Hearing-aid operation method, system and terminal for communication data for old people

Abstract

The invention relates to the technical field of hearing-aid communication terminals, and discloses a hearing-aid operation method, system and terminal for communication data for old people. The hearing-aid operation method is applied to a terminal host integrated with a hearing-aid unit, an audio unit, a communication unit, a positioning unit, an AI processing unit and a wireless connection unit, and specifically comprises the following steps: S101, obtaining a user start request, synchronously collecting an environment simulation sound signal through a terminal microphone array, and converting the environment simulation sound signal into a high-resolution digital audio stream through an ADC module; according to the invention, a hardware end deeply integrates hearing aid, communication and an AI processing unit to integrate an environment sound enhancement function and a family interaction function into the same data processing stream, and a self-adaptive multi-filtering technology based on scene perception is adopted, so that noise reduction, frequency band enhancement and signal shaping parameters can be dynamically adjusted according to a real-time acoustic environment; and speech definition and auditory safety in a complex environment are obviously improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hearing-aid communication terminals, in particular to a hearing-aid operation method, system and terminal for communication data of the elderly. BACKGROUND

[0002] The current market for technology products for the elderly is showing a diversified development trend. The existing technologies and products can be roughly classified into several main directions. The first type is basic safety products. This type of product has relatively mature technology, and the core appeal is to solve the concerns of children about the safety of the elderly. However, the function dimension is extremely single, and it cannot meet the deeper needs of the elderly in daily communication, spiritual comfort and active entertainment. The second type is health management products. They focus on monitoring physiological indicators and medication reminders, and have made some achievements in health data quantification. However, they still regard the elderly as a passive object of care, lack of active output, emotional expression and spiritual interaction with family members, and have obvious tool attributes. The third type is social entertainment products, such as simplified tablets or smart speakers with video functions designed specifically for the elderly. They provide video calls, audio and video playback and other functions, trying to solve the problems of communication and entertainment. However, the operation interface is still relatively complex, the functions are numerous and the logic does not conform to the intuition of the elderly, resulting in high learning costs. Many elderly people eventually abandon them, and they cannot truly integrate into their daily life. The fourth type is the initial form of emotional companion products. Although they realize the importance of emotional value, the functions are often isolated or can only achieve simple recording and playback. They cannot form a sustainable and evolving data loop and emotional positive cycle for the life details and speech expression of the elderly and the effective interaction with family members, and their influence and user stickiness are limited. Although the above existing technologies try to solve the needs of elderly users in different aspects, they all have fundamental limitations and technical problems that have not been overcome. First of all, most products have a serious function island phenomenon. The data between functions is not connected, and the logic is independent, forcing the elderly to switch between different devices or applications, which is fragmented and tedious. Second, the interactive design does not truly achieve zero learning cost. Complex menus and multiple settings are extremely unfriendly to the elderly, creating a use barrier. The deeper problem is that the existing technical paradigm is one-way care, i.e., children monitor the status of the elderly or make greetings in one direction, and there is no effective mechanism to stimulate the elderly to actively express themselves and provide children with high-quality care materials worthy of response, making it difficult to build a sustainable emotional exchange.

[0003] Therefore, there is an urgent need in the market for a communication data terminal for the elderly that can break through these barriers, seamlessly integrate practical functions, and create emotional value through data flow, fundamentally improving the usability and vitality of technology products for the elderly. SUMMARY

[0004] The application aims to provide an old person's communication data hearing aid operation method, system and terminal, which integrates environmental sound enhancement and family interaction function in the same data processing flow through deep integration of hearing aid, communication and AI processing unit in hardware end, adopts adaptive multiple filtering technology based on scene perception, can dynamically adjust noise reduction, frequency band enhancement and signal shaping parameters according to real-time acoustic environment, significantly improves speech intelligibility and auditory safety in complex environment, and aims to solve the problems in the prior art.

[0005] The application also aims to expand the function scene of the terminal by integrating health management ecology and smart home control. Specifically, the terminal can be used as a smart device master control end, access to matching wearable devices such as smart rings, and real-time acquisition of the physical information of the old person (such as heart rate, blood oxygen, step count, etc.), realize daily health monitoring and family circle synchronization, and automatically trigger multi-level alarm linkage when detecting emergency events (such as falling or long-term stillness), improve safety properties. At the same time, the terminal supports voice control of smart home devices, embodies its pivotal role in the home Internet of Things, and realizes zero learning cost through senior design (such as one-key operation, large font interface), and comprehensively improves the experience of old users.

