A hearing aid and its parameter adjustment method and system

Abstract

This invention relates to the field of hearing aid technology, and discloses a hearing aid and a parameter adjustment method and system. The key technical points include a sound acquisition step, a semantic analysis step, and a sound adjustment step. Compared to existing professional parameter adjustment methods, this invention uses an algorithm to automatically adjust the parameters of the hearing aid in different frequency ranges according to the patient's hearing status, eliminating the need for professional parameter adjustment. Furthermore, compared to existing parameter adjustment methods, semantic analysis determines whether the wearer can hear the conversation clearly. Therefore, as the wearing time increases, the accuracy of the hearing aid adjustment can improve in real time, and with long-term use, the hearing aid can adaptively adjust itself in real time according to the wearer's hearing.

Description

Technical Field

[0001] This invention relates to the field of hearing aid technology, and more specifically to a hearing aid and a parameter adjustment method and system. Background Technology

[0002] Hearing aids are essentially small amplifiers that amplify sounds that patients cannot hear or cannot hear clearly, and then use the patient's residual hearing to send the sound to the auditory center of the brain so that the sound can be heard.

[0003] In current technology, hearing aid tuning requires professional medical personnel to measure ear parameters using an audiometer, obtain a spectral curve, mark missing frequencies, adjust various parameters of the hearing aid, and enhance missing or weakened frequencies. This tuning method requires the participation of specialized personnel, thus consuming and relying heavily on medical resources. Furthermore, patients' hearing changes over time, necessitating regular checkups. Therefore, existing hearing aid tuning methods consume significant human and material resources. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a hearing aid and a parameter adjustment method and system to overcome the above-mentioned defects in the existing technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a hearing aid and a parameter tuning method and system, comprising a sound acquisition step, a semantic analysis step, and a sound tuning step; Voice acquisition steps: 1. Generate the first human voice corpus and the second human voice corpus of the wearer's response; Semantic analysis steps include: 2.1 Analyze the first and second voice corpora to obtain the corresponding first and second text content; 2.2 Semantic analysis and matching of the first and second text contents. If the analysis result is good, the analysis stops; if the analysis result is poor, the sound debugging step is initiated. The sound tuning steps include: 3. Identifying the first human voice corpus through a spectrogram model and increasing the default amplitude of at least some frequency ranges.

[0006] Compared to existing methods of parameter adjustment by professionals, this invention uses an algorithm to automatically adjust the parameters of the hearing aid in different frequency ranges based on the patient's hearing status, eliminating the need for professional adjustment. Furthermore, compared to existing adjustment methods, it uses semantic analysis to determine whether the wearer can hear the conversation clearly. Therefore, as the wearing time increases, the accuracy of the hearing aid adjustment can improve in real time, and with long-term use, the hearing aid can adaptively adjust itself in real time according to the wearer's hearing.

[0007] As a further improvement of the present invention, when the analysis result in step 2.2 is good, the semantic analysis step further includes: 2.3.1 Obtain the first human voice corpus and generate the first spectrogram; 2.3.2 Call the spectrogram model to identify the first spectrogram and segment the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 2.3.3 Read the first amplitude values ​​in each frequency range of the spectrogram of several first syllables, compare the several first amplitude values ​​with the default amplitude values ​​in the same frequency range of the spectrogram model. If the first amplitude value is greater than the default amplitude value, the default amplitude value is retained. If the first amplitude value is less than the default amplitude value, the default amplitude value in the frequency range is replaced with the first amplitude value to generate a new default amplitude value.

[0008] In cases where the hearing of the wearer has improved or the amplitude adjustment has been too large, the present invention can adjust back the default amplitude within the frequency range. This not only ensures the accuracy of the default amplitude adjustment, but also enables the hearing aid to adaptively adjust according to the wearer's hearing condition.

[0009] Meanwhile, this invention can divide the first spectrogram into several first syllable spectrograms, making the spectrogram clearer and simpler. At the same time, during the calculation process, multiple spectrograms can be read and recognized separately, reducing the amount of computation in one operation and improving the accuracy of the calculation results.

