentence into the microphone of the terminal. For example, if a user wants to learn the word "Hello," he speaks "Hello" into the terminal.

[0105] Audio input and capture

[0106] The user speaks into the device's microphone. The device's microphone captures what the user says and stores it locally as digital audio data. The audio data is converted to an appropriate format (e.g., WAV, MP3) using an audio library (e.g., PyDub).

[0107] Generate audio data and send it to the server

[0108] The device sends the saved audio data to the server using the HTTP protocol.

[0109] Voice analysis and speech recognition

[0110] The server receives the audio data sent from the device. After receiving the data, it first performs preprocessing on the audio data. This includes noise removal, normalization, and volume adjustment. An audio signal processing library (e.g., LibROSA) is used for this processing.

[0111] Speech Recognition and Feature Extraction

[0112] After preprocessing, the server performs speech recognition on the speech data and converts the speech into text. A speech recognition API (e.g., Google® Cloud Speech-to-Text API) is used for speech recognition. At the same time, speech characteristics (pitch, intensity, tempo, etc.) are extracted. LibROSA is used for feature extraction.

[0113] Compare pronunciations and identify errors

[0114] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0115] Generate feedback

[0116] The server generates feedback to help the user correct their pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0117] Visual feedback: Diagrams and videos showing lip and tongue shape and mouth opening, using video editing software (e.g., Adobe After Effects, Blender).

[0118] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0119] Audio feedback: Includes audio samples of correct pronunciation.

[0120] Sending and Viewing Feedback

[0121] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0122] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention is a means of significantly improving the efficiency of phonetic learning.

[0123] Example 1

[0124] The user says "Hello" and the device's microphone captures the audio. The device converts the captured audio to WAV format and sends it to the server. The server receives the audio data and performs noise reduction and normalization. The server uses the Google Cloud Speech-to-Text API to convert "Hello" to text and simultaneously calculates pitch and intensity. The server compares "Hello" with standard pronunciation data and determines that the "e" sound is closer to a "u." The server generates text feedback such as "Place your tongue behind your upper teeth to pronounce 'L'" along with an audio sample of the correct pronunciation and sends it to the device. The device presents this to the user, who then practices their pronunciation again based on the feedback.

[0125] Prompt Sentence Examples

[0126] "The user said 'Hello'. Given this audio data, please compare it with a standard pronunciation to generate correct feedback."

[0127] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0128] Step 1: Capturing Audio Input

[0129] The user speaks into the microphone of the terminal.

[0130] Input: User's voice, saying "Hello".

[0131] The device captures the spoken voice through a microphone and stores it as digital voice data.

[0132] Output: Captured digital audio data (WAV, MP3 format, etc.).

[0133] What it does: The device's audio API captures audio input from the microphone and saves it as an audio file.

[0134] Step 2: Generate and send audio data

[0135] The terminal transmits the stored voice data to the server.

[0136] Input: Captured digital audio data.

[0137] The terminal uses the HTTP protocol to send the audio data to the server.

[0138] Output: The audio data sent to the server.

[0139] What it does: The device converts the voice data into the appropriate format and sends the data to a server over the Internet.

[0140] Step 3: Preprocessing the audio data

[0141] The server receives the voice data transmitted from the terminal.

[0142] Input: Audio data sent from the device.

[0143] The server performs preprocessing such as noise removal, normalization, and volume adjustment on the received audio data.

[0144] Output: Preprocessed audio data.

[0145] What it does: A server-side audio signal processing library (e.g., LibROSA) removes noise from the audio data and normalizes the volume level to adjust it to an appropriate sound quality.

[0146] Step 4: Speech recognition and feature extraction

[0147] The server analyzes the preprocessed audio data.

[0148] Input: Preprocessed audio data.

[0149] The server performs speech recognition and converts the audio data into text, while simultaneously extracting audio characteristics (pitch, intensity, tempo, etc.).

[0150] Output: Text data and speech characteristics data generated by speech recognition.

[0151] Specific operation: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) to convert the speech into text, and calculates speech feature information using a library such as LibROSA.

[0152] Step 5: Compare pronunciation and identify errors

[0153] The server compares the recognized text with the reference pronunciation data.

[0154] Input: Text data generated by speech recognition, reference pronunciation data.

[0155] The server identifies which parts of the user's pronunciation are incorrect.

[0156] Output: Mispronunciation information.

[0157] Specific operation: The server refers to a dataset of correct pronunciations collected in advance, compares it with the user's pronunciation, identifies incorrect parts, and records them.

[0158] Step 6: Generate feedback

[0159] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[0160] Input: Mispronunciation information.

[0161] The server generates visual, textual and audio feedback.

[0162] Output: Feedback data (visual, text, audio).

[0163] Specific behaviors: Visual feedback is generated using video editing software (e.g., Adobe After Effects), text feedback is written with specific instructions, and audio feedback uses pre-recorded samples of correct pronunciation.

[0164] Step 7: Submit and view feedback

[0165] The server transmits the generated feedback to the terminal.

[0166] Input: Feedback data (visual, text, audio).

[0167] The device analyzes the received feedback and provides it to the user visually and audibly.

[0168] Output: Feedback provided to the user.

[0169] Specific operation: The device presents video, text, and audio data through a user interface, and the user corrects their pronunciation based on the feedback.

[0170] (Application example 1)

[0171] 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."

[0172] Conventional speech analysis systems have primarily been used for personal pronunciation practice. However, effective means for improving the quality of speech communication between store clerks and customers in brick-and-mortar stores are limited. This results in missed opportunities for improving customer satisfaction and improving store clerk skills. The present invention aims to solve this problem by providing a system that evaluates the pronunciation quality of store clerk interactions in real time and provides effective feedback.

[0173] 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.

[0174] In this invention, the server includes means for capturing and analyzing voice data, means for performing voice recognition and extracting voice characteristics, means for comparing the extracted voice characteristics with reference voice characteristics to identify pronunciation errors, means for generating feedback for pronunciation improvement based on the pronunciation errors, means for transmitting the feedback to a terminal and displaying it to the user as visual, text, and voice feedback, and means for analyzing voices spoken by store clerks when interacting with customers and providing feedback, thereby enabling real-time pronunciation improvement so that store clerks can communicate with customers with appropriate pronunciation.

[0175] "Capture" means acquiring information such as sound or video and saving it as digital data.

[0176] "Audio data" is data that represents audio in digital form.

[0177] A "server" is a centralized management system for processing and storing data.

[0178] "Speech recognition" is the technology that analyzes voice data and converts the content into text or other formats.

[0179] "Audio characteristics" are characteristic attributes associated with an audio signal, including pitch, intensity, tempo, etc.

[0180] "Reference speech characteristics" are reference data used to indicate correct pronunciation.

[0181] A "mispronunciation" is a difference observed between the user's pronunciation and the reference pronunciation.

[0182] "Feedback" is information that informs the user of the evaluation results and areas for improvement.

[0183] A "terminal" is a computing device that can be directly operated by a user.

[0184] A "store associate" refers to someone who provides service to customers in a physical store.

[0185] "Customer service" refers to a series of actions and statements that store clerks make to communicate with customers.

[0186] This system analyzes in real time the speech spoken by store clerks when interacting with customers and provides feedback to help improve pronunciation. A specific embodiment of this system will be described below.

[0187] Audio input and capture

[0188] The user (store clerk) speaks into the microphone. For example, the user speaks the phrase "Hello, how can I help you?"

[0189] Generate audio data and send it to the server

[0190] The device captures the user's speech and stores it as digital audio data, which is then converted to WAV or MP3 format and sent to a server.

[0191] Voice analysis and speech recognition

[0192] After receiving the voice data from the device, the server performs preprocessing on the voice data, specifically noise removal, normalization, and volume adjustment. After preprocessing is complete, the server performs speech recognition on the voice data and converts the voice into text. At the same time, it extracts voice characteristics (pitch, intensity, tempo, etc.).

[0193] Compare pronunciations and identify errors

[0194] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0195] Generate feedback

[0196] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0197] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0198] Text feedback: Information in written form with specific instructions, such as "Try to pronounce the letter 'L' with your tongue behind your upper teeth."

[0199] Audio feedback: Includes audio samples of correct pronunciation.

[0200] Sending and Viewing Feedback

[0201] The server sends the generated feedback to the device, which then analyzes the received feedback and provides it to the user visually and audibly, allowing the user to understand the problems with their pronunciation and immediately try to correct them.

[0202] As a concrete example, consider a user saying the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it, identifies pronunciation errors, and provides feedback to help improve. For example, if the system determines that the "o" sound in "Hello" is not pronounced correctly, it immediately provides corrective feedback to the user.

[0203] Also, examples of prompts to use this system effectively include:

[0204] "Please say the following English sentence: 'Hello, how can I help you?'"

[0205] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[0206] It can be provided as such.

[0207] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0208] Step 1:

[0209] The user speaks the target phrase or word into the microphone.

[0210] Input: User's voice

[0211] Output: Analog audio signal

[0212] In this step, audio is captured as the user speaks, for example, "Hello, how can I help you?"

[0213] Step 2:

[0214] The terminal converts the captured analog audio signal into digital audio data and stores it locally.

[0215] Input: Analog audio signal

[0216] Output: Digital audio data (WAV or MP3 format)

[0217] The device converts the analog audio signal received through the microphone into digital format and saves it in WAV or MP3 format.

[0218] Step 3:

[0219] The terminal transmits the converted digital audio data to the server.

[0220] Input: Digital audio data (WAV or MP3 format)

[0221] Output: Audio data sent to the server

[0222] The terminal uploads the voice data to the server via the network.

[0223] Step 4:

[0224] The server preprocesses the received audio data, performing noise removal, normalization, and volume adjustment.

[0225] Input: Transmitted digital audio data

[0226] Output: Preprocessed audio data

[0227] The server cleans the audio data by applying noise reduction filters, normalizing the volume, etc.

[0228] Step 5:

[0229] The server performs speech recognition on the preprocessed audio data, converting it to text and simultaneously extracting audio characteristics (pitch, intensity, tempo).

[0230] Input: Preprocessed audio data

[0231] Output: Text data and voice characteristics data

[0232] The server uses speech recognition algorithms to analyze the audio and convert it to text, while also extracting audio characteristics such as pitch and intensity.

[0233] Step 6:

[0234] The server compares the extracted speech characteristics with reference speech characteristics to identify mispronunciations.

[0235] Input: Voice characteristic data and reference voice characteristic data

[0236] Output: Mispronunciation information

[0237] The server compares and analyzes the reference voice data with the user's voice characteristics to identify incorrect pronunciations.

[0238] Step 7:

[0239] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[0240] Input: Mispronunciation information

[0241] Output: Feedback information (visual, text, audio)

[0242] The server generates specific feedback to help improve mispronunciations, including diagrams, detailed instructions, and audio samples of correct pronunciation.

[0243] Step 8:

[0244] The server transmits the generated feedback to the terminal, which displays the feedback to the user.

[0245] Input: Feedback information

[0246] Output: Visual and auditory feedback to the user

[0247] The feedback information sent from the server to the device is analyzed and displayed on the device. The user receives feedback from the screen and speaker and uses it to improve their pronunciation.

[0248] As a concrete example, consider a user uttering the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it to identify pronunciation errors, and provides real-time feedback for improvement.

[0249] An example prompt is:

[0250] "Please say the following English sentence: 'Hello, how can I help you?'"

[0251] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[0252] 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.

[0253] This system analyzes the user's voice input in real time and provides feedback to help them master correct pronunciation. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions and provides appropriate feedback according to the user's emotions.

[0254] Audio input and capture

[0255] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[0256] Generate audio data and send it to the server

[0257] The device captures the user's speech and stores it locally as digital audio data, which is then converted into an appropriate format (e.g., WAV, MP3) and sent to the server.

[0258] Voice analysis and speech recognition

[0259] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[0260] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[0261] Compare pronunciations and identify errors

[0262] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0263] Emotion recognition

[0264] The server uses an emotion engine to recognize emotions from the user's voice data. The emotion engine analyzes voice characteristics (e.g., pitch, intensity, rate) to identify the user's emotional state (e.g., happy, sad, anger).

[0265] Generate feedback

[0266] The server generates feedback based on the identified pronunciation errors and the user's emotional state. The feedback is provided in the following format:

[0267] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0268] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0269] Audio feedback: Includes audio samples of correct pronunciation.

[0270] If the emotion engine detects negative emotions (e.g., frustration, sadness), the server generates feedback with encouraging messages, such as "Great progress! Try a little harder next time!"

[0271] Sending and Viewing Feedback

[0272] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and their emotional state and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0273] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. Furthermore, by providing feedback that takes into account the user's emotions, it helps maintain motivation. This invention will significantly improve the efficiency and effectiveness of phonetic learning.

[0274] The processing flow will be explained below.

[0275] Step 1:

[0276] User: Speak into the device's microphone. For example, if you want to learn the word "Hello," speak "Hello" into the device.

[0277] Step 2:

[0278] Device: Captures the user's speech and stores it locally as digital audio data.

[0279] Step 3:

[0280] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[0281] Step 4:

[0282] Terminal: Sends the converted audio data to the server.

[0283] Step 5:

[0284] Server: Receives the voice data sent from the terminal.

[0285] Step 6:

[0286] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[0287] Step 7:

[0288] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[0289] Step 8:

[0290] Server: Compares the recognized text with reference pronunciation data.

[0291] Step 9:

[0292] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[0293] Step 10:

[0294] Server: Inputs the user's voice data into the emotion engine and recognizes the user's emotions (e.g., joy, sadness, anger).

[0295] Step 11:

[0296] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors and the user's emotional state. For example, it generates feedback including instructions to correct the "e" sound in "Hello" and an encouraging message if the user is showing negative emotions.

[0297] Step 12:

[0298] Server: Sends the generated feedback to the device.

[0299] Step 13:

[0300] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[0301] Step 14:

[0302] User: Attempts to pronounce again based on the feedback provided, thereby repeating the process and improving pronunciation.

[0303] For example, if a user says "Hello" and the server determines that the "e" sound is close to "u," and the emotion engine recognizes that the user is frustrated, the system will provide feedback such as, "Open your lips a little and lower your tongue a little. Then try again. That's great progress!" In this way, the system effectively supports pronunciation correction while taking the user's emotions into consideration.

[0304] Example 2

[0305] 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."

[0306] In learning pronunciation, it is difficult for users to accurately understand and correct their pronunciation errors. Maintaining motivation during the learning process is also a challenge. Furthermore, there is currently a lack of systems that provide appropriate feedback based on the user's emotional state.

[0307] 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.

[0308] In this invention, the server includes means for capturing input voice and generating voice data, means for transmitting the voice data to the server, and means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server, thereby enabling the server to analyze the voice data, identify pronunciation errors, and further recognize emotions.

[0309] "Means for capturing input voice and generating voice data" refers to the process of capturing voice spoken by a user in real time through the device's microphone and saving it as digital voice data.

[0310] The "means for transmitting the audio data to the server" refers to a method for transmitting digital audio data stored in the terminal to the server via a network.

[0311] "Means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server" refers to the function of performing voice recognition on the voice data received by the server, and analyzing and extracting characteristics such as pitch, intensity, and tempo from the voice data.

[0312] The "means for comparing the characteristics of the voice with reference voice characteristics and identifying pronunciation errors" refers to a process for comparing the characteristics of the analyzed voice with reference correct voice characteristics and identifying the errors.

[0313] "Means for generating feedback for pronunciation improvement based on the pronunciation errors" refers to a method for generating feedback including advice and instructions for the user to pronounce correctly based on the identified pronunciation errors.

[0314] The "means for transmitting the feedback to the terminal and displaying it to the user" refers to a method for transmitting the generated feedback from the server to the terminal and presenting it to the user visually or audibly at the terminal.

[0315] "Means for recognizing emotions from the user's voice and generating appropriate feedback based on the recognized emotions" refers to a process of analyzing the user's voice characteristics to recognize their emotional state and creating feedback according to that emotion.

[0316] This invention is a system that supports users in learning pronunciation in real time. It mainly involves a series of processes: voice input, voice data generation and transmission, voice recognition and analysis, pronunciation comparison and error identification, emotion recognition, and feedback generation and transmission.

[0317] Voice input

[0318] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn "Hello," they speak "Hello" into the device. The device captures this speech in real time.

[0319] Generates audio data and sends it to the server

[0320] The device stores the captured audio as digital data and converts it to an appropriate format, such as WAV or MP3, and transmits the converted audio data over the network to a server, along with metadata such as the user ID and a timestamp.

[0321] Voice analysis and speech recognition

[0322] The server processes the audio data received from the device. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. This preprocessing is a fundamental step to obtain clean audio data. Next, it converts the audio data into text using a speech recognition engine (e.g., Google Cloud Speech-to-Text API). In addition, it analyzes and extracts audio characteristics (e.g., pitch, intensity, tempo).