[0006] The application is implemented as follows: an old person's communication data hearing aid operation method is applied to a terminal host integrated with a hearing aid unit, an audio unit, a communication unit, a positioning unit, an AI processing unit and a wireless connection unit, and specifically includes the following steps: S101: Obtain a user start request, synchronously collect environmental simulation sound signals through a terminal microphone array, convert them into high-resolution digital audio streams through an ADC module, and at the same time, an AI processing unit starts an acoustic scene analysis model to identify the environment type of the high-resolution digital audio stream in real time, to obtain current environment identification parameters, and synchronously monitor whether the communication unit receives a voice message or a call request from a family member, mark the voice message or the call request as a highest priority audio source to be processed; S102: The obtained environment identification parameters are first subjected to spectral noise reduction and Wiener filtering, then a target frequency band is enhanced through an adaptive band-pass filter, a multi-layer filtering algorithm is dynamically configured, and finally signal shaping and anti-aliasing processing are performed, and finally the processed environmental sound and the received voice message or call request from the family member are mixed to generate a composite audio stream; S103: The generated composite audio stream is transmitted to a paired earphone through a Bluetooth low energy protocol with ≤40ms delay, the communication long connection is maintained during the transmission process, real-time family voice insertion is supported, and in combination with positioning and motion data, if the user is in an outdoor mobile state, the output gain is automatically adjusted to balance intelligibility and safety; S104: The earphone converts the incoming composite signal into a sound wave, and the active transmission sound and the passive transmission sound are superimposed in the ear canal through electronic processing to form an enhanced pseudo-environmental sound field, so that the user can naturally perceive the surrounding environment while listening to the enhanced sound; S105: While the hearing aid function is ongoing, the AI processing unit uses ASR to identify and transcribe the collected environmental simulation sound signals into text in real time, and then uses the NLP language processing model to perform semantic analysis, abstract extraction, and sentiment judgment on the text content. Automatically mark the key feature segments with recall value and perform structured processing, then upload the text abstract and key feature segments to the cloud server, send a notification to the family member end through the communication unit, trigger the family member to comment or voice reply, the family member end triggers the comment or voice reply as a new audio content source, feedback to the terminal host through the communication unit, and incorporate the family member's voice message or call request for playback, forming a continuous emotional interaction data loop between the terminal host and the family member end.

[0007] Further, in S101, the user's start request is obtained, and the environmental simulation sound signals are synchronously collected through the terminal microphone array, including: The user sends a start request through a physical button or a voice command, and the terminal host initializes the built-in multi-microphone array into a ready state immediately after responding, waiting to capture the simulation sound signals in the environment, preparing for subsequent synchronous collection and processing; The microphone array continuously collects full-band simulation sound signals in the user's surrounding environment in a synchronous or quasi-synchronous manner, and the environmental sound signals include human voice conversations, vehicle driving sounds, horn sounds, and natural wind sounds, capturing various elements of the current acoustic scene.

[0008] Further, the AI processing unit simultaneously starts the acoustic scene analysis model to identify the environment type of the high-resolution digital audio stream in real time to obtain the current environment identification parameters, including: After the microphone array completes sound signal collection and converts it into a digital audio stream, the built-in AI processing unit is activated, which calls a pre-trained acoustic scene analysis deep learning model to perform real-time analysis and feature mining on the input high-resolution, full-band digital audio data; The acoustic scene analysis deep learning model performs millisecond-level spectral analysis and pattern matching on continuous audio streams, accurately determines the current environment type by extracting acoustic features, and outputs environment identification parameters with confidence scores through the current environment type.

[0009] Further, in S102, the obtained environment identification parameters are first subjected to spectral noise reduction and Wiener filtering, then subjected to adaptive band-pass filtering to enhance the target frequency band, and finally subjected to signal shaping and anti-aliasing processing, including: According to the environmental recognition parameter, the specific parameters of the multi-layer filtering algorithm are dynamically configured, the noise suppression based on the spectral subtraction method is used to effectively separate and suppress the stable and non-stable noise in the environment, and the Wiener filter is used to further optimize the signal waveform, thereby improving the signal-to-noise ratio of the digital audio stream and obtaining the scene analysis result; According to the scene analysis result, the adaptive band-pass filter is started to enhance the target frequency band signal, and then the signal is smoothed and shaped and anti-aliasing filtering is applied to eliminate high-frequency distortion and spectral aliasing. The processed digital signal has high fidelity and accuracy.

[0010] Further, in S103, the output gain is automatically adjusted to balance the clarity and safety while the user is moving outdoors, including: Real-time data from the positioning unit and motion sensor built-in the terminal host are continuously received and fused, and by analyzing the user's speed, trajectory and motion pattern, it can be intelligently judged whether the user is currently in an outdoor moving state; If it is determined that the user is in an outdoor moving environment, the gain parameter of the audio output is automatically fine-tuned according to the preset safety strategy, the clarity of the key warning sound is moderately improved, and the environmental sound masking caused by the overall volume is avoided, thereby ensuring the hearing clarity while maintaining the user's personal safety.

[0011] Further, in S104, the earphone converts the incoming composite signal into sound waves, and the electronic processed sound and the passively transmitted sound are superimposed in the ear canal to form an enhanced pseudo-environmental sound field, including: The loudspeaker unit inside the earphone accurately converts the received composite electronic signal into corresponding analog sound waves, and the conversion process is pre-acoustically calibrated to ensure the high fidelity of the output sound; At the same time, the earphone allows part of the environmental sound to be passively transmitted directly into the ear canal. The natural sound that has not been electronically processed and the processed electronic sound played by the loudspeaker are superimposed and fused at the eardrum to form a pseudo-environmental sound field that is both enhanced and maintains natural auditory experience.

[0012] Further, in S105, the AI processing unit uses ASR recognition to convert the collected environmental analog sound signal into text in real time, and then uses NLP language processing model to perform semantic analysis, abstract extraction and emotion judgment on the text content, automatically marks the key feature segments with recall value and performs structured processing, including: The AI processing unit first processes the collected original environmental sound signal, especially the recording instruction triggered by the user's active key or the environmental sound continuously captured by the device, uses ASR recognition technology to convert the analog audio stream into continuous text stream at high speed and accuracy, and completes the primary conversion from sound to text; The NLP language processing model is called to perform deep semantic analysis, key abstract extraction and sentiment tendency judgment on the transcribed text.