[0010] As a further improvement of the present invention, when reading the first amplitude value in each frequency interval of a plurality of first syllable spectrograms, the method further includes: 2.3.4 Identify the default amplitude state within the same frequency range of the spectrogram model, and mark the frequency range of the default amplitude that is not marked.

[0011] In practical use, this invention marks the default amplitude within the frequency range that the wearer can clearly hear. The amplitude within the marked frequency range does not need to be adjusted again, thereby further increasing the accuracy of amplitude adjustment and reducing the number of amplitude enhancement adjustments. As a further improvement of the present invention, the sound tuning steps include: 3.1 Obtain the first human voice corpus and generate the first spectrogram; 3.2 Call the spectrogram model to identify the first spectrogram and segment the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 3.3 Call the spectrogram sorting algorithm to calculate the spectrogram of the first syllable, and obtain at least some of the first syllable spectrograms that need to be debugged; 3.4 Invoke the debugging algorithm to enhance the amplitude intensity of at least some frequency intervals in the spectrogram of the first syllable, and replace the default amplitude in the spectrogram model.

[0012] This invention can convert human speech corpus into spectrograms, and then divide the spectrograms into spectrograms of several individual syllables. The spectrograms of individual syllables are analyzed and calculated, and at least some frequency ranges in the spectrograms that are more likely to be inaudible are enhanced, thereby reducing the number of amplitude enhancement adjustments.

[0013] As a further improvement of the present invention, step 3.3 also includes: 3.3.1 Identify the frequencies of several first syllable spectrograms and match them with the frequency range of the spectrogram model, and exclude several first syllable spectrograms that are within the marked frequency range; 3.3.2 Sort the frequency intervals appearing in the first syllable spectrogram that are not marked according to the frequency of each frequency interval in the spectrogram model to obtain the frequency intervals to be enhanced.

[0014] This invention uses a spectrogram sorting algorithm to sort out the spectrogram of the marked first syllable and adjust it according to the frequency of occurrence of different sound frequency intervals in the spectrogram model (i.e. the part with the largest proportion of frequency intervals in the spectrogram model), thereby reducing the number of amplitude adjustments.

[0015] As a further improvement of the present invention, the spectrogram ranking algorithm includes:

[0016] in, As an evaluation value, This is a marker value; if it has been marked... Take 1 if it has not been marked. Take 0, Represents a custom frequency range. The corresponding spectral model The number of spectrograms of the first syllable within a frequency range The preset total number of frequency ranges.

[0017] The present invention can obtain the evaluation value of each frequency interval in the first corpus through the above formula. By comparing the magnitude of each evaluation value, the frequency interval with the highest priority can be selected.

[0018] As a further improvement of the present invention, the sound acquisition step includes: 1.1 Acquire the first audio segment and preprocess the audio segment to generate the first human voice corpus of electrical signals; 1.2 The first human voice corpus is analyzed, processed by a spectrogram model, and the processed first human voice corpus is converted into a sound signal by a hearing aid and played to the wearer. 1.3 Obtain the second audio segment of the wearer's response, process the second audio segment, and generate the second human voice corpus of electrical signals.

[0019] As a further improvement of the present invention, a hearing aid parameter adjustment system includes a sound acquisition module, a speech analysis module, and a sound adjustment module. The sound acquisition module is used to acquire a first human voice corpus and a second human voice corpus. The speech analysis module is used to convert the first human voice corpus and the second human voice corpus into text and analyze the semantics. The sound adjustment module adjusts the default amplitude according to the analysis results.

[0020] As a further improvement of the present invention, a hearing aid is provided with a hearing aid parameter adjustment system.

[0021] The beneficial effects of this invention are as follows: Compared to existing methods of parameter adjustment by professionals, this invention utilizes an algorithm to automatically adjust the parameters of the hearing aid in different frequency ranges according to the patient's hearing status, eliminating the need for professional parameter adjustment. Furthermore, compared to existing parameter adjustment methods, it determines whether the wearer can hear the conversation clearly through semantic analysis. Therefore, as the wearing time increases, the accuracy of the hearing aid adjustment can also improve in real time. Moreover, with long-term use, the hearing aid can adaptively adjust itself in real time according to the wearer's hearing.

