Hearing aid fitting method and device, computer equipment and storage medium

Abstract

The invention relates to a hearing aid fitting method and device, computer equipment and a storage medium. The method comprises the following steps: inputting a fitting text to a first language model, and outputting first adjustment data of the hearing aid; inputting the fitting text to a second language model, and outputting second adjustment data of the hearing aid; generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data; and adjusting hearing aid parameters according to a first selection result of the user on the first audio and the second audio. By adopting the method, the problems of high fitting cost and low fitting efficiency of the hearing aid can be solved.

Description

Technical Field

[0001] The present application relates to the field of hearing aids, and in particular to a hearing aid fitting method, apparatus, computer device and storage medium. Background Art

[0002] Before using a hearing aid, it is necessary to optimize the frequency, volume and other parameters of the hearing aid through various evaluations under the guidance of a professional. In the relevant fitting technology, manual fitting is the main method, and users need to perform fitting offline with professional fitters: fitters adjust various parameters of the hearing aid according to the user's condition to debug the user's hearing aid. During offline fitting, users need to go to the designated fitting location and complete the fitting process with professional fitters, which consumes more manpower and material resources.

[0003] In addition, there are some remote fitting methods, where professional fitting personnel can contact users in other places through the Internet to fit hearing aids. This also requires professional fitting technicians to provide one-on-one fitting services. Although online fitting saves some time costs compared to offline fitting, there are still problems of high fitting costs and low efficiency.

[0004] There is currently no effective solution to the problems of high cost and low efficiency in hearing aid fitting in related technologies. Summary of the invention

[0005] Based on this, it is necessary to provide a hearing aid fitting method, apparatus, computer equipment and storage medium that can achieve efficient fitting of hearing aids at low cost in response to the above technical problems.

[0006] In a first aspect, a hearing aid fitting method is provided in this embodiment, the method comprising:

[0007] Inputting the fitting text into the first language model, and outputting the first adjustment data of the hearing aid;

[0008] Inputting the fitting text into a second language model, and outputting second adjustment data of the hearing aid;

[0009] generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data;

[0010] The hearing aid parameters are adjusted according to the first selection result of the user between the first audio and the second audio.

[0011] In some embodiments, after adjusting the hearing aid parameters according to the first selection result of the user between the first audio and the second audio, the method further includes:

[0012] Determining whether the hearing aid meets a preset condition;

[0013] If not, update the fitting text, and repeat the steps of inputting the fitting text into the first language model, outputting the first adjustment data of the hearing aid; inputting the fitting text into the second language model, outputting the second adjustment data of the hearing aid; generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data; and adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, until the hearing aid meets the preset conditions.

[0014] In some embodiments, after inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method includes:

[0015] Obtain a second selection result of the user on the first audio and the second audio generated in the historical iteration process;

[0016] The first adjustment data and the second selection result are combined to obtain new first adjustment data.

[0017] In some embodiments, combining the first adjustment data and the second selection result to obtain new first adjustment data includes:

[0018] Obtaining third adjustment data generated in the historical iteration process according to the audio corresponding to the second selection result;

[0019] The first adjustment data and the third adjustment data under preset weights are combined to obtain the new first adjustment data.

[0020] In some embodiments, generating the first audio based on the first adjustment data includes:

[0021] The audio spectrum in the audio library is adjusted based on the first adjustment data to obtain the first audio.

[0022] In some embodiments, the step of inputting the fitting text to the first language model and outputting first adjustment data of the hearing aid includes:

[0023] The fitting text and the prompt language of the first language model are input into the first language model, and the first adjustment data is output.

[0024] In some embodiments, after inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method further includes:

[0025] inputting the first adjustment data, the second adjustment data and the first selection result into a preset thinking chain, and outputting an evaluation result of the first adjustment data based on the thinking chain;

[0026] In the case that the evaluation result does not meet the predetermined target, the first language model adjusts the output first adjustment data according to the fitting text and the evaluation result.

[0027] In a second aspect, a hearing aid fitting device is provided in this embodiment, the device comprising:

[0028] A first output module, used for inputting a fitting text into a first language model and outputting first adjustment data of the hearing aid;

[0029] A second output module, used for inputting the fitting text into a second language model, and outputting second adjustment data of the hearing aid;

[0030] A generating module, configured to generate a first audio based on the first adjustment data, and generate a second audio based on the second adjustment data;

[0031] The adjustment module is used to adjust the hearing aid parameters according to the first selection result of the user on the first audio and the second audio.

[0032] In a third aspect, a computer device is provided in this embodiment, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the hearing aid fitting method described in the first aspect is implemented.