[0013] Further, the terminal host is also integrated with a health management module and a smart home control module; the health management module is used to access the smart wearable device to obtain the physical information of the old people, realize daily health monitoring and abnormality early warning, and synchronize the health data to the family member end through the communication unit; the smart home control module supports voice instruction to directly control the smart home device.

[0014] Further, the terminal host is also used to execute the health management ecological function: access the smart ring through the wireless connection unit to obtain the physical information such as heart rate, blood oxygen and step count of the old people; the AI processing unit performs long-term monitoring on the physical information, generates a health report and early warning information when detecting an abnormality, and uploads the information to the cloud and pushes the information to the family member APP end through the communication unit; at the same time, based on the positioning unit and the motion sensor data, the AI processing unit automatically triggers an emergency call mode when detecting that the user falls down or is stationary for a long time, including: sending an alarm with a position to the family members and initiating a family circle multi-person call; continuously playing a warning sound to remind people around; if no one responds, automatically connect to the preset emergency center.

[0015] Further, the terminal host is also used to execute the smart home control function: analyze the user voice instruction through the voice recognition module to directly control the accessed smart home IoT device, realize the operation such as turning on the living room light; the terminal host serves as the smart home master control and supports multi-platform smart home ecological access.

[0016] Further, the communication unit supports one-to-many multi-person call function, realizes family circle multi-person call at the same time in an emergency or daily communication, and improves the call timeliness.

[0017] Further, the terminal host adopts an old-age design, including a one-key recording physical button, a large font interface and voice guidance, and realizes zero learning cost operation.

[0018] Compared with the prior art, the hearing aid operation method, system and terminal for old people provided by the application have the following beneficial effects: 1. By deeply integrating hearing aid, communication and AI processing units in hardware, the system integrates environmental sound enhancement and home interaction functions in the same data processing flow. It uses adaptive multi-filtering technology based on scene perception to dynamically adjust noise reduction, frequency band enhancement and signal shaping parameters according to real-time acoustic environment, significantly improving speech intelligibility and auditory safety in complex environments. It also combines ultra-low delay wireless transmission and passive sound transmission structure to create a pseudo environmental sound field at the earphone end that combines electronic enhancement and natural listening experience. This allows elderly users to seamlessly perceive environmental risks and family care without removing the device, fundamentally solving the problem of the separation of traditional hearing aids and communication functions. 2. The system creates an emotional interaction data loop triggered by the old man's original sound. It automatically converts valuable segments in daily speech into structured digital memories through ASR and NLP, and actively pushes them to family members, triggering interactive behaviors such as creation, comments or voice replies. The feedback from family members is then integrated into the hearing audio stream as a high-priority audio source, forming a continuous positive emotional cycle. This not only upgrades the product from a functional tool to a digital extension of oneself, but also innovatively enables cross-generation dialogue, greatly improving the happiness of elderly users and the engagement of family members, creating significant emotional and digital asset value.

[0019] 3. By integrating a health management ecosystem, the system realizes real-time monitoring of the elderly's vital signs and synchronization with the family circle. In the event of an emergency, it automatically triggers multi-level alarm linkage, effectively improving the product's safety and emergency response capabilities. At the same time, the terminal serves as the main control of the smart home, expanding the use scenarios and meeting the needs of elderly users for convenient life. The senior-friendly design ensures zero learning cost, enhancing the product's ease of use and user stickiness.

[0020] The hearing aid system for running communication data for the elderly is used to execute the hearing aid running method described above. The hearing aid running system includes: A sound collection module is used to obtain a user start request and synchronously collect environmental simulation sound signals including human voice conversation, vehicle driving sound, siren sound and natural wind sound through a terminal microphone array. The ADC module converts the signals into high-resolution digital audio streams. A scene analysis module is used to start the acoustic scene analysis model in the AI processing unit to identify the environment type of the digital audio stream in real time, obtain environment identification parameters, and monitor the family member voice messages or call requests received by the communication unit, marking them as the highest priority audio source. An audio processing module is used to perform spectral noise reduction and Wiener filtering, adaptive band-pass filter enhancement of target frequency bands, signal shaping and anti-aliasing processing based on environment identification parameters, and mix the processed environmental sound and family member voice to generate a composite audio stream. A wireless transmission module is configured to transmit the composite audio stream to a paired earphone through a Bluetooth Low Energy protocol with a delay of ≤40 ms, maintain a long connection for communication to support real-time voice insertion, and automatically adjust the output gain in combination with positioning and motion data when moving outdoors; A sound field synthesis module is configured to convert the composite signal received by the earphone into a sound wave, which is superimposed with the passively transmitted environmental sound in the ear canal to form an enhanced pseudo-environmental sound field. An interactive processing module is configured to convert, analyze, and mark key feature segments with recall value in real time through ASR and NLP technologies, upload structured content to the cloud, and trigger interaction on the family member side to form a closed loop of emotional interaction data.

[0021] Specifically, the audio processing module includes: A noise reduction unit is configured to separate stable and unstable noise in the environment by applying a noise suppression based on spectral subtraction and optimize the signal waveform by using Wiener filtering. A filter enhancement unit is configured to start an adaptive band-pass filter to enhance the target frequency band signal according to the scene analysis result. A signal shaping unit is configured to perform smoothing and shaping on the signal and apply anti-aliasing filter processing to eliminate high-frequency distortion and spectral aliasing.