[0022] In cases where the hearing of the wearer has improved or the amplitude adjustment has been too large, the present invention can adjust back the default amplitude within the frequency range. This not only ensures the accuracy of the default amplitude adjustment, but also enables the hearing aid to adaptively adjust according to the wearer's hearing condition. Attached Figure Description

[0023] Figure 1 This is a system module diagram of the present invention; Figure 2 It is a spectrogram of 0-10, where the horizontal axis represents time, the vertical axis represents frequency, and the color intensity represents amplitude.

[0024] Figure labels: 210, Spectrum model; 211, Debugging algorithm; 212, Spectrum sorting algorithm; 213, Speech analysis module; 214, Sound debugging module; 215, Sound acquisition module. Detailed Implementation

[0025] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Identical components are denoted by the same reference numerals. It should be noted that the terms "front," "rear," "left," "right," "upper," and "lower" used in the following description refer to directions in the accompanying drawings, and the terms "bottom surface," "top surface," "inner," and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0026] This embodiment of a hearing aid and parameter tuning method and system includes a sound acquisition step, a semantic analysis step, and a sound tuning step; Voice acquisition steps: 1. Generate the first human voice corpus and the second human voice corpus of the wearer's response; Semantic analysis steps include: 2.1 Analyze the first and second voice corpora to obtain the corresponding first and second text content; 2.2 Semantic analysis and matching of the first and second text contents. If the analysis result is good, the analysis stops; if the analysis result is poor, the sound debugging step is initiated. The sound tuning steps include: 3. Identifying the first human voice corpus through a spectrogram model and increasing the default amplitude of at least some frequency ranges.

[0027] In one embodiment, the generation of the first human voice corpus includes language spoken by others, i.e., questions asked by others when the wearer is actually conversing with them.

[0028] In one embodiment, the hearing aid is equipped with a testing module containing a number of questions. The testing module can convert the questions into first-person speech data and play it to the wearer.

[0029] In one embodiment, when the analysis result in step 2.2 is good, the semantic analysis step further includes: 2.3.1 Obtain the first human voice corpus and generate the first spectrogram; 2.3.2 Call the spectrogram model to identify the first spectrogram and segment the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 2.3.3 Read the first amplitude values ​​in each frequency range of the spectrogram of several first syllables, compare the several first amplitude values ​​with the default amplitude values ​​in the same frequency range of the spectrogram model. If the first amplitude value is greater than the default amplitude value, the default amplitude value is retained. If the first amplitude value is less than the default amplitude value, the default amplitude value in the frequency range is replaced with the first amplitude value to generate a new default amplitude value.

[0030] Specifically, the spectrogram includes three pieces of information: time, amplitude, and frequency, which are represented by coordinate axes, with the horizontal axis representing the amplitude and frequency. The audiogram includes two pieces of information: amplitude and frequency, which are represented by coordinate axes.

[0031] Specifically, the spectrogram model includes default amplitude values ​​for each frequency range. In the initial state, the default amplitude is 25 dB or less (i.e., normal hearing).

[0032] In one embodiment, when reading the first amplitude value within each frequency interval of a plurality of first syllable spectrograms, the method further includes: 2.3.4 Identify the default amplitude state within the same frequency range of the spectrogram model, and mark the frequency range of the default amplitude that is not marked.

[0033] In one embodiment, the sound tuning steps include: 3.1 Obtain the first human voice corpus and generate the first spectrogram; 3.2 Call the spectrogram model to identify the first spectrogram and segment the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 3.3 Call the spectrogram sorting algorithm to calculate the spectrogram of the first syllable, and obtain at least some of the first syllable spectrograms that need to be debugged; 3.4 Invoke the debugging algorithm to enhance the amplitude intensity of at least some frequency intervals in the spectrogram of the first syllable, and replace the default amplitude in the spectrogram model.

[0034] In one embodiment, the debugging algorithm is configured with a default debugging value, which can be adjusted and set by the system.

[0035] Specifically, the default debug value is 5dB.