[0033] In a fourth aspect, a computer-readable storage medium is provided in this embodiment, on which a computer program is stored. When the computer program is executed by a processor, the hearing aid fitting method described in the first aspect is implemented.

[0034] The hearing aid fitting method, apparatus, computer device and storage medium obtain different adjustment data by inputting the fitting text into different first language models and second language models, thereby improving the diversity of the adjustment data output, and simultaneously providing two audios for the user to choose from based on the adjustment data, and then obtaining an adjustment plan for the hearing aid based on the selection result, thereby achieving efficient fitting and reducing the hearing aid fitting cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A diagram of an application environment of a hearing aid fitting method in one embodiment;

[0036] Figure 2 is a flowchart of a hearing aid fitting method in one embodiment;

[0037] Figure 3 is a flowchart of a hearing aid fitting method in another embodiment;

[0038] Figure 4A schematic diagram of language model connection in one embodiment;

[0039] Figure 5 is a structural block diagram of a hearing aid fitting device in one embodiment;

[0040] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0042] The hearing aid fitting method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or it can be placed on the cloud or other network servers. Through the communication set between the terminal, the server or the terminal server, the two language models of the fitting text input are used to realize the output of the hearing aid adjustment data, and the audio data for the user to select is obtained, and the terminal adjusts the hearing aid parameters based on the user's selection result of the audio. Among them, the terminal 102 can be but is not limited to a hearing aid, or various personal computers, laptops, smart phones, tablet computers and portable wearable devices connected to the hearing aid and capable of adjusting the hearing aid parameters. The server 104 can be implemented with an independent server or a server cluster consisting of multiple servers.

[0043] In one embodiment, Figure 2 As shown, a hearing aid fitting method is provided, and the method is applied to Figure 1 The terminal 102 in the example is used as an example to illustrate, and the following steps are included:

[0044] Step S202: input the fitting text into the first language model, and output the first adjustment data of the hearing aid.

[0045] The fitting text is a natural language related to the hearing aid fitting process, which is used to indicate the adjustment direction of the hearing aid. The fitting text includes but is not limited to answers to preset fitting questions, experience feedback on the hearing aid, etc. The first language model is used to understand the fitting text and output the adjustment direction of the hearing aid based on the fitting text. The first adjustment data is used to indicate the adjustment direction of the hearing aid.

[0046] Optionally, after obtaining a fitting text provided by a user and inputting the fitting text into a first language model, first adjustment data output by the first language model is obtained, wherein the first adjustment data indicates an adjustment direction of the hearing aid through natural language.

[0047] Step S204: input the fitting text into the second language model, and output second adjustment data of the hearing aid.

[0048] Among them, the fitting text input into the first language model is the same as the fitting text of the second language model, the model parameters of the second language model and the first language model are different, and the first adjustment data and the second adjustment data are different data. Optionally, the first language model and the second language model are respectively a language model generated after training based on hearing aid text and a language model generated after training based on ordinary text. Alternatively, the first language model and the second language model are two models assigned different random parameters. With such a setting, the first language model and the second language model can infer different adjustment schemes based on the same problem.

[0049] Optionally, the first adjustment data is input into the second language model, and the second language model outputs the second adjustment data by understanding the first adjustment data. The second adjustment data also indicates the adjustment direction of the hearing aid through natural language. Since the model input and model parameters of the second language model and the first language model are different, the first adjustment data and the second adjustment data are different.

[0050] In order to further ensure that the first adjustment data and the second adjustment data are different, the first adjustment data and the fitting text can also be input into the second language model at the same time, and the second language model outputs the second adjustment data different from the first adjustment data by understanding the first adjustment data and the fitting text. Alternatively, while inputting the first adjustment data and the fitting text into the second language model, a preset instruction can also be input, and the preset instruction is used to instruct the second language model to output the second adjustment data different from the first adjustment data. Alternatively, the second language model can be instructed by a preset instruction to output multiple adjustment data with different priorities, and then, based on the order of priority from high to low, each adjustment data is compared with the first adjustment data. When it is determined that the adjustment data is different from the first adjustment data, the comparison is stopped and the current adjustment data is output as the second adjustment data.

[0051] Step S206: generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data.

[0052] The parameters of the first audio and the second audio are different, and the audio contents such as text and melody involved in the audio may be the same or different. Optionally, the parameters of the preset audio may be adjusted by the first adjustment data and the second adjustment data, respectively, and the first audio and the second audio may be outputted according to the adjusted parameters. The first audio and the second audio corresponding to the first adjustment data and the second adjustment data may also be generated based on the trained neural network.

[0053] Step S208: adjusting hearing aid parameters according to the user's first selection result of the first audio and the second audio.