[0022] The hearing aid operation system further includes: A health management module is configured to access the smart wearable device to obtain the physical information of the elderly, perform health monitoring and early warning, and synchronize the data to the family member side through the communication unit. A smart home control module is configured to analyze the user voice command and control the smart home device. An emergency event processing module is configured to detect a fall or long-time stillness event based on the positioning and motion data, and automatically trigger a multi-level alarm linkage.

[0023] The hearing aid operation terminal for the elderly uses communication data, which includes a storage device and a processor. The storage device is used to store a computer program, and the processor runs the computer program to make the hearing aid operation terminal execute the hearing aid operation method described above. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The flowchart of the hearing aid operation method for the elderly using communication data proposed by the present application is shown in the figure. Figure 2 The flowchart of the AI processing unit in the hearing aid operation method for the elderly using communication data proposed by the present application is shown in the figure. The AI processing unit uses ASR to recognize the collected environmental simulation sound signals and convert them into text in real time, and then uses NLP language processing model to perform semantic analysis, abstract extraction, and emotion judgment on the text content. Figure 3The structural schematic diagram of the hearing aid operation system for the old people to communicate data according to the present application is shown in the figure. Figure 4 The structural schematic diagram of the hearing aid operation terminal for the old people to communicate data according to the present application is shown in the figure. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of the present application more clear, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0026] The implementation of the present application is described in detail below in combination with specific examples.

[0027] The same or similar reference numerals in the drawings of the present embodiment correspond to the same or similar components; in the description of the present application, it should be understood that the orientations or positional relationships indicated by the terms "upper", "lower", "left", "right" and the like are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and therefore the terms describing the positional relationships in the drawings are only used for exemplary illustration and cannot be understood as limiting the present application; for those skilled in the art, the specific meanings of the above terms can be understood according to the specific circumstances.

[0028] Referring to Figures 1-2 As shown in the figure, the hearing aid operation method for the old people to communicate data is applied to a terminal host integrated with a hearing aid unit, an audio unit, a communication unit, a positioning unit, an AI processing unit and a wireless connection unit, and specifically includes the following steps: S101: obtaining a user start request, synchronously collecting environment simulation sound signals through a terminal microphone array, converting the environment simulation sound signals into high-resolution digital audio streams through an ADC module, starting an acoustic scene analysis model of an AI processing unit to identify the environment type of the high-resolution digital audio streams in real time to obtain current environment identification parameters, and synchronously monitoring whether the communication unit receives a voice message or a call request of a family member, marking the voice message or the call request as a highest priority audio source to be processed; Among them, obtaining a user start request, synchronously collecting environment simulation sound signals through a terminal microphone array, includes: The user sends a start request through a physical button or a voice instruction, and the terminal host initializes the built-in multi-microphone array into a ready state immediately after responding, and waits to capture the simulation sound wave signals in the environment, so as to prepare for subsequent synchronous collection and processing; The microphone array continuously collects full-band analog sound signals in the user's surrounding environment in a synchronous or quasi-synchronous manner, and the environmental sound signals include human voice conversations, vehicle driving sounds, horn sounds, and natural wind sounds, fully capturing various elements of the current acoustic scene. S102: The acquired environmental identification parameters are first subjected to spectral noise reduction and Wiener filtering, then a dynamic multi-layer filtering algorithm is configured for the target frequency band through an adaptive band-pass filter, and finally signal shaping and anti-aliasing processing are performed, and the processed environmental sound is mixed with the voice message or call request received from the family members to generate a composite audio stream. The acquired environmental identification parameters are first subjected to spectral noise reduction and Wiener filtering, then a dynamic multi-layer filtering algorithm is configured for the target frequency band through an adaptive band-pass filter, and finally signal shaping and anti-aliasing processing are performed, including: The specific parameters of the multi-layer filtering algorithm are dynamically configured according to the environmental identification parameters, the noise suppression based on spectral subtraction effectively separates and suppresses stable and non-stable noise in the environment, and the Wiener filter further optimizes the signal waveform, improving the signal-to-noise ratio of the digital audio stream and obtaining scene analysis results; According to the scene analysis results, the adaptive band-pass filter enhances the target frequency band signal, and then performs smoothing shaping and anti-aliasing filtering processing on the signal to eliminate high-frequency distortion and spectral aliasing. The processed digital signal has high fidelity and accuracy.