[0323] Compare pronunciations and identify errors

[0324] The server compares the parsed text with a reference pronunciation database containing pronunciations from native speakers. The comparison identifies any pronunciation errors the user may have made. For example, if the "e" in "Hello" sounds more like a "u," that part is recorded as an error.

[0325] Emotion recognition

[0326] The server analyzes the voice data through an emotion recognition engine (e.g., IBM Watson® Tone Analyzer) to identify the user's emotional state. It analyzes voice characteristics such as pitch, intensity, and rate to recognize whether the user is happy, sad, or angry.

[0327] Generate feedback

[0328] The server generates feedback based on the identified pronunciation errors and the user's emotional state, and the feedback is provided in visual, text, and audio formats.

[0329] Visual feedback: Diagrams or videos showing lip and tongue position and mouth opening (e.g., educational video link).

[0330] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your top teeth to pronounce the letter 'L'").

[0331] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker).

[0332] If the emotion recognition engine detects a negative emotion, the server adds an encouraging message, such as "Great progress! Try a little harder next time!"

[0333] Sending and Viewing Feedback

[0334] The server sends the generated feedback to the device, which analyzes it and provides it visually and audibly to the user, allowing the user to understand their pronunciation problems and try to correct them in real time.

[0335] As a concrete example, if a user wants to learn the phrase "Thank you," the entire process is carried out sequentially, starting with voice input, followed by generation and transmission of voice data, analysis and recognition, identification of pronunciation errors, emotion recognition, feedback generation, and finally transmission and display of the feedback.

[0336] Example prompts for generative AI models

[0337] Explain the steps to analyze the voice data of a user saying "Thank you" into the device's microphone, recognize pronunciation errors and the user's emotional state, and generate and display appropriate feedback.

[0338] This system allows you to effectively learn correct pronunciation without the need for special equipment or expensive lessons. It also provides emotionally sensitive feedback, making it easier to maintain motivation. This invention is a means to significantly improve the efficiency and effectiveness of phonetic learning.

[0339] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0340] Step 1: Audio Input and Capture

[0341] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn the word "Hello," they speak "Hello." This is the user's voice input. The device captures this voice in real time and saves it as digital voice data. The input is the user's voice, and the output is digital voice data. At this stage, voice capture software is used.

[0342] Step 2: Generate audio data and send it to the server

[0343] The device converts the captured audio into an appropriate format (e.g., WAV, MP3). This converted digital audio data is then sent to the server. The input is the captured audio data, and the output is the converted audio data. The converted audio also contains metadata such as the user ID and a timestamp. Audio format conversion and network communication software is used at this stage.

[0344] Step 3: Voice analysis and speech recognition

[0345] The server analyzes the received audio data. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. Then, it uses a speech recognition engine to convert the audio data to text. At the same time, it analyzes and extracts audio characteristics (pitch, intensity, tempo, etc.). The input is the formatted audio data, and the output is text data and audio characteristic data. At this stage, audio processing and speech recognition software is used. Preprocessing includes noise filtering algorithms and signal normalization algorithms.

[0346] Step 4: Compare pronunciation and identify errors

[0347] The server compares the parsed text to a reference pronunciation database. This reference data contains pronunciations from native speakers. The comparison identifies pronunciation errors. For example, if the "e" in "Hello" is close to a "u," it is marked as an error. The input is text data and voice characteristics data, and the output is data indicating pronunciation errors. This stage involves comparison algorithms and database search software.

[0348] Step 5: Recognize emotions

[0349] The server uses the voice data to recognize the user's emotions. An emotion recognition engine is used to analyze voice characteristics (pitch, intensity, speed, etc.) to identify the user's emotional state. For example, emotions such as joy, sadness, and anger can be identified from the analysis results. The input is the voice characteristic data, and the output is the recognized emotional state. At this stage, an emotion recognition algorithm is used.

[0350] Step 6: Generate feedback

[0351] The server generates feedback based on the identified pronunciation errors and the recognized emotional state. The feedback is provided in the following form:

[0352] Visual feedback: illustrations and videos showing lip and tongue position and mouth opening

[0353] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your upper teeth to pronounce the letter 'L'")

[0354] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker)

[0355] If the emotion recognition engine detects a negative emotion, it adds an encouraging message. The input is the pronunciation error data and the emotional state data, and the output is the generated feedback. At this stage, a feedback generation algorithm is used.

[0356] Step 7: Submit and view feedback

[0357] The server sends the generated feedback to the device. The device receives this feedback and displays it visually and audibly to the user. For example, an on-screen diagram showing the correct lip and tongue positions and an audio playback of the correct pronunciation. The input is the generated feedback, and the output is the visual and audible information presented to the user. This is where communication software and user interface software come into play.

[0358] Through these steps, the system can efficiently improve the user's pronunciation and maintain their motivation to learn.

[0359] (Application example 2)

[0360] 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."

[0361] The present invention relates to a system that improves the efficiency of a user's pronunciation practice and provides appropriate feedback according to the user's emotional state. Conventional pronunciation learning systems can identify pronunciation errors, but they cannot take the user's emotions or state into account, making it difficult to maintain motivation for learning. Therefore, in order to maximize the effectiveness of pronunciation learning, a system that provides feedback according to the user's emotions is needed.

[0362] The identification process by the identification 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 including an emotion engine for recognizing emotions from voice data, means for generating feedback according to the emotions, and means for performing preprocessing such as noise removal, normalization, and volume adjustment. This makes it possible to identify errors in the user's voice pronunciation and provide feedback according to the user's emotional state.

[0363] A "means for generating voice data" is a device or software that allows a user to capture voice input and store it as voice data in digital form.

[0364] The "means for transmitting voice data to a server" refers to a device or software for transferring the generated voice data to a server via a network.

[0365] "Means for analyzing audio data, performing speech recognition, and extracting audio characteristics" refers to devices or software that apply a speech recognition algorithm to audio data within a server and analyze audio characteristics (e.g., pitch, intensity, tempo, etc.).

[0366] The "means for comparing voice characteristics with reference voice characteristics and identifying pronunciation errors" refers to a device or software that compares the user's voice characteristics with preset reference voice characteristics and identifies incorrect pronunciation parts.

[0367] A "means for generating feedback for pronunciation improvement" is a device or software that generates specific instructions or samples for correcting pronunciation errors.

[0368] The "means for transmitting feedback to a terminal and displaying it to a user" refers to a device or software for transmitting the generated feedback to a user's terminal via a network and displaying it.

[0369] The "means including an emotion engine for recognizing emotions" refers to a device or software for analyzing emotions from voice data and identifying the emotional state of a user.

[0370] The "means for generating emotional feedback" refers to a device or software for generating feedback including appropriate encouragement, warnings, and specific instructions based on the user's emotional state.

[0371] The present invention relates to a system that identifies user pronunciation errors and provides real-time emotional feedback. The system generates speech data, transmits it to a server, and analyzes it to perform speech recognition and extract speech characteristics. Furthermore, the system uses an emotion engine to recognize the user's emotions and generate specific feedback accordingly.

[0372] Processing contents to be performed by the program

[0373] The invention uses a "device," such as a microphone or smartphone, to capture a user's voice input. The device stores the captured audio as digital audio data and then transmits this data to a server, where it is converted into an appropriate format (e.g., WAV or MP3).

[0374] The server analyzes the audio data using an installed speech recognition engine (e.g., Google Speech-to-Text API) to convert the speech to text and extracts voice characteristics (e.g., pitch, intensity, tempo). This data is compared with baseline voice characteristics to identify pronunciation errors.

[0375] Furthermore, an emotion recognition engine (e.g., IBM Watson Tone Analyzer) is used to analyze the user's emotions from the voice data, identifying the user's emotional state (e.g., joy, sadness, anger, etc.) and generating feedback accordingly.

[0376] Providing Feedback

[0377] The form of feedback includes at least one of visual feedback, text feedback, and audio feedback. Visual feedback may include animations or videos showing lip movements and tongue positions that indicate correct pronunciation. Text feedback may include specific instructions such as "Put your tongue behind your upper teeth to pronounce 'L'." Audio feedback may include audio samples of correct pronunciation.

[0378] Specific examples

[0379] For example, suppose a factory operator says "start" when operating a robot. This voice input is captured by the device's microphone and sent to a server. The server analyzes the speech of "start" and detects that a specific part (such as the "ta" sound) is not pronounced correctly. It then uses an emotion recognition engine to determine that the operator is slightly nervous. As a result, the system displays an encouraging message along with animations and voice samples showing examples of how to pronounce "ta" correctly.

[0380] Prompt Sentence Examples

[0381] Please tell us about the voice-guided operation support system in factories.

[0382] This system helps factory operators give accurate voice instructions to robots and provides feedback according to the operator's emotional state. It includes steps for voice recognition, pronunciation analysis, emotion recognition, and feedback generation.

[0383] This makes it easier for users to learn correct pronunciation, maintains their motivation to learn, and improves factory operation efficiency.

[0384] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0385] Step 1:

[0386] (audio input and capture)

[0387] The user speaks into a microphone. For example, a factory operator says "start." This voice is captured by the device's microphone and saved as digital voice data. The input is real-time voice data, and the output is saved digital voice data. Specifically, a microphone and voice capture software installed on a smartphone or dedicated device is used.

[0388] Step 2:

[0389] (Sending audio data)

[0390] The device converts the captured digital audio data into an appropriate format (e.g., WAV, MP3) and transmits it to a server via the Internet. The input is the stored digital audio data, and the output is the audio data transmitted to the server. Specifically, a communication module and a data transfer protocol are used.

[0391] Step 3:

[0392] (Audio pre-processing)

[0393] The server performs preprocessing on the received audio data. Specifically, it performs noise removal, normalization, volume adjustment, etc. The input is the audio data sent to the server, and the output is the audio data after preprocessing. As a specific example, an audio processing library (e.g., WebRTC) is used.

[0394] Step 4:

[0395] (Voice Recognition)

[0396] The server performs speech recognition on the preprocessed audio data and converts the audio into text. The input is the preprocessed audio data, and the output is text generated from the audio. Specifically, the server uses the Google Speech-to-Text API or similar.

[0397] Step 5:

[0398] (Extraction of voice characteristics)

[0399] The server extracts audio features (e.g., pitch, intensity, tempo) from audio data. The input is preprocessed audio data, and the output is the extracted audio features. Specifically, it uses an audio analysis library (e.g., Librosa).

[0400] Step 6:

[0401] (Identifying mispronunciations)

[0402] The server compares the extracted speech characteristics with the reference speech characteristics to identify pronunciation errors. The input is the speech characteristics and the reference speech characteristics, and the output is the identified pronunciation errors. Specifically, the comparison process is performed using an algorithm.

[0403] Step 7:

[0404] (Emotion recognition)

[0405] The server uses an emotion engine to recognize emotions from voice data. The input is the voice data, and the output is the recognized emotional state. Specifically, it uses IBM Watson Tone Analyzer.

[0406] Step 8:

[0407] (Generating feedback)

[0408] The server generates feedback based on the identified mispronunciations and the recognized emotional state. The feedback may be in the form of visual, text, or audio. The input is the mispronunciation and the emotional state, and the output is the generated feedback. Specifically, a feedback generation module is used.

[0409] Step 9:

[0410] (Sending and viewing feedback)

[0411] The server sends the generated feedback to the device. The device displays and plays the received feedback to the user. The input is the generated feedback, and the output is the feedback provided to the user. Specifically, a display or speaker is used.

[0412] 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.

[0413] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (registered trademark) (Internet search engine).<URL: https: / / openai.com / blog / chatgpt> ), Gemini (registered trademark) (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.

[0414] 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.

[0415] [Second embodiment]

[0416] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0417] 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.

[0418] 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).

[0419] 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.

[0420] 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.

[0421] 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).

[0422] 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.

[0423] 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.

[0424] 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.

[0425] 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.

[0426] 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.

[0427] 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."

[0428] This system analyzes the user's input speech in real time and provides feedback to help the user master correct pronunciation. A specific embodiment of this system will be described below.

[0429] Audio input and capture

[0430] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[0431] Generate audio data and send it to the server

[0432] The device captures the user's speech, stores it locally as digital audio data, converts it into an appropriate format (e.g., WAV, MP3), and then transmits it to the server.

[0433] Voice analysis and speech recognition

[0434] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[0435] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[0436] Compare pronunciations and identify errors

[0437] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0438] Generate feedback

[0439] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0440] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0441] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0442] Audio feedback: Includes audio samples of correct pronunciation.

[0443] Sending and Viewing Feedback

[0444] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0445] This system allows people to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention will be a means to significantly improve the efficiency of phonetic learning.

[0446] The processing flow will be explained below.

[0447] Step 1:

[0448] User: Speak into the device's microphone. For example, say "Hello."

[0449] Step 2:

[0450] Device: Captures the user's speech and stores it locally as digital audio data.

[0451] Step 3:

[0452] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[0453] Step 4:

[0454] Terminal: Sends the converted audio data to the server.

[0455] Step 5:

[0456] Server: Receives the voice data sent from the terminal.

[0457] Step 6:

[0458] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[0459] Step 7:

[0460] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[0461] Step 8:

[0462] Server: Compares the recognized text with reference pronunciation data.

[0463] Step 9:

[0464] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[0465] Step 10:

[0466] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors. The types of feedback include visual feedback, text feedback, and audio feedback.

[0467] Step 11:

[0468] Server: Sends the generated feedback to the device.

[0469] Step 12:

[0470] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[0471] Step 13:

[0472] User: Based on the feedback provided, they correct and try again, which returns them to step 1 and repeats the process to improve their pronunciation.

[0473] Example 1

[0474] 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."

[0475] Conventional speech recognition systems lack the ability to provide specific feedback to help users master correct pronunciation. Furthermore, they lack the means to suggest specific correction methods for specific pronunciation errors, making it difficult for users to effectively correct their pronunciation.

[0476] 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.

[0477] In this invention, the server includes means for capturing input speech and generating speech data, means for transmitting the speech data to the server, means for analyzing the speech data, performing speech recognition, and extracting speech characteristics in the server, means for comparing the speech characteristics with reference speech characteristics and identifying pronunciation errors, means for generating visual feedback, text feedback, and speech feedback for pronunciation improvement based on the pronunciation errors, and means for transmitting the feedback to a terminal and displaying it to the user, thereby enabling the user to efficiently correct their pronunciation.

[0478] "Input voice" refers to the voice spoken by the user into the microphone.

[0479] "Audio Data" refers to data for storing or transmitting captured audio in digital form.

[0480] "Server" refers to a computer system that analyzes voice data, recognizes voice, and performs post-voice recognition processing.

[0481] "Speech recognition" refers to the process of converting input voice data into character string information.

[0482] "Audio characteristics" refers to characteristic information such as pitch, intensity, and tempo contained in audio data.

[0483] "Reference speech characteristics" refers to speech characteristics information collected and recorded based on correct pronunciation.

[0484] "Mispronunciation" refers to a mispronunciation portion identified by comparing the characteristics of the input speech with the reference speech characteristics.

[0485] "Visual feedback" refers to feedback in the form of diagrams or videos that show lip and tongue shape, mouth opening, etc. to help users improve their pronunciation.

[0486] "Textual feedback" refers to written feedback that includes specific instructions to improve a user's pronunciation.

[0487] "Audio feedback" refers to feedback that provides an audio sample of correct pronunciation.

[0488] "Terminal" refers to a device (smartphone, PC, etc.) through which a user inputs voice.

[0489] "Preprocessing" refers to performing processes such as noise removal, normalization, and volume adjustment on audio data.

[0490] This invention is a system that analyzes user-input speech in real time and provides feedback to help users learn correct pronunciation.

[0491] A specific embodiment of this system is described below: A user speaks a target word or sentence into the microphone of the terminal. For example, if a user wants to learn the word "Hello," he speaks "Hello" into the terminal.

[0492] Audio input and capture

[0493] The user speaks into the device's microphone. The device's microphone captures what the user says and stores it locally as digital audio data. The audio data is converted to an appropriate format (e.g., WAV, MP3) using an audio library (e.g., PyDub).

[0494] Generate audio data and send it to the server

[0495] The device sends the saved audio data to the server using the HTTP protocol.

[0496] Voice analysis and speech recognition

[0497] The server receives the audio data sent from the device. After receiving the data, it first performs preprocessing on the audio data. This includes noise removal, normalization, and volume adjustment. An audio signal processing library (e.g., LibROSA) is used for this processing.