[0054] Optionally, the audio selected by the user is determined according to the first selection result. Since different audios correspond to different parameters, the audio parameters selected by the user are used as parameter adjustment references to adjust the parameters of the current hearing aid.

[0055] Furthermore, based on the same method, multiple language models can be set. Taking the third language model as an example, the output of the first language model, the output of the second language model and the fitting problem are input into the third language model, and the third language model outputs adjustment data different from the other two language models. The input data and fitting problems of other language models are combined as the input of the current language model, and the data adjustment results different from those of other language models are output, so that the output of each language model is different, and multiple audios are obtained for users to choose.

[0056] In the hearing aid fitting method, the first adjustment data and the second adjustment data of the hearing aid are outputted respectively through the first language model and the second language model, and the first language model and the second language model have different parameters, so that the output results of the first language model and the second language model can be diversified, thereby improving the accuracy of the adjustment data output, thereby achieving efficient hearing aid fitting. After the user compares and selects the two audios corresponding to the two adjustment data outputted by the first language model and the second language model, the hearing aid can be adjusted, and the adjustment is convenient and low-cost, thereby solving the problem of high cost and low efficiency of hearing aid adjustment.

[0057] In one embodiment, after adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, the method also includes: determining whether the hearing aid meets the preset conditions; if not, updating the fitting text, and repeatedly inputting the fitting text to the first language model, outputting the first adjustment data of the hearing aid; inputting the fitting text to the second language model, outputting the second adjustment data of the hearing aid; generating the first audio based on the first adjustment data, and generating the second audio based on the second adjustment data; adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, until the hearing aid meets the preset conditions.

[0058] The preset condition is used to indicate that the hearing aid is adjusted. The preset condition can be obtained according to a user instruction. For example, when the user instruction indicates that the user is satisfied with the parameters of the hearing aid, it is determined that the hearing aid meets the preset condition. Alternatively, the preset condition can also be a preset number of iterations, time, and other conditions.

[0059] Wherein, updating the fitting text includes: obtaining a new fitting text based on user feedback or new fitting problems. Furthermore, based on feedback models such as prompt engineering and reflection systems, feedback reasoning is performed on the previous round of model output and user selection results to obtain feedback data, and the new fitting text is updated based on the feedback data.

[0060] Optionally, updating the fitting text includes: after adjusting the hearing aid parameters according to the first selection result, a new fitting text can be obtained based on user feedback or new fitting problems, and the above steps S202 to S208 are repeatedly performed based on the new fitting text until the user is satisfied with the parameters of the hearing aid, or, until the number of iterations reaches a specified number, it is determined whether the hearing aid meets the preset conditions, and the fitting of the hearing aid is stopped.

[0061] In order to further improve the adjustment efficiency of the hearing aid during the iteration process, in one embodiment, after inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method includes: obtaining the second selection result of the user on the first audio and the second audio generated during the historical iteration process; combining the first adjustment data and the second selection result to obtain new first adjustment data.

[0062] The second selection result may be a single selection of the audio by the user during a certain round of iterations, or may be multiple selections made by the user during multiple rounds of historical iterations. Optionally, the initial first adjustment data may be modified based on the third adjustment data; or, the third adjustment data and the first adjustment data may be merged to achieve a combination of the third adjustment data and the first adjustment data, and the weights of the third adjustment data and the first adjustment data are different during the merging process.

[0063] Further, merging the first adjustment data and the second selection result to obtain new first adjustment data includes: obtaining third adjustment data generated in the historical iteration process according to the audio corresponding to the second selection result; and merging the first adjustment data and the third adjustment data under a preset weight to obtain new first adjustment data. The preset weight is a pre-set weight that can be obtained during the training or testing process of the first language model and the second language model.

[0064] Among them, the audio selected by the user in the historical iteration process can be obtained according to the second selection result. Since each audio corresponds to the adjustment data one by one, the third adjustment data output by the first language model in the historical iteration process can be obtained based on the audio. After multiplying the third adjustment data with the preset weight, the third adjustment data after adjusting the weight is added to the first adjustment data to obtain new first adjustment data. It can be understood that the new second adjustment data can be obtained based on the same principle, which will not be repeated here.

[0065] In this embodiment, after the language model recognizes the input text and outputs the adjustment data, the current output is combined with the historical output, and inertia is used to allow the current output to break out of the local optimal solution, thereby finding the global optimal solution. The final adjustment parameters of the hearing aid can be obtained quickly, thereby improving the efficiency of hearing aid adjustment.