[0029] S103: The generated composite audio stream is transmitted to the paired earphone through the Bluetooth protocol with ≤40ms delay, and the communication long connection is maintained during transmission to support real-time insertion of family voice, and combined with positioning and motion data, if the user is in a mobile outdoor state, the output gain is automatically adjusted to balance clarity and safety; The positioning and motion data are combined, and if the user is in a mobile outdoor state, the output gain is automatically adjusted to balance clarity and safety, including: Real-time data from the positioning unit and motion sensor built-in the terminal host are continuously received and fused, and by analyzing the user's speed, trajectory, and motion pattern, it can intelligently determine whether the user is currently in a mobile outdoor state; If it is determined that the user is in a mobile outdoor environment, the gain parameter of the audio output is automatically fine-tuned according to the preset safety strategy, the clarity of the key warning sound is moderately improved, and the environmental sound is shielded due to the overall volume being too large, ensuring auditory clarity while maintaining the user's personal safety; S104: The earphone converts the incoming composite signal into sound waves, and the electronic processed sound and passive transmitted sound are superimposed in the ear canal to form an enhanced pseudo-environmental sound field, allowing the user to naturally perceive the surrounding environment while listening to the enhanced sound. Among them, the earphone converts the incoming composite signal into a sound wave, and the passive sound transmission sound is superimposed in the ear canal to form an enhanced pseudo-environmental sound field, including: The speaker unit inside the earphone accurately converts the received composite electronic signal into corresponding analog sound waves, and the conversion process is pre-acoustically calibrated to ensure high fidelity of the output sound; At the same time, the earphone allows part of the environmental sound to be directly and passively transmitted into the ear canal. This part of the natural sound without electronic processing and the processed electronic sound played by the speaker are superimposed and fused at the eardrum to form a pseudo-environmental sound field that enhances and maintains natural auditory experience; S105: While the hearing aid function is ongoing, the AI processing unit uses ASR to identify and transcribe the collected environmental simulation sound signals into text in real time, and then uses NLP language processing models to analyze the semantic content, extract summaries, and judge emotions. Automatically mark the key feature segments with recall value and structure them, then upload the text summary and key feature segments to the cloud server, send a notification to the family member end through the communication unit, trigger family members to comment or voice reply, and the family member end triggers the comment or voice reply as a new audio content source. Feedback to the terminal host through the communication unit, and integrate the family member's voice message or call request into the data loop of continuous emotional interaction between the terminal host and the family member end, and integrate the environmental sound enhancement and family interaction functions in the same data processing stream through the hardware end. Deeply integrated hearing aid, communication and AI processing unit, using adaptive multi-filtering technology based on scene perception, can dynamically adjust noise reduction, frequency band enhancement and signal shaping parameters according to real-time acoustic environment, significantly improve speech intelligibility and hearing safety in complex environments, and at the same time, through the combination of ultra-low delay ≤40ms wireless transmission and passive sound transmission structure, a pseudo-environmental sound field with electronic enhancement effect and natural auditory experience is constructed at the earphone end, allowing the elderly user to seamlessly perceive environmental risks and family care without removing the device, fundamentally solving the pain point of the traditional hearing aid device and communication function fragmentation.

[0030] In this embodiment, the terminal host also integrates a health management module and a smart home control module. The health management module accesses the smart ring through a wireless connection unit (such as Bluetooth) to obtain real-time vital information of the old person, such as heart rate, blood oxygen, and step count. The AI processing unit monitors and analyzes these information for a long time. When an abnormal value (such as high heart rate or low blood oxygen) is detected, a health report and a warning information are automatically generated and uploaded to the cloud server through the communication unit, and are pushed to the family member APP end, realizing health monitoring and intervention of the old person. In addition, based on the positioning unit and the motion sensor data, the AI processing unit can detect user falls (through the motion sensor) or long-time stillness (through the positioning and motion data), and automatically trigger an emergency call mode: priority to send a location-based alarm to family members and initiate a family circle multi-person call; continuously play a warning sound to remind people around; if no one responds, automatically connect to a preset emergency center. This multi-level alarm linkage mechanism enhances the safety properties of the product and solves the sudden risks of the elderly when they are alone.

[0031] The smart home control module supports voice recognition, and the old person can directly control the smart home IoT device (such as “turn on the living room light”) through voice. The terminal host as the smart home master controls accesses multiple platform smart home ecosystems, expands the use scenarios, and embodies the hub role of the product in the “home Internet of Things”.

[0032] The communication unit supports one-to-many multi-person call function, which realizes multi-person call at the same time in the family circle in emergency or daily communication. Compared with the one-to-one broadcast call form on the market, it is more time-efficient in process.

[0033] The terminal host adopts senior-friendly design, including one-key recording physical button, large font interface, and voice guidance, to ensure zero learning cost operation and reduce the use threshold of the elderly users.

[0034] In S101 of this embodiment, the AI processing unit simultaneously starts the acoustic scene analysis model to identify the environment type of the high-resolution digital audio stream in real time to obtain the current environment recognition parameter, including: After the microphone array completes the sound signal collection and converts it into a digital audio stream, the built-in AI processing unit is activated. The AI processing unit calls the pre-trained acoustic scene analysis deep learning model to perform real-time analysis and feature mining on the input high-resolution, full-band digital audio data. The acoustic scene analysis deep learning model performs millisecond-level spectral analysis and pattern matching on the continuous audio stream, accurately determines the current environment type by extracting acoustic features, and outputs the environment recognition parameter with confidence score through the current environment type.

[0035] In S105 of the present embodiment, the AI processing unit converts the collected environmental simulation sound signals into text in real time using ASR recognition, and then uses an NLP language processing model to perform semantic analysis, abstract extraction, and sentiment judgment on the text content, automatically marks key feature segments with recall value, and performs structured processing, including: The AI processing unit first processes the collected original environmental sound signals, especially the recording instructions triggered by the user's active key or the continuously captured environmental sound, using ASR recognition technology to convert the analog audio stream into continuous text stream at high speed and accurately, completing the primary conversion from sound to text; The NLP language processing model is called to perform deep semantic analysis, key abstract extraction, and sentiment judgment on the converted text. Through analysis of the text content, the NLP language processing model can automatically identify and mark key feature segments with emotions or memorial value, and perform structured processing for subsequent use.

[0036] The technical solution constructs an emotional interaction data closed loop triggered by the old man's original sound. Through ASR and NLP, valuable segments in daily speech are automatically converted into structured digital memories and actively pushed to family members, triggering interactive behaviors such as creation, comments, or voice replies. The feedback content of family members is then real-time integrated into the hearing aid audio stream as a high-priority audio source, forming a continuous positive emotional cycle. This not only upgrades the product from a functional tool to a digital self-extension, but also innovatively realizes cross-generation two-way dialogue, greatly improving the happiness of the elderly user and the participation of family members, creating significant emotional value and digital asset value.