[0498] Speech Recognition and Feature Extraction

[0499] After preprocessing is complete, the server performs speech recognition on the audio data and converts the audio into text. A speech recognition API (e.g., Google Cloud Speech-to-Text API) is used for speech recognition. At the same time, voice characteristics (pitch, intensity, tempo, etc.) are extracted. LibROSA is used for feature extraction.

[0500] Compare pronunciations and identify errors

[0501] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0502] Generate feedback

[0503] The server generates feedback to help the user correct their pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0504] Visual feedback: Diagrams and videos showing lip and tongue shape and mouth opening, using video editing software (e.g., Adobe After Effects, Blender).

[0505] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0506] Audio feedback: Includes audio samples of correct pronunciation.

[0507] Sending and Viewing Feedback

[0508] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0509] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention is a means of significantly improving the efficiency of phonetic learning.

[0510] Example 1

[0511] The user says "Hello" and the device's microphone captures the audio. The device converts the captured audio to WAV format and sends it to the server. The server receives the audio data and performs noise reduction and normalization. The server uses the Google Cloud Speech-to-Text API to convert "Hello" to text and simultaneously calculates pitch and intensity. The server compares "Hello" with standard pronunciation data and determines that the "e" sound is closer to a "u." The server generates text feedback such as "Place your tongue behind your upper teeth to pronounce 'L'" along with an audio sample of the correct pronunciation and sends it to the device. The device presents this to the user, who then practices their pronunciation again based on the feedback.

[0512] Prompt Sentence Examples

[0513] "The user said 'Hello'. Given this audio data, please compare it with a standard pronunciation to generate correct feedback."

[0514] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0515] Step 1: Capturing Audio Input

[0516] The user speaks into the microphone of the terminal.

[0517] Input: User's voice, saying "Hello".

[0518] The device captures the spoken voice through a microphone and stores it as digital voice data.

[0519] Output: Captured digital audio data (WAV, MP3 format, etc.).

[0520] What it does: The device's audio API captures audio input from the microphone and saves it as an audio file.

[0521] Step 2: Generate and send audio data

[0522] The terminal transmits the stored voice data to the server.

[0523] Input: Captured digital audio data.

[0524] The terminal uses the HTTP protocol to send the audio data to the server.

[0525] Output: The audio data sent to the server.

[0526] What it does: The device converts the voice data into the appropriate format and sends the data to a server over the Internet.

[0527] Step 3: Preprocessing the audio data

[0528] The server receives the voice data transmitted from the terminal.

[0529] Input: Audio data sent from the device.

[0530] The server performs preprocessing such as noise removal, normalization, and volume adjustment on the received audio data.

[0531] Output: Preprocessed audio data.

[0532] What it does: A server-side audio signal processing library (e.g., LibROSA) removes noise from the audio data and normalizes the volume level to adjust it to an appropriate sound quality.

[0533] Step 4: Speech recognition and feature extraction

[0534] The server analyzes the preprocessed audio data.

[0535] Input: Preprocessed audio data.

[0536] The server performs speech recognition and converts the audio data into text, while simultaneously extracting audio characteristics (pitch, intensity, tempo, etc.).

[0537] Output: Text data and speech characteristics data generated by speech recognition.

[0538] Specific operation: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) to convert the speech into text, and calculates speech feature information using a library such as LibROSA.

[0539] Step 5: Compare pronunciation and identify errors

[0540] The server compares the recognized text with the reference pronunciation data.

[0541] Input: Text data generated by speech recognition, reference pronunciation data.

[0542] The server identifies which parts of the user's pronunciation are incorrect.

[0543] Output: Mispronunciation information.

[0544] Specific operation: The server refers to a dataset of correct pronunciations collected in advance, compares it with the user's pronunciation, identifies incorrect parts, and records them.

[0545] Step 6: Generate feedback

[0546] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[0547] Input: Mispronunciation information.

[0548] The server generates visual, textual and audio feedback.

[0549] Output: Feedback data (visual, text, audio).

[0550] Specific behaviors: Visual feedback is generated using video editing software (e.g., Adobe After Effects), text feedback is written with specific instructions, and audio feedback uses pre-recorded samples of correct pronunciation.

[0551] Step 7: Submit and view feedback

[0552] The server transmits the generated feedback to the terminal.

[0553] Input: Feedback data (visual, text, audio).

[0554] The device analyzes the received feedback and provides it to the user visually and audibly.

[0555] Output: Feedback provided to the user.

[0556] Specific operation: The device presents video, text, and audio data through a user interface, and the user corrects their pronunciation based on the feedback.

[0557] (Application example 1)

[0558] 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."

[0559] Conventional speech analysis systems have primarily been used for personal pronunciation practice. However, effective means for improving the quality of speech communication between store clerks and customers in brick-and-mortar stores are limited. This results in missed opportunities for improving customer satisfaction and improving store clerk skills. The present invention aims to solve this problem by providing a system that evaluates the pronunciation quality of store clerk interactions in real time and provides effective feedback.

[0560] 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.

[0561] In this invention, the server includes means for capturing and analyzing voice data, means for performing voice recognition and extracting voice characteristics, means for comparing the extracted voice characteristics with reference voice characteristics to identify pronunciation errors, means for generating feedback for pronunciation improvement based on the pronunciation errors, means for transmitting the feedback to a terminal and displaying it to the user as visual, text, and voice feedback, and means for analyzing voices spoken by store clerks when interacting with customers and providing feedback, thereby enabling real-time pronunciation improvement so that store clerks can communicate with customers with appropriate pronunciation.

[0562] "Capture" means acquiring information such as sound or video and saving it as digital data.

[0563] "Audio data" is data that represents audio in digital form.

[0564] A "server" is a centralized management system for processing and storing data.

[0565] "Speech recognition" is the technology that analyzes voice data and converts the content into text or other formats.

[0566] "Audio characteristics" are characteristic attributes associated with an audio signal, including pitch, intensity, tempo, etc.

[0567] "Reference speech characteristics" are reference data used to indicate correct pronunciation.

[0568] A "mispronunciation" is a difference observed between the user's pronunciation and the reference pronunciation.

[0569] "Feedback" is information that informs the user of the evaluation results and areas for improvement.

[0570] A "terminal" is a computing device that can be directly operated by a user.

[0571] A "store associate" refers to someone who provides service to customers in a physical store.

[0572] "Customer service" refers to a series of actions and statements that store clerks make to communicate with customers.

[0573] This system analyzes in real time the speech spoken by store clerks when interacting with customers and provides feedback to help improve pronunciation. A specific embodiment of this system will be described below.

[0574] Audio input and capture

[0575] The user (store clerk) speaks into the microphone. For example, the user speaks the phrase "Hello, how can I help you?"

[0576] Generate audio data and send it to the server

[0577] The device captures the user's speech and stores it as digital audio data, which is then converted to WAV or MP3 format and sent to a server.

[0578] Voice analysis and speech recognition

[0579] After receiving the voice data from the device, the server performs preprocessing on the voice data, specifically noise removal, normalization, and volume adjustment. After preprocessing is complete, the server performs speech recognition on the voice data and converts the voice into text. At the same time, it extracts voice characteristics (pitch, intensity, tempo, etc.).

[0580] Compare pronunciations and identify errors

[0581] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0582] Generate feedback

[0583] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0584] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0585] Text feedback: Information in written form with specific instructions, such as "Try to pronounce the letter 'L' with your tongue behind your upper teeth."

[0586] Audio feedback: Includes audio samples of correct pronunciation.

[0587] Sending and Viewing Feedback

[0588] The server sends the generated feedback to the device, which then analyzes the received feedback and provides it to the user visually and audibly, allowing the user to understand the problems with their pronunciation and immediately try to correct them.

[0589] As a concrete example, consider a user saying the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it, identifies pronunciation errors, and provides feedback to help improve. For example, if the system determines that the "o" sound in "Hello" is not pronounced correctly, it immediately provides corrective feedback to the user.

[0590] Also, examples of prompts to use this system effectively include:

[0591] "Please say the following English sentence: 'Hello, how can I help you?'"

[0592] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[0593] It can be provided as such.

[0594] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0595] Step 1:

[0596] The user speaks the target phrase or word into the microphone.

[0597] Input: User's voice

[0598] Output: Analog audio signal

[0599] In this step, audio is captured as the user speaks, for example, "Hello, how can I help you?"

[0600] Step 2:

[0601] The terminal converts the captured analog audio signal into digital audio data and stores it locally.

[0602] Input: Analog audio signal

[0603] Output: Digital audio data (WAV or MP3 format)

[0604] The device converts the analog audio signal received through the microphone into digital format and saves it in WAV or MP3 format.

[0605] Step 3:

[0606] The terminal transmits the converted digital audio data to the server.

[0607] Input: Digital audio data (WAV or MP3 format)

[0608] Output: Audio data sent to the server

[0609] The terminal uploads the voice data to the server via the network.

[0610] Step 4:

[0611] The server preprocesses the received audio data, performing noise removal, normalization, and volume adjustment.

[0612] Input: Transmitted digital audio data

[0613] Output: Preprocessed audio data

[0614] The server cleans the audio data by applying noise reduction filters, normalizing the volume, etc.

[0615] Step 5:

[0616] The server performs speech recognition on the preprocessed audio data, converting it to text and simultaneously extracting audio characteristics (pitch, intensity, tempo).

[0617] Input: Preprocessed audio data

[0618] Output: Text data and voice characteristics data

[0619] The server uses speech recognition algorithms to analyze the audio and convert it to text, while also extracting audio characteristics such as pitch and intensity.

[0620] Step 6:

[0621] The server compares the extracted speech characteristics with reference speech characteristics to identify mispronunciations.

[0622] Input: Voice characteristic data and reference voice characteristic data

[0623] Output: Mispronunciation information

[0624] The server compares and analyzes the reference voice data with the user's voice characteristics to identify incorrect pronunciations.

[0625] Step 7:

[0626] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[0627] Input: Mispronunciation information

[0628] Output: Feedback information (visual, text, audio)

[0629] The server generates specific feedback to help improve mispronunciations, including diagrams, detailed instructions, and audio samples of correct pronunciation.

[0630] Step 8:

[0631] The server transmits the generated feedback to the terminal, which displays the feedback to the user.

[0632] Input: Feedback information

[0633] Output: Visual and auditory feedback to the user

[0634] The feedback information sent from the server to the device is analyzed and displayed on the device. The user receives feedback from the screen and speaker and uses it to improve their pronunciation.

[0635] As a concrete example, consider a user uttering the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it to identify pronunciation errors, and provides real-time feedback for improvement.

[0636] An example prompt is:

[0637] "Please say the following English sentence: 'Hello, how can I help you?'"

[0638] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[0639] 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.

[0640] This system analyzes the user's voice input in real time and provides feedback to help them master correct pronunciation. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions and provides appropriate feedback according to the user's emotions.

[0641] Audio input and capture

[0642] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[0643] Generate audio data and send it to the server

[0644] The device captures the user's speech and stores it locally as digital audio data, which is then converted into an appropriate format (e.g., WAV, MP3) and sent to the server.

[0645] Voice analysis and speech recognition

[0646] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[0647] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[0648] Compare pronunciations and identify errors

[0649] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0650] Emotion recognition

[0651] The server uses an emotion engine to recognize emotions from the user's voice data. The emotion engine analyzes voice characteristics (e.g., pitch, intensity, rate) to identify the user's emotional state (e.g., happy, sad, anger).

[0652] Generate feedback

[0653] The server generates feedback based on the identified pronunciation errors and the user's emotional state. The feedback is provided in the following format:

[0654] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0655] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0656] Audio feedback: Includes audio samples of correct pronunciation.

[0657] If the emotion engine detects negative emotions (e.g., frustration, sadness), the server generates feedback with encouraging messages, such as "Great progress! Try a little harder next time!"

[0658] Sending and Viewing Feedback

[0659] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and their emotional state and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0660] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. Furthermore, by providing feedback that takes into account the user's emotions, it helps maintain motivation. This invention will significantly improve the efficiency and effectiveness of phonetic learning.

[0661] The processing flow will be explained below.

[0662] Step 1:

[0663] User: Speak into the device's microphone. For example, if you want to learn the word "Hello," speak "Hello" into the device.

[0664] Step 2:

[0665] Device: Captures the user's speech and stores it locally as digital audio data.

[0666] Step 3:

[0667] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[0668] Step 4:

[0669] Terminal: Sends the converted audio data to the server.

[0670] Step 5:

[0671] Server: Receives the voice data sent from the terminal.

[0672] Step 6:

[0673] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[0674] Step 7:

[0675] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[0676] Step 8:

[0677] Server: Compares the recognized text with reference pronunciation data.

[0678] Step 9:

[0679] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[0680] Step 10:

[0681] Server: Inputs the user's voice data into the emotion engine and recognizes the user's emotions (e.g., joy, sadness, anger).

[0682] Step 11:

[0683] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors and the user's emotional state. For example, it generates feedback including instructions to correct the "e" sound in "Hello" and an encouraging message if the user is showing negative emotions.

[0684] Step 12:

[0685] Server: Sends the generated feedback to the device.

[0686] Step 13:

[0687] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[0688] Step 14:

[0689] User: Attempts to pronounce again based on the feedback provided, thereby repeating the process and improving pronunciation.

[0690] For example, if a user says "Hello" and the server determines that the "e" sound is close to "u," and the emotion engine recognizes that the user is frustrated, the system will provide feedback such as, "Open your lips a little and lower your tongue a little. Then try again. That's great progress!" In this way, the system effectively supports pronunciation correction while taking the user's emotions into consideration.

[0691] Example 2

[0692] 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."

[0693] In learning pronunciation, it is difficult for users to accurately understand and correct their pronunciation errors. Maintaining motivation during the learning process is also a challenge. Furthermore, there is currently a lack of systems that provide appropriate feedback based on the user's emotional state.

[0694] 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.

[0695] In this invention, the server includes means for capturing input voice and generating voice data, means for transmitting the voice data to the server, and means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server, thereby enabling the server to analyze the voice data, identify pronunciation errors, and further recognize emotions.

[0696] "Means for capturing input voice and generating voice data" refers to the process of capturing voice spoken by a user in real time through the device's microphone and saving it as digital voice data.

[0697] The "means for transmitting the audio data to the server" refers to a method for transmitting digital audio data stored in the terminal to the server via a network.

[0698] "Means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server" refers to the function of performing voice recognition on the voice data received by the server, and analyzing and extracting characteristics such as pitch, intensity, and tempo from the voice data.

[0699] The "means for comparing the characteristics of the voice with reference voice characteristics and identifying pronunciation errors" refers to a process for comparing the characteristics of the analyzed voice with reference correct voice characteristics and identifying the errors.

[0700] "Means for generating feedback for pronunciation improvement based on the pronunciation errors" refers to a method for generating feedback including advice and instructions for the user to pronounce correctly based on the identified pronunciation errors.

[0701] The "means for transmitting the feedback to the terminal and displaying it to the user" refers to a method for transmitting the generated feedback from the server to the terminal and presenting it to the user visually or audibly at the terminal.

[0702] "Means for recognizing emotions from the user's voice and generating appropriate feedback based on the recognized emotions" refers to a process of analyzing the user's voice characteristics to recognize their emotional state and creating feedback according to that emotion.

[0703] This invention is a system that supports users in learning pronunciation in real time. It mainly involves a series of processes: voice input, voice data generation and transmission, voice recognition and analysis, pronunciation comparison and error identification, emotion recognition, and feedback generation and transmission.

[0704] Voice input

[0705] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn "Hello," they speak "Hello" into the device. The device captures this speech in real time.

[0706] Generates audio data and sends it to the server

[0707] The device stores the captured audio as digital data and converts it to an appropriate format, such as WAV or MP3, and transmits the converted audio data over the network to a server, along with metadata such as the user ID and a timestamp.

[0708] Voice analysis and speech recognition

[0709] The server processes the audio data received from the device. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. This preprocessing is a fundamental step to obtain clean audio data. Next, it converts the audio data into text using a speech recognition engine (e.g., Google Cloud Speech-to-Text API). In addition, it analyzes and extracts audio characteristics (e.g., pitch, intensity, tempo).

[0710] Compare pronunciations and identify errors

[0711] The server compares the parsed text with a reference pronunciation database containing pronunciations from native speakers. The comparison identifies any pronunciation errors the user may have made. For example, if the "e" in "Hello" sounds more like a "u," that part is recorded as an error.

[0712] Emotion recognition

[0713] The server analyzes the voice data through an emotion recognition engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. It analyzes voice characteristics such as pitch, intensity, and speed to determine whether the user is happy, sad, or angry.

[0714] Generate feedback

[0715] The server generates feedback based on the identified pronunciation errors and the user's emotional state, and the feedback is provided in visual, text, and audio formats.

[0716] Visual feedback: Diagrams or videos showing lip and tongue position and mouth opening (e.g., educational video link).