[0036] In one embodiment, step 3.3 further includes: 3.3.1 Identify the frequencies of several first syllable spectrograms and match them with the frequency range of the spectrogram model, and exclude several first syllable spectrograms that are within the marked frequency range; 3.3.2 Sort the frequency intervals appearing in the first syllable spectrogram that are not marked according to the frequency of each frequency interval in the spectrogram model to obtain the frequency intervals to be enhanced.

[0037] In one embodiment, several spectrogram models are provided, including but not limited to a Mandarin Chinese spectrogram model and an English spectrogram model.

[0038] Specifically, for example, in the Mandarin model, the spectrograms of each first syllable are divided according to the pinyin syllables, and each syllable has a corresponding spectrogram, thus corresponding to words in different frequency ranges.

[0039] Furthermore, the specific numerical values ​​of the frequency range in the spectrogram of each syllable can be determined based on the spectrograms of most people's syllables.

[0040] In one embodiment, the spectrogram ranking algorithm includes:

[0041] in, As an evaluation value, This is a marker value; if it has been marked... Take 1 if it has not been marked. Take 0, Represents a custom frequency range. The corresponding spectral model The number of spectrograms of the first syllable within a frequency range (for example, in the Mandarin model, how many pinyin syllables contain this frequency in the frequency range of 200~500). The preset total number of frequency ranges.

[0042] Specifically, in the process of calculating the evaluation value, the frequency of the words appearing in the first corpus needs to be calculated multiple times according to the corresponding frequency range.

[0043] Specifically, the frequency range can be adjusted as needed, and the total frequency range is 100Hz~10000Hz by default.

[0044] Specifically, The meaning is as follows: t1~t2, where t1 represents the frequency end value of the frequency range, and t2 represents the other end value of the frequency range.

[0045] In one embodiment, the sound acquisition step includes: 1.1 Acquire the first audio segment and preprocess the audio segment to generate the first human voice corpus of electrical signals; 1.2 The first human voice corpus is analyzed and processed through a spectrogram model (i.e., if the amplitude of the sound part in the frequency range is less than the default amplitude, it is amplified to the default amplitude). The processed first human voice corpus is converted into a sound signal through the hearing aid and played to the wearer. 1.3 Obtain the second audio segment of the wearer's response, process the second audio segment, and generate the second human voice corpus of electrical signals.

[0046] In one embodiment, if the frequency ranges in the first corpus are all marked and the semantic analysis result is poor, the marking is removed. With this setting, if the wearer's hearing deteriorates further, the adjusted amplitude can be readjusted.

[0047] Specifically, when the flag is canceled, the default amplitude will increase from the previous default amplitude. (For example, in the 200Hz~500Hz frequency range, the original default amplitude was 25dB, and the readjustment will increase it from 25dB).

[0048] In one embodiment, a hearing aid parameter tuning system includes a sound acquisition module, a speech analysis module, and a sound tuning module. The sound acquisition module is used to acquire a first human voice corpus and a second human voice corpus. The speech analysis module is used to convert the first human voice corpus and the second human voice corpus into text and analyze the semantics. The sound tuning module adjusts the default amplitude based on the analysis results.

[0049] Specifically, the system is equipped with artificial intelligence algorithms such as ASR and NLP. The ASR algorithm can convert the first and second voice corpora into corresponding text content, while NLP can analyze the semantics of the first and second voice corpora to determine whether the semantics of the first and second voice corpora match.

[0050] In one embodiment, a hearing aid includes a hearing aid tuning system, comprising a processor, memory, a microphone, an amplifier, an earpiece, and a power supply. The system is stored in the memory, and the processor is used to run and compute the system in the memory.

[0051] Specifically, the hearing aid also has a built-in transmission module that communicates with the server (via WiFi, SIM card, etc.), allowing professionals to know the wearer's hearing status through communication.