[0037] Referring to Figure 3As shown, the old communication data hearing aid operation system is used to execute the above hearing aid operation method, and the hearing aid operation system comprises: a sound acquisition module, which is used to acquire a user starting request, synchronously acquires environment simulation sound signals including human voice conversation, vehicle driving sound, whistle sound and natural wind sound through a terminal microphone array, and converts the environment simulation sound signals into high-resolution digital audio streams through an ADC module; a scene analysis module, which is used to start an acoustic scene analysis model in an AI processing unit to identify the environment type of the digital audio stream in real time, acquire environment identification parameters, and monitor a family member voice message or a call request received by a communication unit, and mark the family member voice message or the call request as a highest priority audio source; an audio processing module, which is used to sequentially perform spectral noise reduction and Wiener filtering, enhance a target frequency band through an adaptive band-pass filter based on the environment identification parameters, perform signal shaping and anti-aliasing processing, and mix the processed environment sound and the family member voice to generate a composite audio stream; a wireless transmission module, which is used to transmit the composite audio stream to a paired earphone through a Bluetooth low energy protocol with a delay of ≤40 ms, maintain a communication long connection to support real-time voice insertion, and automatically adjust output gain when moving outdoors in combination with positioning and motion data; a sound field synthesis module, which is used to convert the composite signal received by the earphone into a sound wave, superimpose the sound wave with passively transmitted environment sound in an ear canal to form an enhanced pseudo-environment sound field; and an interactive processing module, which is used to convert, analyze and mark key feature segments with recall value in real time through ASR and NLP technologies on the collected audio signal, upload structured content to the cloud and trigger family member end interaction to form an emotional interaction data closed loop, and combine ultra-low delay ≤40 ms wireless transmission with a passive sound transmission structure to build a pseudo-environment sound field with both electronic enhancement effect and natural hearing experience on the earphone end, so that the old user can seamlessly perceive environmental risks and family care without removing the device, and fundamentally solve the pain point of the split between traditional hearing aids and communication functions.

[0038] In the embodiment, the audio processing module comprises: a noise reduction unit, which is used to separate stable and unstable noises in the environment by applying noise suppression based on spectral subtraction, and optimize the signal waveform through Wiener filtering; a filter enhancement unit, which is used to enhance the target frequency band signal through an adaptive band-pass filter according to the scene analysis result; and a signal shaping unit, which is used to perform smoothing shaping and anti-aliasing filtering processing on the signal to eliminate high-frequency distortion and spectral aliasing. Through deep fusion of hearing aid, communication and AI processing units on the hardware end, the environment sound enhancement and family interaction functions are integrated in the same data processing flow, and adaptive multiple filtering technology based on scene perception is adopted to dynamically adjust the noise reduction, frequency band enhancement and signal shaping parameters according to the real-time acoustic environment, thereby significantly improving the speech intelligibility and auditory safety in complex environments.

[0039] The hearing aid operation system further comprises a health management module, a smart home control module and an emergency event processing module. The health management module accesses the smart wearable device through the wireless connection unit, obtains the physical information of the old people, performs health monitoring and early warning, and synchronizes the data to the family member end through the communication unit. The smart home control module analyzes the user voice instruction and controls the smart home device to realize home automation. The emergency event processing module detects the fall or long-time stillness event based on the positioning and motion data, automatically triggers multi-level alarm linkage, including family circle call, warning sound playing and emergency center connection; for solving the existing health monitoring devices on the market, which mainly focus on physiological index collection, but lack of deep integration with the family social circle, resulting in that the health data cannot be synchronized to the family members in real time, and the emergency event response mechanism is single, usually only supports one-to-one broadcast call, the response timeliness is poor, and the multi-level alarm linkage cannot be triggered quickly when the old people fall or are still for a long time, and the smart home control function is often independent of the old people's communication terminal, forcing the old users to switch between different devices, increasing the operation complexity, which violates the principle of old design.

[0040] Referring to Figure 4 As shown in the hearing aid operation terminal of the old people, the hearing aid operation terminal comprises a storage device and a processor, the storage device is used for storing a computer program, and the processor runs the computer program to enable the hearing aid operation terminal to execute the hearing aid operation method.

[0041] The technical scheme automatically converts the valuable fragments in daily voice into structured digital memories through ASR and NLP, forms a continuous emotional positive cycle, which not only upgrades the product from a functional tool to a digital self-extension, but also innovatively realizes the cross-generation bidirectional dialogue, and has the following functions: In terms of hardware functions, the recording function supports one-key operation and can upload to the cloud while recording, but does not have the function of canceling recording and uploading, and each recording can be up to 60 minutes long; the hearing aid function provides four hearing aid modes: custom, strong noise reduction, indoor, and outdoor, which can be selected as needed; the FM function can realize local station searching, meeting the needs of listening to the radio; the call function can be one-to-many, similar to group chat and conference functions, facilitating communication and interaction among multiple people; the positioning function is only turned on when calling, and the location is not displayed at other times; the AI conversation function can chat with AI; the message push function process is that the user first records a voice message on the APP and uploads it to the cloud, then the cloud pushes the latest unread message to the time machine hardware display, if there is only one message pushed, the user reads it and the information is not stored, if there are multiple messages, the time queue is confirmed to be read and the next unread message is pushed, if the user does not read the current displayed message, the subsequent new message will not be displayed and the current message will always be displayed; the device binding code function can be used to bind devices to establish a family circle; the system function includes low battery reminder, charging progress prompt, volume increase and decrease, and power on and off prompt, etc., to facilitate the user to master the device state; the vibration prompt function will give corresponding prompt when the key is operated; there is also network support through E-sim card; In terms of software functions, the AI debugging function can realize AI creation and support AI conversation; the recording transcription debugging function is very powerful and can perform domestic and foreign voice transcription and dialect transcription; the homepage function includes device binding, family circle creation / joining, information viewing, family circle member invitation, and call and positioning operations; the homepage-device can perform hearing aid custom mode debugging, device information setting, and related operations such as unbinding; the recording function not only involves recording content permission management, but also can transcribe, summarize, generate to-do items, create memoirs, and even perform voice message on recording content; the note function supports generating to-do items through recording, creating to-do items by the user, and generating family to-do items using AI; the "My" panel includes account management, subscription package, recycle bin, and other related settings, facilitating the user to manage and operate their own usage, etc.