[0717] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your top teeth to pronounce the letter 'L'").

[0718] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker).

[0719] If the emotion recognition engine detects a negative emotion, the server adds an encouraging message, such as "Great progress! Try a little harder next time!"

[0720] Sending and Viewing Feedback

[0721] The server sends the generated feedback to the device, which analyzes it and provides it visually and audibly to the user, allowing the user to understand their pronunciation problems and try to correct them in real time.

[0722] As a concrete example, if a user wants to learn the phrase "Thank you," the entire process is carried out sequentially, starting with voice input, followed by generation and transmission of voice data, analysis and recognition, identification of pronunciation errors, emotion recognition, feedback generation, and finally transmission and display of the feedback.

[0723] Example prompts for generative AI models

[0724] Explain the steps to analyze the voice data of a user saying "Thank you" into the device's microphone, recognize pronunciation errors and the user's emotional state, and generate and display appropriate feedback.

[0725] This system allows you to effectively learn correct pronunciation without the need for special equipment or expensive lessons. It also provides emotionally sensitive feedback, making it easier to maintain motivation. This invention is a means to significantly improve the efficiency and effectiveness of phonetic learning.

[0726] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0727] Step 1: Audio Input and Capture

[0728] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn the word "Hello," they speak "Hello." This is the user's voice input. The device captures this voice in real time and saves it as digital voice data. The input is the user's voice, and the output is digital voice data. At this stage, voice capture software is used.

[0729] Step 2: Generate audio data and send it to the server

[0730] The device converts the captured audio into an appropriate format (e.g., WAV, MP3). This converted digital audio data is then sent to the server. The input is the captured audio data, and the output is the converted audio data. The converted audio also contains metadata such as the user ID and a timestamp. Audio format conversion and network communication software is used at this stage.

[0731] Step 3: Voice analysis and speech recognition

[0732] The server analyzes the received audio data. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. Then, it uses a speech recognition engine to convert the audio data to text. At the same time, it analyzes and extracts audio characteristics (pitch, intensity, tempo, etc.). The input is the formatted audio data, and the output is text data and audio characteristic data. At this stage, audio processing and speech recognition software is used. Preprocessing includes noise filtering algorithms and signal normalization algorithms.

[0733] Step 4: Compare pronunciation and identify errors

[0734] The server compares the parsed text to a reference pronunciation database. This reference data contains pronunciations from native speakers. The comparison identifies pronunciation errors. For example, if the "e" in "Hello" is close to a "u," it is marked as an error. The input is text data and voice characteristics data, and the output is data indicating pronunciation errors. This stage involves comparison algorithms and database search software.

[0735] Step 5: Recognize emotions

[0736] The server uses the voice data to recognize the user's emotions. An emotion recognition engine is used to analyze voice characteristics (pitch, intensity, speed, etc.) to identify the user's emotional state. For example, emotions such as joy, sadness, and anger can be identified from the analysis results. The input is the voice characteristic data, and the output is the recognized emotional state. At this stage, an emotion recognition algorithm is used.

[0737] Step 6: Generate feedback

[0738] The server generates feedback based on the identified pronunciation errors and the recognized emotional state. The feedback is provided in the following form:

[0739] Visual feedback: illustrations and videos showing lip and tongue position and mouth opening

[0740] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your upper teeth to pronounce the letter 'L'")

[0741] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker)

[0742] If the emotion recognition engine detects a negative emotion, it adds an encouraging message. The input is the pronunciation error data and the emotional state data, and the output is the generated feedback. At this stage, a feedback generation algorithm is used.

[0743] Step 7: Submit and view feedback

[0744] The server sends the generated feedback to the device. The device receives this feedback and displays it visually and audibly to the user. For example, an on-screen diagram showing the correct lip and tongue positions and an audio playback of the correct pronunciation. The input is the generated feedback, and the output is the visual and audible information presented to the user. This is where communication software and user interface software come into play.

[0745] Through these steps, the system can efficiently improve the user's pronunciation and maintain their motivation to learn.

[0746] (Application example 2)

[0747] 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."

[0748] The present invention relates to a system that improves the efficiency of a user's pronunciation practice and provides appropriate feedback according to the user's emotional state. Conventional pronunciation learning systems can identify pronunciation errors, but they cannot take the user's emotions or state into account, making it difficult to maintain motivation for learning. Therefore, in order to maximize the effectiveness of pronunciation learning, a system that provides feedback according to the user's emotions is needed.

[0749] The identification process by the identification 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 including an emotion engine for recognizing emotions from voice data, means for generating feedback according to the emotions, and means for performing preprocessing such as noise removal, normalization, and volume adjustment. This makes it possible to identify errors in the user's voice pronunciation and provide feedback according to the user's emotional state.

[0750] A "means for generating voice data" is a device or software that allows a user to capture voice input and store it as voice data in digital form.

[0751] The "means for transmitting voice data to a server" refers to a device or software for transferring the generated voice data to a server via a network.

[0752] "Means for analyzing audio data, performing speech recognition, and extracting audio characteristics" refers to devices or software that apply a speech recognition algorithm to audio data within a server and analyze audio characteristics (e.g., pitch, intensity, tempo, etc.).

[0753] The "means for comparing voice characteristics with reference voice characteristics and identifying pronunciation errors" refers to a device or software that compares the user's voice characteristics with preset reference voice characteristics and identifies incorrect pronunciation parts.

[0754] A "means for generating feedback for pronunciation improvement" is a device or software that generates specific instructions or samples for correcting pronunciation errors.

[0755] The "means for transmitting feedback to a terminal and displaying it to a user" refers to a device or software for transmitting the generated feedback to a user's terminal via a network and displaying it.

[0756] The "means including an emotion engine for recognizing emotions" refers to a device or software for analyzing emotions from voice data and identifying the emotional state of a user.

[0757] The "means for generating emotional feedback" refers to a device or software for generating feedback including appropriate encouragement, warnings, and specific instructions based on the user's emotional state.

[0758] The present invention relates to a system that identifies user pronunciation errors and provides real-time emotional feedback. The system generates speech data, transmits it to a server, and analyzes it to perform speech recognition and extract speech characteristics. Furthermore, the system uses an emotion engine to recognize the user's emotions and generate specific feedback accordingly.

[0759] Processing contents to be performed by the program

[0760] The invention uses a "device," such as a microphone or smartphone, to capture a user's voice input. The device stores the captured audio as digital audio data and then transmits this data to a server, where it is converted into an appropriate format (e.g., WAV or MP3).

[0761] The server analyzes the audio data using an installed speech recognition engine (e.g., Google Speech-to-Text API) to convert the speech to text and extracts voice characteristics (e.g., pitch, intensity, tempo). This data is compared with baseline voice characteristics to identify pronunciation errors.

[0762] Furthermore, an emotion recognition engine (e.g., IBM Watson Tone Analyzer) is used to analyze the user's emotions from the voice data, identifying the user's emotional state (e.g., joy, sadness, anger, etc.) and generating feedback accordingly.

[0763] Providing Feedback

[0764] The form of feedback includes at least one of visual feedback, text feedback, and audio feedback. Visual feedback may include animations or videos showing lip movements and tongue positions that indicate correct pronunciation. Text feedback may include specific instructions such as "Put your tongue behind your upper teeth to pronounce 'L'." Audio feedback may include audio samples of correct pronunciation.

[0765] Specific examples

[0766] For example, suppose a factory operator says "start" when operating a robot. This voice input is captured by the device's microphone and sent to a server. The server analyzes the speech of "start" and detects that a specific part (such as the "ta" sound) is not pronounced correctly. It then uses an emotion recognition engine to determine that the operator is slightly nervous. As a result, the system displays an encouraging message along with animations and voice samples showing examples of how to pronounce "ta" correctly.

[0767] Prompt Sentence Examples

[0768] Please tell us about the voice-guided operation support system in factories.

[0769] This system helps factory operators give accurate voice instructions to robots and provides feedback according to the operator's emotional state. It includes steps for voice recognition, pronunciation analysis, emotion recognition, and feedback generation.

[0770] This makes it easier for users to learn correct pronunciation, maintains their motivation to learn, and improves factory operation efficiency.

[0771] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0772] Step 1:

[0773] (audio input and capture)

[0774] The user speaks into a microphone. For example, a factory operator says "start." This voice is captured by the device's microphone and saved as digital voice data. The input is real-time voice data, and the output is saved digital voice data. Specifically, a microphone and voice capture software installed on a smartphone or dedicated device is used.

[0775] Step 2:

[0776] (Sending audio data)

[0777] The device converts the captured digital audio data into an appropriate format (e.g., WAV, MP3) and transmits it to a server via the Internet. The input is the stored digital audio data, and the output is the audio data transmitted to the server. Specifically, a communication module and a data transfer protocol are used.

[0778] Step 3:

[0779] (Audio pre-processing)

[0780] The server performs preprocessing on the received audio data. Specifically, it performs noise removal, normalization, volume adjustment, etc. The input is the audio data sent to the server, and the output is the audio data after preprocessing. As a specific example, an audio processing library (e.g., WebRTC) is used.

[0781] Step 4:

[0782] (Voice Recognition)

[0783] The server performs speech recognition on the preprocessed audio data and converts the audio into text. The input is the preprocessed audio data, and the output is text generated from the audio. Specifically, the server uses the Google Speech-to-Text API or similar.

[0784] Step 5:

[0785] (Extraction of voice characteristics)

[0786] The server extracts audio features (e.g., pitch, intensity, tempo) from audio data. The input is preprocessed audio data, and the output is the extracted audio features. Specifically, it uses an audio analysis library (e.g., Librosa).

[0787] Step 6:

[0788] (Identifying mispronunciations)

[0789] The server compares the extracted speech characteristics with the reference speech characteristics to identify pronunciation errors. The input is the speech characteristics and the reference speech characteristics, and the output is the identified pronunciation errors. Specifically, the comparison process is performed using an algorithm.

[0790] Step 7:

[0791] (Emotion recognition)

[0792] The server uses an emotion engine to recognize emotions from voice data. The input is the voice data, and the output is the recognized emotional state. Specifically, it uses IBM Watson Tone Analyzer.

[0793] Step 8:

[0794] (Generating feedback)

[0795] The server generates feedback based on the identified mispronunciations and the recognized emotional state. The feedback may be in the form of visual, text, or audio. The input is the mispronunciation and the emotional state, and the output is the generated feedback. Specifically, a feedback generation module is used.

[0796] Step 9:

[0797] (Sending and viewing feedback)

[0798] The server sends the generated feedback to the device. The device displays and plays the received feedback to the user. The input is the generated feedback, and the output is the feedback provided to the user. Specifically, a display or speaker is used.

[0799] 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.

[0800] 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.

[0801] 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.

[0802] [Third embodiment]

[0803] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[0804] 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.

[0805] 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).

[0806] 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.

[0807] 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.

[0808] 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).

[0809] 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.

[0810] 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.

[0811] 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.

[0812] 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.

[0813] 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.

[0814] 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."

[0815] This system analyzes the user's input speech in real time and provides feedback to help the user master correct pronunciation. A specific embodiment of this system will be described below.

[0816] Audio input and capture

[0817] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[0818] Generate audio data and send it to the server

[0819] The device captures the user's speech, stores it locally as digital audio data, converts it into an appropriate format (e.g., WAV, MP3), and then transmits it to the server.

[0820] Voice analysis and speech recognition

[0821] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[0822] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[0823] Compare pronunciations and identify errors

[0824] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0825] Generate feedback

[0826] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0827] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0828] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0829] Audio feedback: Includes audio samples of correct pronunciation.

[0830] Sending and Viewing Feedback

[0831] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0832] This system allows people to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention will be a means to significantly improve the efficiency of phonetic learning.

[0833] The processing flow will be explained below.

[0834] Step 1:

[0835] User: Speak into the device's microphone. For example, say "Hello."

[0836] Step 2:

[0837] Device: Captures the user's speech and stores it locally as digital audio data.

[0838] Step 3:

[0839] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[0840] Step 4:

[0841] Terminal: Sends the converted audio data to the server.

[0842] Step 5:

[0843] Server: Receives the voice data sent from the terminal.

[0844] Step 6:

[0845] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[0846] Step 7:

[0847] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[0848] Step 8:

[0849] Server: Compares the recognized text with reference pronunciation data.

[0850] Step 9:

[0851] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[0852] Step 10:

[0853] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors. The types of feedback include visual feedback, text feedback, and audio feedback.

[0854] Step 11:

[0855] Server: Sends the generated feedback to the device.

[0856] Step 12:

[0857] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[0858] Step 13:

[0859] User: Based on the feedback provided, they correct and try again, which returns them to step 1 and repeats the process to improve their pronunciation.

[0860] Example 1

[0861] 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."

[0862] Conventional speech recognition systems lack the ability to provide specific feedback to help users master correct pronunciation. Furthermore, they lack the means to suggest specific correction methods for specific pronunciation errors, making it difficult for users to effectively correct their pronunciation.

[0863] 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.

[0864] In this invention, the server includes means for capturing input speech and generating speech data, means for transmitting the speech data to the server, means for analyzing the speech data, performing speech recognition, and extracting speech characteristics in the server, means for comparing the speech characteristics with reference speech characteristics and identifying pronunciation errors, means for generating visual feedback, text feedback, and speech feedback for pronunciation improvement based on the pronunciation errors, and means for transmitting the feedback to a terminal and displaying it to the user, thereby enabling the user to efficiently correct their pronunciation.

[0865] "Input voice" refers to the voice spoken by the user into the microphone.

[0866] "Audio Data" refers to data for storing or transmitting captured audio in digital form.

[0867] "Server" refers to a computer system that analyzes voice data, recognizes voice, and performs post-voice recognition processing.

[0868] "Speech recognition" refers to the process of converting input voice data into character string information.

[0869] "Audio characteristics" refers to characteristic information such as pitch, intensity, and tempo contained in audio data.

[0870] "Reference speech characteristics" refers to speech characteristics information collected and recorded based on correct pronunciation.

[0871] "Mispronunciation" refers to a mispronunciation portion identified by comparing the characteristics of the input speech with the reference speech characteristics.

[0872] "Visual feedback" refers to feedback in the form of diagrams or videos that show lip and tongue shape, mouth opening, etc. to help users improve their pronunciation.

[0873] "Textual feedback" refers to written feedback that includes specific instructions to improve a user's pronunciation.

[0874] "Audio feedback" refers to feedback that provides an audio sample of correct pronunciation.

[0875] "Terminal" refers to a device (smartphone, PC, etc.) through which a user inputs voice.

[0876] "Preprocessing" refers to performing processes such as noise removal, normalization, and volume adjustment on audio data.

[0877] This invention is a system that analyzes user-input speech in real time and provides feedback to help users learn correct pronunciation.

[0878] A specific embodiment of this system is described below: A user speaks a target word or sentence into the microphone of the terminal. For example, if a user wants to learn the word "Hello," he speaks "Hello" into the terminal.

[0879] Audio input and capture

[0880] The user speaks into the device's microphone. The device's microphone captures what the user says and stores it locally as digital audio data. The audio data is converted to an appropriate format (e.g., WAV, MP3) using an audio library (e.g., PyDub).

[0881] Generate audio data and send it to the server

[0882] The device sends the saved audio data to the server using the HTTP protocol.

[0883] Voice analysis and speech recognition

[0884] The server receives the audio data sent from the device. After receiving the data, it first performs preprocessing on the audio data. This includes noise removal, normalization, and volume adjustment. An audio signal processing library (e.g., LibROSA) is used for this processing.

[0885] Speech Recognition and Feature Extraction

[0886] After preprocessing is complete, the server performs speech recognition on the audio data and converts the audio into text. A speech recognition API (e.g., Google Cloud Speech-to-Text API) is used for speech recognition. At the same time, voice characteristics (pitch, intensity, tempo, etc.) are extracted. LibROSA is used for feature extraction.

[0887] Compare pronunciations and identify errors

[0888] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0889] Generate feedback

[0890] The server generates feedback to help the user correct their pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0891] Visual feedback: Diagrams and videos showing lip and tongue shape and mouth opening, using video editing software (e.g., Adobe After Effects, Blender).

[0892] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[0893] Audio feedback: Includes audio samples of correct pronunciation.

[0894] Sending and Viewing Feedback

[0895] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[0896] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention is a means of significantly improving the efficiency of phonetic learning.

[0897] Example 1

[0898] The user says "Hello" and the device's microphone captures the audio. The device converts the captured audio to WAV format and sends it to the server. The server receives the audio data and performs noise reduction and normalization. The server uses the Google Cloud Speech-to-Text API to convert "Hello" to text and simultaneously calculates pitch and intensity. The server compares "Hello" with standard pronunciation data and determines that the "e" sound is closer to a "u." The server generates text feedback such as "Place your tongue behind your upper teeth to pronounce 'L'" along with an audio sample of the correct pronunciation and sends it to the device. The device presents this to the user, who then practices their pronunciation again based on the feedback.