[0052] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for adjusting parameters of a hearing aid, characterized in that: This includes the sound acquisition step, the semantic analysis step, and the sound tuning step; Voice acquisition steps:

1. Generate the first human voice corpus and the second human voice corpus of the wearer's response; The semantic analysis steps include: 2.1 Analyze the first and second voice corpora to obtain the corresponding first and second text content; 2.2 Semantic analysis and matching of the first and second text contents. If the analysis result is good, the analysis stops; if the analysis result is poor, the sound debugging step is initiated. The sound tuning steps include:

3. Identifying the first human voice corpus through a spectrogram model (210) and increasing the default amplitude of at least some frequency ranges; If the analysis result in step 2.2 is good, the semantic analysis step also includes: 2.3.1 Obtain the first human voice corpus and generate the first spectrogram; 2.3.2 Call the spectrogram model (210) to identify the first spectrogram and divide the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 2.3.3 Read the first amplitude values ​​in each frequency range of several first syllable spectrograms, compare several first amplitude values ​​with the default amplitude values ​​in the same frequency range of the spectrogram model (210), if the first amplitude value is greater than the default amplitude value, then retain the default amplitude value, if the first amplitude value is less than the default amplitude value, then replace the default amplitude value in that frequency range with the first amplitude value to generate a new default amplitude value.

2. The hearing aid parameter adjustment method according to claim 1, characterized in that: When reading the first amplitude value within each frequency range of several first syllable spectrograms, it also includes: 2.3.4 Identify the default amplitude state within the same frequency range of the spectrum model (210) and mark the frequency range of the default amplitude that is not marked.

3. The hearing aid parameter adjustment method according to claim 2, characterized in that: The sound tuning steps include: 3.1 Obtain the first human voice corpus and generate the first spectrogram; 3.2 Call the spectrogram model (210) to identify the first spectrogram and divide the first spectrogram into several spectrograms of the first syllables with individual pronunciations; 3.3 Call the spectrogram sorting algorithm (212) to calculate the spectrogram of the first syllable, and obtain at least some of the first syllable spectrograms that need to be debugged; 3.4 Invoke the debugging algorithm (211) to enhance the amplitude intensity of at least some of the frequency intervals in the spectrogram of the first syllable and replace the default amplitude in the spectrogram model (210).

4. The hearing aid parameter adjustment method according to claim 3, characterized in that: Step 3.3 also includes: 3.3.1 Identify the frequencies of several first syllable spectrograms and match them with the frequency range of the spectrogram model (210), and exclude several first syllable spectrograms that are in the marked frequency range; 3.3.2 Sort the frequency intervals appearing in the first syllable spectrogram that are not marked according to the frequency of each frequency interval in the spectrogram model (210) to obtain the frequency intervals to be enhanced.

5. A hearing aid parameter adjustment method according to claim 3 or 4, characterized in that: The spectrogram sorting algorithm includes: in, As an evaluation value, This is a marker value; if it has been marked... Take 1 if it has not been marked. Take 0, Represents a custom frequency range. The corresponding spectral model (210) represents The number of spectrograms of the first syllable within a frequency range. The preset total number of frequency ranges.

6. The hearing aid parameter adjustment method according to claim 1, characterized in that: The steps for acquiring sound include: 1.1 Acquire the first audio segment and preprocess the audio segment to generate the first human voice corpus of electrical signals; 1.2 Analyze the first human voice data, process the first human voice data through the spectrogram model (210), and convert the processed first human voice data into a sound signal through the hearing aid and play it to the wearer; 1.3 Obtain the second audio segment of the wearer's response, process the second audio segment, and generate the second human voice corpus of electrical signals.

7. A hearing aid parameter adjustment system, characterized in that: The device includes a sound acquisition module (215), a speech analysis module (213), and a sound tuning module (214). The sound acquisition module (215) is used to acquire a first human voice corpus and a second human voice corpus. The speech analysis module (213) is used to convert the first human voice corpus and the second human voice corpus into text and analyze the semantics. The sound tuning module (214) adjusts the default amplitude according to the analysis results. The sound acquisition module (215), the speech analysis module (213), and the sound tuning module (214) can respectively execute the sound acquisition step, the semantic analysis step, and the sound tuning step in the hearing aid parameter tuning method according to any one of claims 1-6.

8. A hearing aid, characterized in that: It is equipped with a hearing aid parameter adjustment system as described in claim 7.

Images