[0042] In summary, the local machine provides a multi-dimensional, convenient and practical user experience through the cooperation of hardware and software functions.

[0043] In this embodiment, the entire operation process can be controlled by a computer to provide signal feedback, and the steps are performed in sequence. These are all conventional knowledge of automatic control, which will not be described in detail in this embodiment.

[0044] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement and improvement made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A hearing aid operating method for communicating data with an elderly person, characterized in that, The application is applied to a terminal host integrated with a hearing aid unit, an audio unit, a communication unit, a positioning unit, an AI processing unit and a wireless connection unit, and specifically comprises the following steps: S101: obtaining a user start request, synchronously collecting environment simulation sound signals through a terminal microphone array, converting the environment simulation sound signals into high-resolution digital audio streams through an ADC module, starting an acoustic scene analysis model of an AI processing unit to identify the type of the high-resolution digital audio streams in real time to obtain current environment identification parameters, and synchronously monitoring whether a voice message or a call request of a family member is received by the communication unit, marking the voice message or the call request as a highest priority audio source to be processed; S102: performing spectral noise reduction and Wiener filtering on the obtained environment identification parameters, enhancing a target frequency band through a self-adaptive band-pass filter, configuring a multi-layer filtering algorithm, performing signal shaping and anti-aliasing processing, mixing the processed environment sound and the received voice message or call request of the family member to generate a composite audio stream; S103: transmitting the generated composite audio stream to a paired earphone through a Bluetooth low energy protocol with a delay of ≤40ms, maintaining a communication long connection during the transmission process, supporting real-time insertion of family voice, and automatically adjusting the output gain to balance the intelligibility and safety if the user is in an outdoor mobile state; S104: converting the transmitted composite signal into a sound wave by the earphone, superimposing the electronic processed sound and the passive transmitted sound in the ear canal to form an enhanced pseudo environment sound field, so that the user can naturally perceive the surrounding environment while listening to the enhanced sound; S105: while the hearing aid function is continuously performed, the AI processing unit uses ASR recognition to convert the collected environment simulation sound signals into text in real time, uses an NLP language processing model to perform semantic analysis, abstract extraction and emotion judgment on the text content, automatically marks key feature segments with recall value and performs structured processing, uploads the text abstract and key feature segments to a cloud server, sends a notification to a family member end through the communication unit, triggers the family member to comment or voice reply, and the comment or voice reply of the family member end is fed back to the terminal host as a new audio content source and merged into the voice message or call request of the family member to be played, forming a continuous emotional interaction data closed loop of the terminal host and the family member end.

2. The hearing aid operation method for communicating data for the elderly according to claim 1, wherein In S101, the user start request is obtained, and environment simulation sound signals are synchronously collected through a terminal microphone array, including: The user sends a start request through a physical button or a voice instruction, and the terminal host initializes the built-in multi-microphone array into a ready state immediately after responding, and waits to capture analog sound wave signals in the environment, preparing for subsequent synchronous collection and processing; The microphone array continuously collects full-band analog sound signals in the user's surrounding environment in a synchronous or quasi-synchronous manner, and the environment sound signals include human voice conversation, vehicle driving sound, siren sound and natural wind sound, which comprehensively capture various elements of the current acoustic scene.

3. The hearing aid operation method for communicating data for the elderly according to claim 2, wherein In the meantime, the AI processing unit starts the acoustic scene analysis model to identify the environment type of the high-resolution digital audio stream in real time to obtain the current environment recognition parameters, including: After the microphone array completes the sound signal collection and converts it into a digital audio stream, the built-in AI processing unit is activated, which calls the pre-trained acoustic scene analysis deep learning model to perform real-time analysis and feature mining on the input high-resolution, full-band digital audio data; The acoustic scene analysis deep learning model performs millisecond-level spectral analysis and pattern matching on continuous audio streams, accurately determines the current environment type by extracting acoustic features, and outputs environment recognition parameters with confidence scores.

4. The hearing aid operation method for communicating data for the elderly according to claim 3, wherein In S102, the obtained environment recognition parameters are first subjected to spectral noise reduction and Wiener filtering, then subjected to adaptive band-pass filter enhancement of target frequency bands, and finally subjected to signal shaping and anti-aliasing processing, including: According to the specific parameters of the dynamically configured multi-layer filtering algorithm based on the environment recognition parameters, the noise suppression based on spectral subtraction is used to effectively separate and suppress stable and unstable noise in the environment, and the Wiener filter is used to further optimize the signal waveform, thereby improving the signal-to-noise ratio of the digital audio stream and obtaining the scene analysis result; According to the scene analysis result, the adaptive band-pass filter is started to enhance the target frequency band signal, and then the signal is smoothed and subjected to anti-aliasing filtering to eliminate high-frequency distortion and spectral aliasing. The processed digital signal has high fidelity and accuracy.