[0899] Prompt Sentence Examples

[0900] "The user said 'Hello'. Given this audio data, please compare it with a standard pronunciation to generate correct feedback."

[0901] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0902] Step 1: Capturing Audio Input

[0903] The user speaks into the microphone of the terminal.

[0904] Input: User's voice, saying "Hello".

[0905] The device captures the spoken voice through a microphone and stores it as digital voice data.

[0906] Output: Captured digital audio data (WAV, MP3 format, etc.).

[0907] What it does: The device's audio API captures audio input from the microphone and saves it as an audio file.

[0908] Step 2: Generate and send audio data

[0909] The terminal transmits the stored voice data to the server.

[0910] Input: Captured digital audio data.

[0911] The terminal uses the HTTP protocol to send the audio data to the server.

[0912] Output: The audio data sent to the server.

[0913] What it does: The device converts the voice data into the appropriate format and sends the data to a server over the Internet.

[0914] Step 3: Preprocessing the audio data

[0915] The server receives the voice data transmitted from the terminal.

[0916] Input: Audio data sent from the device.

[0917] The server performs preprocessing such as noise removal, normalization, and volume adjustment on the received audio data.

[0918] Output: Preprocessed audio data.

[0919] What it does: A server-side audio signal processing library (e.g., LibROSA) removes noise from the audio data and normalizes the volume level to adjust it to an appropriate sound quality.

[0920] Step 4: Speech recognition and feature extraction

[0921] The server analyzes the preprocessed audio data.

[0922] Input: Preprocessed audio data.

[0923] The server performs speech recognition and converts the audio data into text, while simultaneously extracting audio characteristics (pitch, intensity, tempo, etc.).

[0924] Output: Text data and speech characteristics data generated by speech recognition.

[0925] Specific operation: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) to convert the speech into text, and calculates speech feature information using a library such as LibROSA.

[0926] Step 5: Compare pronunciation and identify errors

[0927] The server compares the recognized text with the reference pronunciation data.

[0928] Input: Text data generated by speech recognition, reference pronunciation data.

[0929] The server identifies which parts of the user's pronunciation are incorrect.

[0930] Output: Mispronunciation information.

[0931] Specific operation: The server refers to a dataset of correct pronunciations collected in advance, compares it with the user's pronunciation, identifies incorrect parts, and records them.

[0932] Step 6: Generate feedback

[0933] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[0934] Input: Mispronunciation information.

[0935] The server generates visual, textual and audio feedback.

[0936] Output: Feedback data (visual, text, audio).

[0937] Specific behaviors: Visual feedback is generated using video editing software (e.g., Adobe After Effects), text feedback is written with specific instructions, and audio feedback uses pre-recorded samples of correct pronunciation.

[0938] Step 7: Submit and view feedback

[0939] The server transmits the generated feedback to the terminal.

[0940] Input: Feedback data (visual, text, audio).

[0941] The device analyzes the received feedback and provides it to the user visually and audibly.

[0942] Output: Feedback provided to the user.

[0943] Specific operation: The device presents video, text, and audio data through a user interface, and the user corrects their pronunciation based on the feedback.

[0944] (Application example 1)

[0945] 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."

[0946] Conventional speech analysis systems have primarily been used for personal pronunciation practice. However, effective means for improving the quality of speech communication between store clerks and customers in brick-and-mortar stores are limited. This results in missed opportunities for improving customer satisfaction and improving store clerk skills. The present invention aims to solve this problem by providing a system that evaluates the pronunciation quality of store clerk interactions in real time and provides effective feedback.

[0947] 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.

[0948] In this invention, the server includes means for capturing and analyzing voice data, means for performing voice recognition and extracting voice characteristics, means for comparing the extracted voice characteristics with reference voice characteristics to identify pronunciation errors, means for generating feedback for pronunciation improvement based on the pronunciation errors, means for transmitting the feedback to a terminal and displaying it to the user as visual, text, and voice feedback, and means for analyzing voices spoken by store clerks when interacting with customers and providing feedback, thereby enabling real-time pronunciation improvement so that store clerks can communicate with customers with appropriate pronunciation.

[0949] "Capture" means acquiring information such as sound or video and saving it as digital data.

[0950] "Audio data" is data that represents audio in digital form.

[0951] A "server" is a centralized management system for processing and storing data.

[0952] "Speech recognition" is the technology that analyzes voice data and converts the content into text or other formats.

[0953] "Audio characteristics" are characteristic attributes associated with an audio signal, including pitch, intensity, tempo, etc.

[0954] "Reference speech characteristics" are reference data used to indicate correct pronunciation.

[0955] A "mispronunciation" is a difference observed between the user's pronunciation and the reference pronunciation.

[0956] "Feedback" is information that informs the user of the evaluation results and areas for improvement.

[0957] A "terminal" is a computing device that can be directly operated by a user.

[0958] A "store associate" refers to someone who provides service to customers in a physical store.

[0959] "Customer service" refers to a series of actions and statements that store clerks make to communicate with customers.

[0960] This system analyzes in real time the speech spoken by store clerks when interacting with customers and provides feedback to help improve pronunciation. A specific embodiment of this system will be described below.

[0961] Audio input and capture

[0962] The user (store clerk) speaks into the microphone. For example, the user speaks the phrase "Hello, how can I help you?"

[0963] Generate audio data and send it to the server

[0964] The device captures the user's speech and stores it as digital audio data, which is then converted to WAV or MP3 format and sent to a server.

[0965] Voice analysis and speech recognition

[0966] After receiving the voice data from the device, the server performs preprocessing on the voice data, specifically noise removal, normalization, and volume adjustment. After preprocessing is complete, the server performs speech recognition on the voice data and converts the voice into text. At the same time, it extracts voice characteristics (pitch, intensity, tempo, etc.).

[0967] Compare pronunciations and identify errors

[0968] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[0969] Generate feedback

[0970] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[0971] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[0972] Text feedback: Information in written form with specific instructions, such as "Try to pronounce the letter 'L' with your tongue behind your upper teeth."

[0973] Audio feedback: Includes audio samples of correct pronunciation.

[0974] Sending and Viewing Feedback

[0975] The server sends the generated feedback to the device, which then analyzes the received feedback and provides it to the user visually and audibly, allowing the user to understand the problems with their pronunciation and immediately try to correct them.

[0976] As a concrete example, consider a user saying the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it, identifies pronunciation errors, and provides feedback to help improve. For example, if the system determines that the "o" sound in "Hello" is not pronounced correctly, it immediately provides corrective feedback to the user.

[0977] Also, examples of prompts to use this system effectively include:

[0978] "Please say the following English sentence: 'Hello, how can I help you?'"

[0979] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[0980] It can be provided as such.

[0981] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0982] Step 1:

[0983] The user speaks the target phrase or word into the microphone.

[0984] Input: User's voice

[0985] Output: Analog audio signal

[0986] In this step, audio is captured as the user speaks, for example, "Hello, how can I help you?"

[0987] Step 2:

[0988] The terminal converts the captured analog audio signal into digital audio data and stores it locally.

[0989] Input: Analog audio signal

[0990] Output: Digital audio data (WAV or MP3 format)

[0991] The device converts the analog audio signal received through the microphone into digital format and saves it in WAV or MP3 format.

[0992] Step 3:

[0993] The terminal transmits the converted digital audio data to the server.

[0994] Input: Digital audio data (WAV or MP3 format)

[0995] Output: Audio data sent to the server

[0996] The terminal uploads the voice data to the server via the network.

[0997] Step 4:

[0998] The server preprocesses the received audio data, performing noise removal, normalization, and volume adjustment.

[0999] Input: Transmitted digital audio data

[1000] Output: Preprocessed audio data

[1001] The server cleans the audio data by applying noise reduction filters, normalizing the volume, etc.

[1002] Step 5:

[1003] The server performs speech recognition on the preprocessed audio data, converting it to text and simultaneously extracting audio characteristics (pitch, intensity, tempo).

[1004] Input: Preprocessed audio data

[1005] Output: Text data and voice characteristics data

[1006] The server uses speech recognition algorithms to analyze the audio and convert it to text, while also extracting audio characteristics such as pitch and intensity.

[1007] Step 6:

[1008] The server compares the extracted speech characteristics with reference speech characteristics to identify mispronunciations.

[1009] Input: Voice characteristic data and reference voice characteristic data

[1010] Output: Mispronunciation information

[1011] The server compares and analyzes the reference voice data with the user's voice characteristics to identify incorrect pronunciations.

[1012] Step 7:

[1013] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[1014] Input: Mispronunciation information

[1015] Output: Feedback information (visual, text, audio)

[1016] The server generates specific feedback to help improve mispronunciations, including diagrams, detailed instructions, and audio samples of correct pronunciation.

[1017] Step 8:

[1018] The server transmits the generated feedback to the terminal, which displays the feedback to the user.

[1019] Input: Feedback information

[1020] Output: Visual and auditory feedback to the user

[1021] The feedback information sent from the server to the device is analyzed and displayed on the device. The user receives feedback from the screen and speaker and uses it to improve their pronunciation.

[1022] As a concrete example, consider a user uttering the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it to identify pronunciation errors, and provides real-time feedback for improvement.

[1023] An example prompt is:

[1024] "Please say the following English sentence: 'Hello, how can I help you?'"

[1025] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[1026] 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.

[1027] This system analyzes the user's voice input in real time and provides feedback to help them master correct pronunciation. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions and provides appropriate feedback according to the user's emotions.

[1028] Audio input and capture

[1029] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[1030] Generate audio data and send it to the server

[1031] The device captures the user's speech and stores it locally as digital audio data, which is then converted into an appropriate format (e.g., WAV, MP3) and sent to the server.

[1032] Voice analysis and speech recognition

[1033] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[1034] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[1035] Compare pronunciations and identify errors

[1036] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[1037] Emotion recognition

[1038] The server uses an emotion engine to recognize emotions from the user's voice data. The emotion engine analyzes voice characteristics (e.g., pitch, intensity, rate) to identify the user's emotional state (e.g., happy, sad, anger).

[1039] Generate feedback

[1040] The server generates feedback based on the identified pronunciation errors and the user's emotional state. The feedback is provided in the following format:

[1041] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[1042] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[1043] Audio feedback: Includes audio samples of correct pronunciation.

[1044] If the emotion engine detects negative emotions (e.g., frustration, sadness), the server generates feedback with encouraging messages, such as "Great progress! Try a little harder next time!"

[1045] Sending and Viewing Feedback

[1046] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and their emotional state and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[1047] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. Furthermore, by providing feedback that takes into account the user's emotions, it helps maintain motivation. This invention will significantly improve the efficiency and effectiveness of phonetic learning.

[1048] The processing flow will be explained below.

[1049] Step 1:

[1050] User: Speak into the device's microphone. For example, if you want to learn the word "Hello," speak "Hello" into the device.

[1051] Step 2:

[1052] Device: Captures the user's speech and stores it locally as digital audio data.

[1053] Step 3:

[1054] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[1055] Step 4:

[1056] Terminal: Sends the converted audio data to the server.

[1057] Step 5:

[1058] Server: Receives the voice data sent from the terminal.

[1059] Step 6:

[1060] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[1061] Step 7:

[1062] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[1063] Step 8:

[1064] Server: Compares the recognized text with reference pronunciation data.

[1065] Step 9:

[1066] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[1067] Step 10:

[1068] Server: Inputs the user's voice data into the emotion engine and recognizes the user's emotions (e.g., joy, sadness, anger).

[1069] Step 11:

[1070] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors and the user's emotional state. For example, it generates feedback including instructions to correct the "e" sound in "Hello" and an encouraging message if the user is showing negative emotions.

[1071] Step 12:

[1072] Server: Sends the generated feedback to the device.

[1073] Step 13:

[1074] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[1075] Step 14:

[1076] User: Attempts to pronounce again based on the feedback provided, thereby repeating the process and improving pronunciation.

[1077] For example, if a user says "Hello" and the server determines that the "e" sound is close to "u," and the emotion engine recognizes that the user is frustrated, the system will provide feedback such as, "Open your lips a little and lower your tongue a little. Then try again. That's great progress!" In this way, the system effectively supports pronunciation correction while taking the user's emotions into consideration.

[1078] Example 2

[1079] 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."

[1080] In learning pronunciation, it is difficult for users to accurately understand and correct their pronunciation errors. Maintaining motivation during the learning process is also a challenge. Furthermore, there is currently a lack of systems that provide appropriate feedback based on the user's emotional state.

[1081] 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.

[1082] In this invention, the server includes means for capturing input voice and generating voice data, means for transmitting the voice data to the server, and means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server, thereby enabling the server to analyze the voice data, identify pronunciation errors, and further recognize emotions.

[1083] "Means for capturing input voice and generating voice data" refers to the process of capturing voice spoken by a user in real time through the device's microphone and saving it as digital voice data.

[1084] The "means for transmitting the audio data to the server" refers to a method for transmitting digital audio data stored in the terminal to the server via a network.

[1085] "Means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server" refers to the function of performing voice recognition on the voice data received by the server, and analyzing and extracting characteristics such as pitch, intensity, and tempo from the voice data.

[1086] The "means for comparing the characteristics of the voice with reference voice characteristics and identifying pronunciation errors" refers to a process for comparing the characteristics of the analyzed voice with reference correct voice characteristics and identifying the errors.

[1087] "Means for generating feedback for pronunciation improvement based on the pronunciation errors" refers to a method for generating feedback including advice and instructions for the user to pronounce correctly based on the identified pronunciation errors.

[1088] The "means for transmitting the feedback to the terminal and displaying it to the user" refers to a method for transmitting the generated feedback from the server to the terminal and presenting it to the user visually or audibly at the terminal.

[1089] "Means for recognizing emotions from the user's voice and generating appropriate feedback based on the recognized emotions" refers to a process of analyzing the user's voice characteristics to recognize their emotional state and creating feedback according to that emotion.

[1090] This invention is a system that supports users in learning pronunciation in real time. It mainly involves a series of processes: voice input, voice data generation and transmission, voice recognition and analysis, pronunciation comparison and error identification, emotion recognition, and feedback generation and transmission.

[1091] Voice input

[1092] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn "Hello," they speak "Hello" into the device. The device captures this speech in real time.

[1093] Generates audio data and sends it to the server

[1094] The device stores the captured audio as digital data and converts it to an appropriate format, such as WAV or MP3, and transmits the converted audio data over the network to a server, along with metadata such as the user ID and a timestamp.

[1095] Voice analysis and speech recognition

[1096] The server processes the audio data received from the device. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. This preprocessing is a fundamental step to obtain clean audio data. Next, it converts the audio data into text using a speech recognition engine (e.g., Google Cloud Speech-to-Text API). In addition, it analyzes and extracts audio characteristics (e.g., pitch, intensity, tempo).

[1097] Compare pronunciations and identify errors

[1098] The server compares the parsed text with a reference pronunciation database containing pronunciations from native speakers. The comparison identifies any pronunciation errors the user may have made. For example, if the "e" in "Hello" sounds more like a "u," that part is recorded as an error.

[1099] Emotion recognition

[1100] The server analyzes the voice data through an emotion recognition engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. It analyzes voice characteristics such as pitch, intensity, and speed to determine whether the user is happy, sad, or angry.

[1101] Generate feedback

[1102] The server generates feedback based on the identified pronunciation errors and the user's emotional state, and the feedback is provided in visual, text, and audio formats.

[1103] Visual feedback: Diagrams or videos showing lip and tongue position and mouth opening (e.g., educational video link).

[1104] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your top teeth to pronounce the letter 'L'").

[1105] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker).

[1106] If the emotion recognition engine detects a negative emotion, the server adds an encouraging message, such as "Great progress! Try a little harder next time!"

[1107] Sending and Viewing Feedback

[1108] The server sends the generated feedback to the device, which analyzes it and provides it visually and audibly to the user, allowing the user to understand their pronunciation problems and try to correct them in real time.

[1109] As a concrete example, if a user wants to learn the phrase "Thank you," the entire process is carried out sequentially, starting with voice input, followed by generation and transmission of voice data, analysis and recognition, identification of pronunciation errors, emotion recognition, feedback generation, and finally transmission and display of the feedback.

[1110] Example prompts for generative AI models

[1111] Explain the steps to analyze the voice data of a user saying "Thank you" into the device's microphone, recognize pronunciation errors and the user's emotional state, and generate and display appropriate feedback.

[1112] This system allows you to effectively learn correct pronunciation without the need for special equipment or expensive lessons. It also provides emotionally sensitive feedback, making it easier to maintain motivation. This invention is a means to significantly improve the efficiency and effectiveness of phonetic learning.

[1113] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1114] Step 1: Audio Input and Capture

[1115] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn the word "Hello," they speak "Hello." This is the user's voice input. The device captures this voice in real time and saves it as digital voice data. The input is the user's voice, and the output is digital voice data. At this stage, voice capture software is used.

[1116] Step 2: Generate audio data and send it to the server

[1117] The device converts the captured audio into an appropriate format (e.g., WAV, MP3). This converted digital audio data is then sent to the server. The input is the captured audio data, and the output is the converted audio data. The converted audio also contains metadata such as the user ID and a timestamp. Audio format conversion and network communication software is used at this stage.

[1118] Step 3: Voice analysis and speech recognition

[1119] The server analyzes the received audio data. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. Then, it uses a speech recognition engine to convert the audio data to text. At the same time, it analyzes and extracts audio characteristics (pitch, intensity, tempo, etc.). The input is the formatted audio data, and the output is text data and audio characteristic data. At this stage, audio processing and speech recognition software is used. Preprocessing includes noise filtering algorithms and signal normalization algorithms.

[1120] Step 4: Compare pronunciation and identify errors

[1121] The server compares the parsed text to a reference pronunciation database. This reference data contains pronunciations from native speakers. The comparison identifies pronunciation errors. For example, if the "e" in "Hello" is close to a "u," it is marked as an error. The input is text data and voice characteristics data, and the output is data indicating pronunciation errors. This stage involves comparison algorithms and database search software.

[1122] Step 5: Recognize emotions

[1123] The server uses the voice data to recognize the user's emotions. An emotion recognition engine is used to analyze voice characteristics (pitch, intensity, speed, etc.) to identify the user's emotional state. For example, emotions such as joy, sadness, and anger can be identified from the analysis results. The input is the voice characteristic data, and the output is the recognized emotional state. At this stage, an emotion recognition algorithm is used.

[1124] Step 6: Generate feedback

[1125] The server generates feedback based on the identified pronunciation errors and the recognized emotional state. The feedback is provided in the following form:

[1126] Visual feedback: illustrations and videos showing lip and tongue position and mouth opening

[1127] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your upper teeth to pronounce the letter 'L'")

[1128] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker)

[1129] If the emotion recognition engine detects a negative emotion, it adds an encouraging message. The input is the pronunciation error data and the emotional state data, and the output is the generated feedback. At this stage, a feedback generation algorithm is used.

[1130] Step 7: Submit and view feedback

[1131] The server sends the generated feedback to the device. The device receives this feedback and displays it visually and audibly to the user. For example, an on-screen diagram showing the correct lip and tongue positions and an audio playback of the correct pronunciation. The input is the generated feedback, and the output is the visual and audible information presented to the user. This is where communication software and user interface software come into play.

[1132] Through these steps, the system can efficiently improve the user's pronunciation and maintain their motivation to learn.

[1133] (Application example 2)

[1134] 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."

[1135] The present invention relates to a system that improves the efficiency of a user's pronunciation practice and provides appropriate feedback according to the user's emotional state. Conventional pronunciation learning systems can identify pronunciation errors, but they cannot take the user's emotions or state into account, making it difficult to maintain motivation for learning. Therefore, in order to maximize the effectiveness of pronunciation learning, a system that provides feedback according to the user's emotions is needed.

[1136] The identification process by the identification 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 including an emotion engine for recognizing emotions from voice data, means for generating feedback according to the emotions, and means for performing preprocessing such as noise removal, normalization, and volume adjustment. This makes it possible to identify errors in the user's voice pronunciation and provide feedback according to the user's emotional state.

[1137] A "means for generating voice data" is a device or software that allows a user to capture voice input and store it as voice data in digital form.

[1138] The "means for transmitting voice data to a server" refers to a device or software for transferring the generated voice data to a server via a network.

[1139] "Means for analyzing audio data, performing speech recognition, and extracting audio characteristics" refers to devices or software that apply a speech recognition algorithm to audio data within a server and analyze audio characteristics (e.g., pitch, intensity, tempo, etc.).

[1140] The "means for comparing voice characteristics with reference voice characteristics and identifying pronunciation errors" refers to a device or software that compares the user's voice characteristics with preset reference voice characteristics and identifies incorrect pronunciation parts.

[1141] A "means for generating feedback for pronunciation improvement" is a device or software that generates specific instructions or samples for correcting pronunciation errors.

[1142] The "means for transmitting feedback to a terminal and displaying it to a user" refers to a device or software for transmitting the generated feedback to a user's terminal via a network and displaying it.

[1143] The "means including an emotion engine for recognizing emotions" refers to a device or software for analyzing emotions from voice data and identifying the emotional state of a user.

[1144] The "means for generating emotional feedback" refers to a device or software for generating feedback including appropriate encouragement, warnings, and specific instructions based on the user's emotional state.

[1145] The present invention relates to a system that identifies user pronunciation errors and provides real-time emotional feedback. The system generates speech data, transmits it to a server, and analyzes it to perform speech recognition and extract speech characteristics. Furthermore, the system uses an emotion engine to recognize the user's emotions and generate specific feedback accordingly.

[1146] Processing contents to be performed by the program

[1147] The invention uses a "device," such as a microphone or smartphone, to capture a user's voice input. The device stores the captured audio as digital audio data and then transmits this data to a server, where it is converted into an appropriate format (e.g., WAV or MP3).

[1148] The server analyzes the audio data using an installed speech recognition engine (e.g., Google Speech-to-Text API) to convert the speech to text and extracts voice characteristics (e.g., pitch, intensity, tempo). This data is compared with baseline voice characteristics to identify pronunciation errors.

[1149] Furthermore, an emotion recognition engine (e.g., IBM Watson Tone Analyzer) is used to analyze the user's emotions from the voice data, identifying the user's emotional state (e.g., joy, sadness, anger, etc.) and generating feedback accordingly.

[1150] Providing Feedback

[1151] The form of feedback includes at least one of visual feedback, text feedback, and audio feedback. Visual feedback may include animations or videos showing lip movements and tongue positions that indicate correct pronunciation. Text feedback may include specific instructions such as "Put your tongue behind your upper teeth to pronounce 'L'." Audio feedback may include audio samples of correct pronunciation.

[1152] Specific examples

[1153] For example, suppose a factory operator says "start" when operating a robot. This voice input is captured by the device's microphone and sent to a server. The server analyzes the speech of "start" and detects that a specific part (such as the "ta" sound) is not pronounced correctly. It then uses an emotion recognition engine to determine that the operator is slightly nervous. As a result, the system displays an encouraging message along with animations and voice samples showing examples of how to pronounce "ta" correctly.

[1154] Prompt Sentence Examples

[1155] Please tell us about the voice-guided operation support system in factories.

[1156] This system helps factory operators give accurate voice instructions to robots and provides feedback according to the operator's emotional state. It includes steps for voice recognition, pronunciation analysis, emotion recognition, and feedback generation.

[1157] This makes it easier for users to learn correct pronunciation, maintains their motivation to learn, and improves factory operation efficiency.

[1158] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1159] Step 1:

[1160] (audio input and capture)

[1161] The user speaks into a microphone. For example, a factory operator says "start." This voice is captured by the device's microphone and saved as digital voice data. The input is real-time voice data, and the output is saved digital voice data. Specifically, a microphone and voice capture software installed on a smartphone or dedicated device is used.

[1162] Step 2:

[1163] (Sending audio data)

[1164] The device converts the captured digital audio data into an appropriate format (e.g., WAV, MP3) and transmits it to a server via the Internet. The input is the stored digital audio data, and the output is the audio data transmitted to the server. Specifically, a communication module and a data transfer protocol are used.

[1165] Step 3:

[1166] (Audio pre-processing)

[1167] The server performs preprocessing on the received audio data. Specifically, it performs noise removal, normalization, volume adjustment, etc. The input is the audio data sent to the server, and the output is the audio data after preprocessing. As a specific example, an audio processing library (e.g., WebRTC) is used.

[1168] Step 4:

[1169] (Voice Recognition)

[1170] The server performs speech recognition on the preprocessed audio data and converts the audio into text. The input is the preprocessed audio data, and the output is text generated from the audio. Specifically, the server uses the Google Speech-to-Text API or similar.

[1171] Step 5:

[1172] (Extraction of voice characteristics)

[1173] The server extracts audio features (e.g., pitch, intensity, tempo) from audio data. The input is preprocessed audio data, and the output is the extracted audio features. Specifically, it uses an audio analysis library (e.g., Librosa).

[1174] Step 6:

[1175] (Identifying mispronunciations)

[1176] The server compares the extracted speech characteristics with the reference speech characteristics to identify pronunciation errors. The input is the speech characteristics and the reference speech characteristics, and the output is the identified pronunciation errors. Specifically, the comparison process is performed using an algorithm.

[1177] Step 7:

[1178] (Emotion recognition)

[1179] The server uses an emotion engine to recognize emotions from voice data. The input is the voice data, and the output is the recognized emotional state. Specifically, it uses IBM Watson Tone Analyzer.

[1180] Step 8:

[1181] (Generating feedback)

[1182] The server generates feedback based on the identified mispronunciations and the recognized emotional state. The feedback may be in the form of visual, text, or audio. The input is the mispronunciation and the emotional state, and the output is the generated feedback. Specifically, a feedback generation module is used.

[1183] Step 9:

[1184] (Sending and viewing feedback)

[1185] The server sends the generated feedback to the device. The device displays and plays the received feedback to the user. The input is the generated feedback, and the output is the feedback provided to the user. Specifically, a display or speaker is used.

[1186] 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.

[1187] 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.

[1188] 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.

[1189] [Fourth embodiment]

[1190] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1191] 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.

[1192] 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).

[1193] 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.

[1194] 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.

[1195] 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).

[1196] 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.

[1197] 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.

[1198] 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.

[1199] 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.

[1200] 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.

[1201] 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.

[1202] 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."

[1203] This system analyzes the user's input speech in real time and provides feedback to help the user master correct pronunciation. A specific embodiment of this system will be described below.

[1204] Audio input and capture

[1205] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[1206] Generate audio data and send it to the server

[1207] The device captures the user's speech, stores it locally as digital audio data, converts it into an appropriate format (e.g., WAV, MP3), and then transmits it to the server.

[1208] Voice analysis and speech recognition

[1209] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[1210] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[1211] Compare pronunciations and identify errors

[1212] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[1213] Generate feedback

[1214] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[1215] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[1216] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[1217] Audio feedback: Includes audio samples of correct pronunciation.

[1218] Sending and Viewing Feedback

[1219] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[1220] This system allows people to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention will be a means to significantly improve the efficiency of phonetic learning.

[1221] The processing flow will be explained below.

[1222] Step 1:

[1223] User: Speak into the device's microphone. For example, say "Hello."

[1224] Step 2:

[1225] Device: Captures the user's speech and stores it locally as digital audio data.

[1226] Step 3:

[1227] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[1228] Step 4:

[1229] Terminal: Sends the converted audio data to the server.

[1230] Step 5:

[1231] Server: Receives the voice data sent from the terminal.

[1232] Step 6:

[1233] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[1234] Step 7:

[1235] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[1236] Step 8:

[1237] Server: Compares the recognized text with reference pronunciation data.

[1238] Step 9:

[1239] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[1240] Step 10:

[1241] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors. The types of feedback include visual feedback, text feedback, and audio feedback.

[1242] Step 11:

[1243] Server: Sends the generated feedback to the device.

[1244] Step 12:

[1245] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[1246] Step 13:

[1247] User: Based on the feedback provided, they correct and try again, which returns them to step 1 and repeats the process to improve their pronunciation.

[1248] Example 1

[1249] 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."

[1250] Conventional speech recognition systems lack the ability to provide specific feedback to help users master correct pronunciation. Furthermore, they lack the means to suggest specific correction methods for specific pronunciation errors, making it difficult for users to effectively correct their pronunciation.

[1251] 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.

[1252] In this invention, the server includes means for capturing input speech and generating speech data, means for transmitting the speech data to the server, means for analyzing the speech data, performing speech recognition, and extracting speech characteristics in the server, means for comparing the speech characteristics with reference speech characteristics and identifying pronunciation errors, means for generating visual feedback, text feedback, and speech feedback for pronunciation improvement based on the pronunciation errors, and means for transmitting the feedback to a terminal and displaying it to the user, thereby enabling the user to efficiently correct their pronunciation.

[1253] "Input voice" refers to the voice spoken by the user into the microphone.

[1254] "Audio Data" refers to data for storing or transmitting captured audio in digital form.

[1255] "Server" refers to a computer system that analyzes voice data, recognizes voice, and performs post-voice recognition processing.

[1256] "Speech recognition" refers to the process of converting input voice data into character string information.

[1257] "Audio characteristics" refers to characteristic information such as pitch, intensity, and tempo contained in audio data.

[1258] "Reference speech characteristics" refers to speech characteristics information collected and recorded based on correct pronunciation.

[1259] "Mispronunciation" refers to a mispronunciation portion identified by comparing the characteristics of the input speech with the reference speech characteristics.

[1260] "Visual feedback" refers to feedback in the form of diagrams or videos that show lip and tongue shape, mouth opening, etc. to help users improve their pronunciation.

[1261] "Textual feedback" refers to written feedback that includes specific instructions to improve a user's pronunciation.

[1262] "Audio feedback" refers to feedback that provides an audio sample of correct pronunciation.

[1263] "Terminal" refers to a device (smartphone, PC, etc.) through which a user inputs voice.

[1264] "Preprocessing" refers to performing processes such as noise removal, normalization, and volume adjustment on audio data.

[1265] This invention is a system that analyzes user-input speech in real time and provides feedback to help users learn correct pronunciation.

[1266] A specific embodiment of this system is described below: A user speaks a target word or sentence into the microphone of the terminal. For example, if a user wants to learn the word "Hello," he speaks "Hello" into the terminal.

[1267] Audio input and capture

[1268] The user speaks into the device's microphone. The device's microphone captures what the user says and stores it locally as digital audio data. The audio data is converted to an appropriate format (e.g., WAV, MP3) using an audio library (e.g., PyDub).

[1269] Generate audio data and send it to the server

[1270] The device sends the saved audio data to the server using the HTTP protocol.

[1271] Voice analysis and speech recognition

[1272] The server receives the audio data sent from the device. After receiving the data, it first performs preprocessing on the audio data. This includes noise removal, normalization, and volume adjustment. An audio signal processing library (e.g., LibROSA) is used for this processing.

[1273] Speech Recognition and Feature Extraction

[1274] After preprocessing is complete, the server performs speech recognition on the audio data and converts the audio into text. A speech recognition API (e.g., Google Cloud Speech-to-Text API) is used for speech recognition. At the same time, voice characteristics (pitch, intensity, tempo, etc.) are extracted. LibROSA is used for feature extraction.

[1275] Compare pronunciations and identify errors

[1276] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[1277] Generate feedback

[1278] The server generates feedback to help the user correct their pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[1279] Visual feedback: Diagrams and videos showing lip and tongue shape and mouth opening, using video editing software (e.g., Adobe After Effects, Blender).

[1280] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[1281] Audio feedback: Includes audio samples of correct pronunciation.

[1282] Sending and Viewing Feedback

[1283] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[1284] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. This invention is a means of significantly improving the efficiency of phonetic learning.

[1285] Example 1

[1286] The user says "Hello" and the device's microphone captures the audio. The device converts the captured audio to WAV format and sends it to the server. The server receives the audio data and performs noise reduction and normalization. The server uses the Google Cloud Speech-to-Text API to convert "Hello" to text and simultaneously calculates pitch and intensity. The server compares "Hello" with standard pronunciation data and determines that the "e" sound is closer to a "u." The server generates text feedback such as "Place your tongue behind your upper teeth to pronounce 'L'" along with an audio sample of the correct pronunciation and sends it to the device. The device presents this to the user, who then practices their pronunciation again based on the feedback.

[1287] Prompt Sentence Examples

[1288] "The user said 'Hello'. Given this audio data, please compare it with a standard pronunciation to generate correct feedback."

[1289] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1290] Step 1: Capturing Audio Input

[1291] The user speaks into the microphone of the terminal.

[1292] Input: User's voice, saying "Hello".

[1293] The device captures the spoken voice through a microphone and stores it as digital voice data.

[1294] Output: Captured digital audio data (WAV, MP3 format, etc.).

[1295] What it does: The device's audio API captures audio input from the microphone and saves it as an audio file.

[1296] Step 2: Generate and send audio data

[1297] The terminal transmits the stored voice data to the server.

[1298] Input: Captured digital audio data.

[1299] The terminal uses the HTTP protocol to send the audio data to the server.

[1300] Output: The audio data sent to the server.