5. The hearing aid operation method for communicating data for the elderly according to claim 4, wherein In S103, the output gain is automatically adjusted to balance the clarity and safety if the user is in a mobile outdoor state, including: Real-time data from the built-in positioning unit and motion sensor of the terminal host are continuously received and fused, and by analyzing the user's speed, trajectory, and motion pattern, it can be intelligently determined whether the user is currently in a mobile outdoor state; If it is determined that the user is in a mobile outdoor environment, the gain parameter of the audio output is automatically fine-tuned according to the preset safety strategy, the clarity of the key warning sound is moderately improved, and the overall volume is prevented from being too large to cause environmental sound masking, thereby ensuring auditory clarity while maintaining the user's personal safety.

6. The hearing aid operation method for communicating data for the elderly according to claim 5, wherein In S104, the earphone converts the incoming composite signal into sound waves, and the electronic processed sound and passively transmitted sound are superimposed in the ear canal to form an enhanced pseudo-environmental sound field, including: The loudspeaker unit inside the earphone accurately converts the received composite electronic signal into corresponding analog sound waves, and the conversion process is pre-acoustically calibrated to ensure high fidelity of the output sound; At the same time, the earphone allows part of the environmental sound to be passively transmitted directly into the ear canal. This part of the environmental sound is not electronically processed and is superimposed with the processed electronic sound played by the loudspeaker in the eardrum to form a pseudo-environmental sound field that is enhanced and maintains natural auditory experience.

7. The hearing aid operation method for communicating data for the elderly according to claim 6, wherein In S105, the AI processing unit uses ASR recognition to convert the collected environmental analog sound signals into text in real time, and then uses NLP language processing models to perform semantic analysis, abstract extraction, and sentiment judgment on the text content, automatically marking key feature segments with recall value and performing structured processing, including: The AI processing unit first processes the collected original environmental sound signals, especially the recording instructions triggered by the user's active button or the continuously captured environmental sound, using ASR recognition technology to convert the analog audio stream into continuous text stream at high speed and accuracy, completing the primary conversion from sound to text. The NLP language processing model is called to perform deep semantic analysis, key abstract extraction, and sentiment judgment on the transcribed text. Through the analysis of the text content, the NLP language processing model can automatically identify and mark the key feature segments that contain emotions or have memorial value, and structure them for subsequent use.

8. A hearing aid operating system using communication data for the elderly, characterized by, The hearing aid operation system for executing the hearing aid operation method of any one of claims 1-7 comprises: A sound collection module is used to obtain a user's start request, synchronously collect environmental simulation sound signals including human voice conversation, vehicle driving sound, horn sound, and natural wind sound through a terminal microphone array, and convert them into high-resolution digital audio streams through an ADC module; A scene analysis module is used to start the acoustic scene analysis model in the AI processing unit to identify the environmental type of the digital audio stream in real time, obtain environmental identification parameters, and monitor the voice messages or call requests of family members received by the communication unit, marking them as the highest priority audio source; An audio processing module is used to perform spectral noise reduction and Wiener filtering, adaptive band-pass filter enhancement of target frequency bands, signal shaping and anti-aliasing processing based on environmental identification parameters, and mix the processed environmental sound with the voice of family members to generate a composite audio stream; A wireless transmission module is used to transmit the composite audio stream to a paired earphone through a Bluetooth protocol with a delay of ≤40ms, maintain a long connection for real-time voice insertion, and automatically adjust the output gain when moving outdoors in combination with positioning and motion data; A sound field synthesis module is used to convert the composite signal received by the earphone into sound waves, which are superimposed with the passively transmitted environmental sound in the ear canal to form an enhanced pseudo-environmental sound field; An interactive processing module is used to transcribe, analyze, and mark key feature segments with recall value in real time through ASR and NLP technology, upload structured content to the cloud, and trigger family member interaction to form a closed loop of emotional interaction data; A health management module is used to access smart wearable devices to obtain old people's vital signs, perform health monitoring and early warning, and synchronize data to the family member end through the communication unit; An intelligent home control module is used to analyze user voice commands and control smart home devices; An emergency processing module is used to detect user falls or long-term stillness events based on positioning unit and motion sensor data, and automatically trigger multi-level alarm linkage. The multi-level alarm linkage includes: sending an alarm with a location to family members and initiating a family circle multi-person call, continuously playing a warning sound to remind surrounding people, and automatically connecting to a preset emergency center if no one responds.

9. The hearing aid operation system for communication data of the elderly according to claim 8, wherein, The audio processing module includes: A noise reduction unit is used to separate stable and unstable noise in the environment using spectral subtraction-based noise suppression, and optimize the signal waveform using Wiener filtering. A filter enhancement unit is used to start adaptive band-pass filter to enhance target frequency band signal according to scene analysis result; A signal shaping unit is used to smooth and shape the signal and apply anti-aliasing filter processing to eliminate high frequency distortion and spectrum aliasing.

10. A hearing aid operating terminal for the elderly using communication data, characterized by The hearing aid running terminal comprises a storage device and a processor, the storage device is used for storing a computer program, and the processor runs the computer program to enable the hearing aid running terminal to execute the hearing aid running method in any one of claims 1-7.

Images