[1301] What it does: The device converts the voice data into the appropriate format and sends the data to a server over the Internet.

[1302] Step 3: Preprocessing the audio data

[1303] The server receives the voice data transmitted from the terminal.

[1304] Input: Audio data sent from the device.

[1305] The server performs preprocessing such as noise removal, normalization, and volume adjustment on the received audio data.

[1306] Output: Preprocessed audio data.

[1307] What it does: A server-side audio signal processing library (e.g., LibROSA) removes noise from the audio data and normalizes the volume level to adjust it to an appropriate sound quality.

[1308] Step 4: Speech recognition and feature extraction

[1309] The server analyzes the preprocessed audio data.

[1310] Input: Preprocessed audio data.

[1311] The server performs speech recognition and converts the audio data into text, while simultaneously extracting audio characteristics (pitch, intensity, tempo, etc.).

[1312] Output: Text data and speech characteristics data generated by speech recognition.

[1313] Specific operation: The server uses a speech recognition API (e.g., Google Cloud Speech-to-Text API) to convert the speech into text, and calculates speech feature information using a library such as LibROSA.

[1314] Step 5: Compare pronunciation and identify errors

[1315] The server compares the recognized text with the reference pronunciation data.

[1316] Input: Text data generated by speech recognition, reference pronunciation data.

[1317] The server identifies which parts of the user's pronunciation are incorrect.

[1318] Output: Mispronunciation information.

[1319] Specific operation: The server refers to a dataset of correct pronunciations collected in advance, compares it with the user's pronunciation, identifies incorrect parts, and records them.

[1320] Step 6: Generate feedback

[1321] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[1322] Input: Mispronunciation information.

[1323] The server generates visual, textual and audio feedback.

[1324] Output: Feedback data (visual, text, audio).

[1325] Specific behaviors: Visual feedback is generated using video editing software (e.g., Adobe After Effects), text feedback is written with specific instructions, and audio feedback uses pre-recorded samples of correct pronunciation.

[1326] Step 7: Submit and view feedback

[1327] The server transmits the generated feedback to the terminal.

[1328] Input: Feedback data (visual, text, audio).

[1329] The device analyzes the received feedback and provides it to the user visually and audibly.

[1330] Output: Feedback provided to the user.

[1331] Specific operation: The device presents video, text, and audio data through a user interface, and the user corrects their pronunciation based on the feedback.

[1332] (Application example 1)

[1333] 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."

[1334] Conventional speech analysis systems have primarily been used for personal pronunciation practice. However, effective means for improving the quality of speech communication between store clerks and customers in brick-and-mortar stores are limited. This results in missed opportunities for improving customer satisfaction and improving store clerk skills. The present invention aims to solve this problem by providing a system that evaluates the pronunciation quality of store clerk interactions in real time and provides effective feedback.

[1335] 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.

[1336] In this invention, the server includes means for capturing and analyzing voice data, means for performing voice recognition and extracting voice characteristics, means for comparing the extracted voice characteristics with reference voice characteristics to identify pronunciation errors, means for generating feedback for pronunciation improvement based on the pronunciation errors, means for transmitting the feedback to a terminal and displaying it to the user as visual, text, and voice feedback, and means for analyzing voices spoken by store clerks when interacting with customers and providing feedback, thereby enabling real-time pronunciation improvement so that store clerks can communicate with customers with appropriate pronunciation.

[1337] "Capture" means acquiring information such as sound or video and saving it as digital data.

[1338] "Audio data" is data that represents audio in digital form.

[1339] A "server" is a centralized management system for processing and storing data.

[1340] "Speech recognition" is the technology that analyzes voice data and converts the content into text or other formats.

[1341] "Audio characteristics" are characteristic attributes associated with an audio signal, including pitch, intensity, tempo, etc.

[1342] "Reference speech characteristics" are reference data used to indicate correct pronunciation.

[1343] A "mispronunciation" is a difference observed between the user's pronunciation and the reference pronunciation.

[1344] "Feedback" is information that informs the user of the evaluation results and areas for improvement.

[1345] A "terminal" is a computing device that can be directly operated by a user.

[1346] A "store associate" refers to someone who provides service to customers in a physical store.

[1347] "Customer service" refers to a series of actions and statements that store clerks make to communicate with customers.

[1348] This system analyzes in real time the speech spoken by store clerks when interacting with customers and provides feedback to help improve pronunciation. A specific embodiment of this system will be described below.

[1349] Audio input and capture

[1350] The user (store clerk) speaks into the microphone. For example, the user speaks the phrase "Hello, how can I help you?"

[1351] Generate audio data and send it to the server

[1352] The device captures the user's speech and stores it as digital audio data, which is then converted to WAV or MP3 format and sent to a server.

[1353] Voice analysis and speech recognition

[1354] After receiving the voice data from the device, the server performs preprocessing on the voice data, specifically noise removal, normalization, and volume adjustment. After preprocessing is complete, the server performs speech recognition on the voice data and converts the voice into text. At the same time, it extracts voice characteristics (pitch, intensity, tempo, etc.).

[1355] Compare pronunciations and identify errors

[1356] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[1357] Generate feedback

[1358] The server generates feedback to correct the user's pronunciation based on the identified pronunciation errors. The feedback is provided in the following format:

[1359] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[1360] Text feedback: Information in written form with specific instructions, such as "Try to pronounce the letter 'L' with your tongue behind your upper teeth."

[1361] Audio feedback: Includes audio samples of correct pronunciation.

[1362] Sending and Viewing Feedback

[1363] The server sends the generated feedback to the device, which then analyzes the received feedback and provides it to the user visually and audibly, allowing the user to understand the problems with their pronunciation and immediately try to correct them.

[1364] As a concrete example, consider a user saying the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it, identifies pronunciation errors, and provides feedback to help improve. For example, if the system determines that the "o" sound in "Hello" is not pronounced correctly, it immediately provides corrective feedback to the user.

[1365] Also, examples of prompts to use this system effectively include:

[1366] "Please say the following English sentence: 'Hello, how can I help you?'"

[1367] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[1368] It can be provided as such.

[1369] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1370] Step 1:

[1371] The user speaks the target phrase or word into the microphone.

[1372] Input: User's voice

[1373] Output: Analog audio signal

[1374] In this step, audio is captured as the user speaks, for example, "Hello, how can I help you?"

[1375] Step 2:

[1376] The terminal converts the captured analog audio signal into digital audio data and stores it locally.

[1377] Input: Analog audio signal

[1378] Output: Digital audio data (WAV or MP3 format)

[1379] The device converts the analog audio signal received through the microphone into digital format and saves it in WAV or MP3 format.

[1380] Step 3:

[1381] The terminal transmits the converted digital audio data to the server.

[1382] Input: Digital audio data (WAV or MP3 format)

[1383] Output: Audio data sent to the server

[1384] The terminal uploads the voice data to the server via the network.

[1385] Step 4:

[1386] The server preprocesses the received audio data, performing noise removal, normalization, and volume adjustment.

[1387] Input: Transmitted digital audio data

[1388] Output: Preprocessed audio data

[1389] The server cleans the audio data by applying noise reduction filters, normalizing the volume, etc.

[1390] Step 5:

[1391] The server performs speech recognition on the preprocessed audio data, converting it to text and simultaneously extracting audio characteristics (pitch, intensity, tempo).

[1392] Input: Preprocessed audio data

[1393] Output: Text data and voice characteristics data

[1394] The server uses speech recognition algorithms to analyze the audio and convert it to text, while also extracting audio characteristics such as pitch and intensity.

[1395] Step 6:

[1396] The server compares the extracted speech characteristics with reference speech characteristics to identify mispronunciations.

[1397] Input: Voice characteristic data and reference voice characteristic data

[1398] Output: Mispronunciation information

[1399] The server compares and analyzes the reference voice data with the user's voice characteristics to identify incorrect pronunciations.

[1400] Step 7:

[1401] The server generates feedback for correcting the user's pronunciation based on the identified pronunciation errors.

[1402] Input: Mispronunciation information

[1403] Output: Feedback information (visual, text, audio)

[1404] The server generates specific feedback to help improve mispronunciations, including diagrams, detailed instructions, and audio samples of correct pronunciation.

[1405] Step 8:

[1406] The server transmits the generated feedback to the terminal, which displays the feedback to the user.

[1407] Input: Feedback information

[1408] Output: Visual and auditory feedback to the user

[1409] The feedback information sent from the server to the device is analyzed and displayed on the device. The user receives feedback from the screen and speaker and uses it to improve their pronunciation.

[1410] As a concrete example, consider a user uttering the phrase "Hello, how can I help you?" The system captures this pronunciation, analyzes it to identify pronunciation errors, and provides real-time feedback for improvement.

[1411] An example prompt is:

[1412] "Please say the following English sentence: 'Hello, how can I help you?'"

[1413] "Practice what you should say to your customers. Start with, 'Welcome to our store!'"

[1414] 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.

[1415] This system analyzes the user's voice input in real time and provides feedback to help them master correct pronunciation. Furthermore, this system is combined with an emotion engine that recognizes the user's emotions and provides appropriate feedback according to the user's emotions.

[1416] Audio input and capture

[1417] The user speaks the target word or sentence into the device's microphone. For example, if the user wants to learn the word "Hello," they speak "Hello" into the device.

[1418] Generate audio data and send it to the server

[1419] The device captures the user's speech and stores it locally as digital audio data, which is then converted into an appropriate format (e.g., WAV, MP3) and sent to the server.

[1420] Voice analysis and speech recognition

[1421] The server receives the voice data sent from the terminal. After receiving the data, it first performs preprocessing on the voice data, such as noise removal, normalization, and volume adjustment.

[1422] After the preprocessing is complete, the server performs speech recognition and converts the speech to text, while simultaneously extracting speech characteristics (pitch, intensity, tempo, etc.).

[1423] Compare pronunciations and identify errors

[1424] The server compares the recognized text with reference pronunciation data. This comparison identifies which parts of the user's pronunciation are incorrect. For example, if the "e" in "Hello" sounds more like a "u," it will be recorded as a mispronunciation.

[1425] Emotion recognition

[1426] The server uses an emotion engine to recognize emotions from the user's voice data. The emotion engine analyzes voice characteristics (e.g., pitch, intensity, rate) to identify the user's emotional state (e.g., happy, sad, anger).

[1427] Generate feedback

[1428] The server generates feedback based on the identified pronunciation errors and the user's emotional state. The feedback is provided in the following format:

[1429] Visual feedback: Diagrams and videos showing lip and tongue shape, mouth opening, etc.

[1430] Text feedback: Information in the form of written instructions with specific instructions, such as "Put your tongue behind your upper teeth to say 'L'."

[1431] Audio feedback: Includes audio samples of correct pronunciation.

[1432] If the emotion engine detects negative emotions (e.g., frustration, sadness), the server generates feedback with encouraging messages, such as "Great progress! Try a little harder next time!"

[1433] Sending and Viewing Feedback

[1434] The server sends the generated feedback to the device. The device analyzes the received feedback and provides it visually and audibly to the user. This allows the user to understand the problems with their pronunciation and their emotional state and try to correct them immediately. For example, if the user pronounces "Hello" again, the system will repeat the same process and continue providing feedback.

[1435] This system allows users to effectively learn correct pronunciation without the need for special equipment or expensive lessons. Furthermore, by providing feedback that takes into account the user's emotions, it helps maintain motivation. This invention will significantly improve the efficiency and effectiveness of phonetic learning.

[1436] The processing flow will be explained below.

[1437] Step 1:

[1438] User: Speak into the device's microphone. For example, if you want to learn the word "Hello," speak "Hello" into the device.

[1439] Step 2:

[1440] Device: Captures the user's speech and stores it locally as digital audio data.

[1441] Step 3:

[1442] On your device: Convert the stored audio data to the appropriate format (e.g., WAV, MP3).

[1443] Step 4:

[1444] Terminal: Sends the converted audio data to the server.

[1445] Step 5:

[1446] Server: Receives the voice data sent from the terminal.

[1447] Step 6:

[1448] Server: Preprocesses the audio data, specifically performing noise removal, normalization, volume adjustment, etc.

[1449] Step 7:

[1450] Server: The preprocessed speech data is input into a speech recognition engine, which converts it to text and extracts speech characteristics (pitch, intensity, tempo, etc.).

[1451] Step 8:

[1452] Server: Compares the recognized text with reference pronunciation data.

[1453] Step 9:

[1454] Server: Identifies mispronunciations based on speech characteristics and comparing them to standards of correct pronunciation. For example, determining that the "e" in "Hello" sounds similar to a "u."

[1455] Step 10:

[1456] Server: Inputs the user's voice data into the emotion engine and recognizes the user's emotions (e.g., joy, sadness, anger).

[1457] Step 11:

[1458] Server: Generates feedback for pronunciation improvement based on the identified pronunciation errors and the user's emotional state. For example, it generates feedback including instructions to correct the "e" sound in "Hello" and an encouraging message if the user is showing negative emotions.

[1459] Step 12:

[1460] Server: Sends the generated feedback to the device.

[1461] Step 13:

[1462] Terminal: Analyzes the feedback received from the server and provides it to the user visually and audibly, for example by playing diagrams or videos, playing audio samples, or displaying text messages.

[1463] Step 14:

[1464] User: Attempts to pronounce again based on the feedback provided, thereby repeating the process and improving pronunciation.

[1465] For example, if a user says "Hello" and the server determines that the "e" sound is close to "u," and the emotion engine recognizes that the user is frustrated, the system will provide feedback such as, "Open your lips a little and lower your tongue a little. Then try again. That's great progress!" In this way, the system effectively supports pronunciation correction while taking the user's emotions into consideration.

[1466] Example 2

[1467] 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."

[1468] In learning pronunciation, it is difficult for users to accurately understand and correct their pronunciation errors. Maintaining motivation during the learning process is also a challenge. Furthermore, there is currently a lack of systems that provide appropriate feedback based on the user's emotional state.

[1469] 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.

[1470] In this invention, the server includes means for capturing input voice and generating voice data, means for transmitting the voice data to the server, and means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server, thereby enabling the server to analyze the voice data, identify pronunciation errors, and further recognize emotions.

[1471] "Means for capturing input voice and generating voice data" refers to the process of capturing voice spoken by a user in real time through the device's microphone and saving it as digital voice data.

[1472] The "means for transmitting the audio data to the server" refers to a method for transmitting digital audio data stored in the terminal to the server via a network.

[1473] "Means for analyzing the voice data, performing voice recognition, and extracting voice characteristics in the server" refers to the function of performing voice recognition on the voice data received by the server, and analyzing and extracting characteristics such as pitch, intensity, and tempo from the voice data.

[1474] The "means for comparing the characteristics of the voice with reference voice characteristics and identifying pronunciation errors" refers to a process for comparing the characteristics of the analyzed voice with reference correct voice characteristics and identifying the errors.

[1475] "Means for generating feedback for pronunciation improvement based on the pronunciation errors" refers to a method for generating feedback including advice and instructions for the user to pronounce correctly based on the identified pronunciation errors.

[1476] The "means for transmitting the feedback to the terminal and displaying it to the user" refers to a method for transmitting the generated feedback from the server to the terminal and presenting it to the user visually or audibly at the terminal.

[1477] "Means for recognizing emotions from the user's voice and generating appropriate feedback based on the recognized emotions" refers to a process of analyzing the user's voice characteristics to recognize their emotional state and creating feedback according to that emotion.

[1478] This invention is a system that supports users in learning pronunciation in real time. It mainly involves a series of processes: voice input, voice data generation and transmission, voice recognition and analysis, pronunciation comparison and error identification, emotion recognition, and feedback generation and transmission.

[1479] Voice input

[1480] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn "Hello," they speak "Hello" into the device. The device captures this speech in real time.

[1481] Generates audio data and sends it to the server

[1482] The device stores the captured audio as digital data and converts it to an appropriate format, such as WAV or MP3, and transmits the converted audio data over the network to a server, along with metadata such as the user ID and a timestamp.

[1483] Voice analysis and speech recognition

[1484] The server processes the audio data received from the device. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. This preprocessing is a fundamental step to obtain clean audio data. Next, it converts the audio data into text using a speech recognition engine (e.g., Google Cloud Speech-to-Text API). In addition, it analyzes and extracts audio characteristics (e.g., pitch, intensity, tempo).

[1485] Compare pronunciations and identify errors

[1486] The server compares the parsed text with a reference pronunciation database containing pronunciations from native speakers. The comparison identifies any pronunciation errors the user may have made. For example, if the "e" in "Hello" sounds more like a "u," that part is recorded as an error.

[1487] Emotion recognition

[1488] The server analyzes the voice data through an emotion recognition engine (e.g., IBM Watson Tone Analyzer) to identify the user's emotional state. It analyzes voice characteristics such as pitch, intensity, and speed to determine whether the user is happy, sad, or angry.

[1489] Generate feedback

[1490] The server generates feedback based on the identified pronunciation errors and the user's emotional state, and the feedback is provided in visual, text, and audio formats.

[1491] Visual feedback: Diagrams or videos showing lip and tongue position and mouth opening (e.g., educational video link).

[1492] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your top teeth to pronounce the letter 'L'").

[1493] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker).

[1494] If the emotion recognition engine detects a negative emotion, the server adds an encouraging message, such as "Great progress! Try a little harder next time!"

[1495] Sending and Viewing Feedback

[1496] The server sends the generated feedback to the device, which analyzes it and provides it visually and audibly to the user, allowing the user to understand their pronunciation problems and try to correct them in real time.

[1497] As a concrete example, if a user wants to learn the phrase "Thank you," the entire process is carried out sequentially, starting with voice input, followed by generation and transmission of voice data, analysis and recognition, identification of pronunciation errors, emotion recognition, feedback generation, and finally transmission and display of the feedback.

[1498] Example prompts for generative AI models

[1499] Explain the steps to analyze the voice data of a user saying "Thank you" into the device's microphone, recognize pronunciation errors and the user's emotional state, and generate and display appropriate feedback.

[1500] This system allows you to effectively learn correct pronunciation without the need for special equipment or expensive lessons. It also provides emotionally sensitive feedback, making it easier to maintain motivation. This invention is a means to significantly improve the efficiency and effectiveness of phonetic learning.

[1501] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1502] Step 1: Audio Input and Capture

[1503] The user speaks the words or sentences they want to learn into the device's microphone. For example, if they want to learn the word "Hello," they speak "Hello." This is the user's voice input. The device captures this voice in real time and saves it as digital voice data. The input is the user's voice, and the output is digital voice data. At this stage, voice capture software is used.

[1504] Step 2: Generate audio data and send it to the server

[1505] The device converts the captured audio into an appropriate format (e.g., WAV, MP3). This converted digital audio data is then sent to the server. The input is the captured audio data, and the output is the converted audio data. The converted audio also contains metadata such as the user ID and a timestamp. Audio format conversion and network communication software is used at this stage.

[1506] Step 3: Voice analysis and speech recognition

[1507] The server analyzes the received audio data. First, it performs preprocessing such as noise removal, normalization, and volume adjustment. Then, it uses a speech recognition engine to convert the audio data to text. At the same time, it analyzes and extracts audio characteristics (pitch, intensity, tempo, etc.). The input is the formatted audio data, and the output is text data and audio characteristic data. At this stage, audio processing and speech recognition software is used. Preprocessing includes noise filtering algorithms and signal normalization algorithms.

[1508] Step 4: Compare pronunciation and identify errors

[1509] The server compares the parsed text to a reference pronunciation database. This reference data contains pronunciations from native speakers. The comparison identifies pronunciation errors. For example, if the "e" in "Hello" is close to a "u," it is marked as an error. The input is text data and voice characteristics data, and the output is data indicating pronunciation errors. This stage involves comparison algorithms and database search software.

[1510] Step 5: Recognize emotions

[1511] The server uses the voice data to recognize the user's emotions. An emotion recognition engine is used to analyze voice characteristics (pitch, intensity, speed, etc.) to identify the user's emotional state. For example, emotions such as joy, sadness, and anger can be identified from the analysis results. The input is the voice characteristic data, and the output is the recognized emotional state. At this stage, an emotion recognition algorithm is used.

[1512] Step 6: Generate feedback

[1513] The server generates feedback based on the identified pronunciation errors and the recognized emotional state. The feedback is provided in the following form:

[1514] Visual feedback: illustrations and videos showing lip and tongue position and mouth opening

[1515] Text feedback: Sentences with specific instructions (e.g., "Put your tongue behind your upper teeth to pronounce the letter 'L'")

[1516] Audio feedback: Audio samples of correct pronunciation (e.g., from a native speaker)

[1517] If the emotion recognition engine detects a negative emotion, it adds an encouraging message. The input is the pronunciation error data and the emotional state data, and the output is the generated feedback. At this stage, a feedback generation algorithm is used.

[1518] Step 7: Submit and view feedback

[1519] The server sends the generated feedback to the device. The device receives this feedback and displays it visually and audibly to the user. For example, an on-screen diagram showing the correct lip and tongue positions and an audio playback of the correct pronunciation. The input is the generated feedback, and the output is the visual and audible information presented to the user. This is where communication software and user interface software come into play.

[1520] Through these steps, the system can efficiently improve the user's pronunciation and maintain their motivation to learn.

[1521] (Application example 2)

[1522] 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."

[1523] The present invention relates to a system that improves the efficiency of a user's pronunciation practice and provides appropriate feedback according to the user's emotional state. Conventional pronunciation learning systems can identify pronunciation errors, but they cannot take the user's emotions or state into account, making it difficult to maintain motivation for learning. Therefore, in order to maximize the effectiveness of pronunciation learning, a system that provides feedback according to the user's emotions is needed.

[1524] The identification process by the identification 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 including an emotion engine for recognizing emotions from voice data, means for generating feedback according to the emotions, and means for performing preprocessing such as noise removal, normalization, and volume adjustment. This makes it possible to identify errors in the user's voice pronunciation and provide feedback according to the user's emotional state.

[1525] A "means for generating voice data" is a device or software that allows a user to capture voice input and store it as voice data in digital form.

[1526] The "means for transmitting voice data to a server" refers to a device or software for transferring the generated voice data to a server via a network.

[1527] "Means for analyzing audio data, performing speech recognition, and extracting audio characteristics" refers to devices or software that apply a speech recognition algorithm to audio data within a server and analyze audio characteristics (e.g., pitch, intensity, tempo, etc.).

[1528] The "means for comparing voice characteristics with reference voice characteristics and identifying pronunciation errors" refers to a device or software that compares the user's voice characteristics with preset reference voice characteristics and identifies incorrect pronunciation parts.

[1529] A "means for generating feedback for pronunciation improvement" is a device or software that generates specific instructions or samples for correcting pronunciation errors.

[1530] The "means for transmitting feedback to a terminal and displaying it to a user" refers to a device or software for transmitting the generated feedback to a user's terminal via a network and displaying it.

[1531] The "means including an emotion engine for recognizing emotions" refers to a device or software for analyzing emotions from voice data and identifying the emotional state of a user.

[1532] The "means for generating emotional feedback" refers to a device or software for generating feedback including appropriate encouragement, warnings, and specific instructions based on the user's emotional state.

[1533] The present invention relates to a system that identifies user pronunciation errors and provides real-time emotional feedback. The system generates speech data, transmits it to a server, and analyzes it to perform speech recognition and extract speech characteristics. Furthermore, the system uses an emotion engine to recognize the user's emotions and generate specific feedback accordingly.

[1534] Processing contents to be performed by the program

[1535] The invention uses a "device," such as a microphone or smartphone, to capture a user's voice input. The device stores the captured audio as digital audio data and then transmits this data to a server, where it is converted into an appropriate format (e.g., WAV or MP3).

[1536] The server analyzes the audio data using an installed speech recognition engine (e.g., Google Speech-to-Text API) to convert the speech to text and extracts voice characteristics (e.g., pitch, intensity, tempo). This data is compared with baseline voice characteristics to identify pronunciation errors.

[1537] Furthermore, an emotion recognition engine (e.g., IBM Watson Tone Analyzer) is used to analyze the user's emotions from the voice data, identifying the user's emotional state (e.g., joy, sadness, anger, etc.) and generating feedback accordingly.

[1538] Providing Feedback

[1539] The form of feedback includes at least one of visual feedback, text feedback, and audio feedback. Visual feedback may include animations or videos showing lip movements and tongue positions that indicate correct pronunciation. Text feedback may include specific instructions such as "Put your tongue behind your upper teeth to pronounce 'L'." Audio feedback may include audio samples of correct pronunciation.

[1540] Specific examples

[1541] For example, suppose a factory operator says "start" when operating a robot. This voice input is captured by the device's microphone and sent to a server. The server analyzes the speech of "start" and detects that a specific part (such as the "ta" sound) is not pronounced correctly. It then uses an emotion recognition engine to determine that the operator is slightly nervous. As a result, the system displays an encouraging message along with animations and voice samples showing examples of how to pronounce "ta" correctly.

[1542] Prompt Sentence Examples

[1543] Please tell us about the voice-guided operation support system in factories.

[1544] This system helps factory operators give accurate voice instructions to robots and provides feedback according to the operator's emotional state. It includes steps for voice recognition, pronunciation analysis, emotion recognition, and feedback generation.

[1545] This makes it easier for users to learn correct pronunciation, maintains their motivation to learn, and improves factory operation efficiency.

[1546] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1547] Step 1:

[1548] (audio input and capture)

[1549] The user speaks into a microphone. For example, a factory operator says "start." This voice is captured by the device's microphone and saved as digital voice data. The input is real-time voice data, and the output is saved digital voice data. Specifically, a microphone and voice capture software installed on a smartphone or dedicated device is used.

[1550] Step 2:

[1551] (Sending audio data)

[1552] The device converts the captured digital audio data into an appropriate format (e.g., WAV, MP3) and transmits it to a server via the Internet. The input is the stored digital audio data, and the output is the audio data transmitted to the server. Specifically, a communication module and a data transfer protocol are used.

[1553] Step 3:

[1554] (Audio pre-processing)

[1555] The server performs preprocessing on the received audio data. Specifically, it performs noise removal, normalization, volume adjustment, etc. The input is the audio data sent to the server, and the output is the audio data after preprocessing. As a specific example, an audio processing library (e.g., WebRTC) is used.

[1556] Step 4:

[1557] (Voice Recognition)

[1558] The server performs speech recognition on the preprocessed audio data and converts the audio into text. The input is the preprocessed audio data, and the output is text generated from the audio. Specifically, the server uses the Google Speech-to-Text API or similar.

[1559] Step 5:

[1560] (Extraction of voice characteristics)

[1561] The server extracts audio features (e.g., pitch, intensity, tempo) from audio data. The input is preprocessed audio data, and the output is the extracted audio features. Specifically, it uses an audio analysis library (e.g., Librosa).

[1562] Step 6:

[1563] (Identifying mispronunciations)

[1564] The server compares the extracted speech characteristics with the reference speech characteristics to identify pronunciation errors. The input is the speech characteristics and the reference speech characteristics, and the output is the identified pronunciation errors. Specifically, the comparison process is performed using an algorithm.

[1565] Step 7:

[1566] (Emotion recognition)

[1567] The server uses an emotion engine to recognize emotions from voice data. The input is the voice data, and the output is the recognized emotional state. Specifically, it uses IBM Watson Tone Analyzer.

[1568] Step 8:

[1569] (Generating feedback)

[1570] The server generates feedback based on the identified mispronunciations and the recognized emotional state. The feedback may be in the form of visual, text, or audio. The input is the mispronunciation and the emotional state, and the output is the generated feedback. Specifically, a feedback generation module is used.

[1571] Step 9:

[1572] (Sending and viewing feedback)

[1573] The server sends the generated feedback to the device. The device displays and plays the received feedback to the user. The input is the generated feedback, and the output is the feedback provided to the user. Specifically, a display or speaker is used.

[1574] 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.

[1575] 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.

[1576] 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.

[1577] 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.

[1578] FIG. 9 is a diagram illustrating an emotion map 400 on which multiple emotions are mapped. In the emotion map 400, emotions are arranged in concentric circles radiating from the center. Emotions closer to the center of the concentric circles are more primitive. Emotions representing states and actions arising from a state of mind are arranged on the outer edges of the concentric circles. The concept of emotion includes both affect and mental states. Emotions generally generated from reactions occurring in the brain are arranged on the left side of the concentric circles. Emotions generally induced by situational judgment are arranged on the right side of the concentric circles. Emotions generally generated from reactions occurring in the brain and induced by situational judgment are arranged on the upper and lower sides of the concentric circles. Furthermore, the emotion of "pleasure" is arranged on the upper side of the concentric circles, and the emotion of "discomfort" is arranged on the lower side. In this way, in the emotion map 400, multiple emotions are mapped based on the structure by which emotions are generated, and emotions that tend to occur simultaneously are mapped close to each other.

[1579] 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.

[1580] 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).

[1581] 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.

[1582] 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."

[1583] 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.

[1584] 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).

[1585] 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.

[1586] 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.

[1587] 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.

[1588] 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.

[1589] 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.

[1590] 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.

[1591] 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.

[1592] 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.

[1593] 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.

[1594] 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.

[1595] The following is further disclosed regarding the above embodiment.

[1596] (Claim 1)

[1597] means for capturing input voice and generating voice data;

[1598] means for transmitting the voice data to a server;

[1599] means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics;

[1600] means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors;

[1601] means for generating feedback for improving pronunciation based on the pronunciation errors;

[1602] means for transmitting said feedback to a terminal and displaying said feedback to a user;

[1603] A system including:

[1604] (Claim 2)

[1605] 10. The system of claim 1, wherein the feedback comprises at least one of visual feedback, text feedback, and audio feedback.

[1606] (Claim 3)

[1607] The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

[1608] "Example 1"

[1609] (Claim 1)

[1610] means for capturing input voice and generating voice data;

[1611] means for transmitting the voice data to a server;

[1612] means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics;

[1613] means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors;

[1614] means for generating visual, text and audio feedback for pronunciation improvement based on the pronunciation errors;

[1615] means for transmitting said feedback to a terminal and displaying said feedback to a user;

[1616] A system including:

[1617] (Claim 2)

[1618] 10. The system of claim 1, wherein the feedback includes a diagram or animation showing lip and tongue shape and mouth opening.

[1619] (Claim 3)

[1620] The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

[1621] "Application Example 1"

[1622] (Claim 1)

[1623] means for capturing input voice and generating voice data;

[1624] means for transmitting the voice data to a server;

[1625] means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics;

[1626] means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors;

[1627] means for generating feedback for improving pronunciation based on the pronunciation errors;

[1628] means for transmitting said feedback to a terminal and displaying said feedback to a user;

[1629] A means for analyzing the speech spoken by store staff when interacting with customers and providing feedback;

[1630] A system including:

[1631] (Claim 2)

[1632] 10. The system of claim 1, wherein the feedback comprises at least one of visual feedback, text feedback, and audio feedback.

[1633] (Claim 3)

[1634] The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

[1635] "Example 2: Combining Emotion Engines"

[1636] (Claim 1)

[1637] means for capturing input voice and generating voice data;

[1638] means for transmitting the voice data to a server;

[1639] means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics;

[1640] means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors;

[1641] means for generating feedback for improving pronunciation based on the pronunciation errors;

[1642] means for transmitting said feedback to a terminal and displaying said feedback to a user;

[1643] means for recognizing emotions from the user's voice and generating appropriate feedback based on the recognized emotions;

[1644] A system including:

[1645] (Claim 2)

[1646] 10. The system of claim 1, wherein the feedback comprises at least one of visual feedback, text feedback, and audio feedback.

[1647] (Claim 3)

[1648] The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

[1649] "Application example 2 when combining emotion engines"

[1650] (Claim 1)

[1651] means for capturing input voice and generating voice data;

[1652] means for transmitting the voice data to a server;

[1653] means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics;

[1654] means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors;

[1655] means for generating feedback for improving pronunciation based on the pronunciation errors;

[1656] means for transmitting the feedback to a terminal and displaying the feedback to a user;

[1657] means including an emotion engine for recognizing emotions from audio data;

[1658] a means for generating feedback according to the emotion;

[1659] A system including:

[1660] (Claim 2)

[1661] 10. The system of claim 1, wherein the feedback comprises at least one of visual feedback, text feedback, and audio feedback.

[1662] (Claim 3)

[1663] The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

[1664] (Claim 4)

[1665] 10. The system of claim 1, wherein the system generates feedback including qualitative and emotional information according to the user's state. [Explanation of symbols]

[1666] 10, 210, 310, 410 Data Processing Systems 12 Data Processing Device 14 Smart Devices 214 Smart Glasses 314 Headset-type terminal 414 Robot< / url:> < / url:> < / url:> < / url:>

Claims

1. means for capturing input voice and generating voice data; means for transmitting the voice data to a server; means in the server for analyzing the voice data, performing voice recognition, and extracting voice characteristics; means for comparing the characteristics of the speech with reference speech characteristics to identify pronunciation errors; means for generating feedback for improving pronunciation based on the pronunciation errors; means for transmitting said feedback to a terminal and displaying said feedback to a user; A system including:

2. The system of claim 1 , wherein the feedback comprises at least one of visual feedback, text feedback, and audio feedback.

3. The system of claim 1, wherein the server includes means for pre-processing noise removal, normalization and volume adjustment.

Images