[0066] In one embodiment, generating a first audio based on first adjustment data includes: adjusting an audio spectrum in an audio library based on the first adjustment data to obtain the first audio. The audio library includes one or more preset audios. Optionally, after obtaining a preset mapping relationship between a textual description and a momentum parameter of the audio, the first adjustment data is mapped based on the preset mapping relationship to obtain a momentum parameter corresponding to the first adjustment data, and the spectrum of the audio in the audio library is adjusted based on the momentum parameter to obtain the first audio. Based on the same principle, generating a second audio includes: adjusting an audio spectrum in an audio library based on the second adjustment data to obtain a second audio, which will not be elaborated here.

[0067] Compared with the method in which the user is asked to compare the audio outputted in the previous round with the current audio and make a selection in the traditional fitting technology, in this embodiment, the spectrum of the audio in the audio library is adjusted by adjusting the data, so that the user can obtain the first audio outputted by the first language model and the second audio outputted by the second language model at the same time, and the adjustment range and direction are more intuitive. It can improve the user experience and enable the user to make an accurate selection, thereby achieving the effect of improving the fitting efficiency and accuracy.

[0068] In one embodiment, inputting the fitting text into the first language model and outputting first adjustment data of the hearing aid includes: inputting the fitting text and the prompt language of the first language model into the first language model, and outputting the first adjustment data.

[0069] The prompt language of the first language model may be a natural language text used to prompt the current role of the first language model. Alternatively, the prompt language may be obtained according to the second selection result made by the user in the historical iteration process. Optionally, the natural language text corresponding to the second selection result is used as the prompt language, so that the first language model can use it as a reference to understand the fitting text and perform reasoning to output the first adjustment book.

[0070] In this embodiment, by dynamically generating prompt language and adding the prompt language to the input of the first language model, the hearing aid fitting method can be made more flexible and have the function of self-optimizing the reasoning results to a certain extent, so that the output of the first language model becomes more accurate as the number of conversation rounds increases, thereby improving the fitting efficiency.

[0071] Furthermore, after inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method also includes: inputting the first adjustment data, the second adjustment data and the first selection result into a preset thinking chain, and outputting an evaluation result of the first adjustment data based on the thinking chain; when the evaluation result does not meet the predetermined target, the first language model adjusts the output first adjustment data according to the fitting text and the evaluation result.

[0072] The preset thought chain is used to review the output of the first language model and the second language model in this round, as well as the user selection results, so as to achieve the result of reflection and optimization of past dialogue results and act on subsequent answers. The evaluation result is used to verify whether the first adjustment data is reasonable, and the predetermined goal can be a pre-set evaluation parameter, evaluation conclusion, and other goals.

[0073] Optionally, if the evaluation result does not meet the predetermined target, in the next iteration, the fitting text and the evaluation result are input into the first language model together, so that the first language model can adjust the output based on the evaluation result to obtain new first adjustment data. Alternatively, if the evaluation result does not meet the predetermined target, the first language model is controlled to re-output the first adjustment data based on the fitting text and the evaluation result until the first adjustment data meets the predetermined target.

[0074] In this embodiment, by setting up a thought chain to review and feedback each iteration, the future model output is optimized based on the user's choice in the historical iteration process to increase the credibility and accuracy of the model output. Based on the same principle, the input and output of the second language model can also be optimized through prompt language and / or thought chain, which will not be described in detail.

[0075] Furthermore, the first adjustment data outputted can also be adjusted in combination with the preset thinking chain and the prompt language. Exemplarily, the fitting text and the prompt language of the first language model are input into the first language model, and the first adjustment data is outputted; the fitting text and the prompt language of the second language model are input into the second language model, and the second adjustment data is outputted. The first adjustment data, the prompt language of the first language model, and the first selection result are input into the preset thinking chain, and the evaluation result of the first adjustment data is outputted through the thinking chain. Similarly, the second adjustment data and the prompt language of the second language model are input into the preset thinking chain, and the evaluation result of the second adjustment data is outputted through the thinking chain. In the next round of iteration, the two language models adjust the outputs according to the corresponding evaluation results respectively.

[0076] In one embodiment, a hearing aid intelligent fitting system for performing hearing aid fitting is provided, wherein the hearing aid intelligent fitting system uses a large language model to replace a professional fitter to complete the task of understanding user semantics and solving user needs. As an intelligent agent with learning, reasoning, and decision-making capabilities, the large language model can make mutual collaborative decisions in a multi-agent system. Compared with the traditional multi-agent system based on reinforcement learning, the multi-agent system based on the large model has better learning and migration capabilities, can more effectively complete some high-dimensional tasks, and can save manpower, material resources, and learning and time costs. Among them, the language model parameters in the hearing aid intelligent fitting system are different. Optionally, one language model can adopt a conventional language model, such as the llama3 general language model; another language model can adopt a language model applied to the fitting field, such as the medical model of SenseTime.

[0077] Based on this hearing aid intelligent fitting system, Figure 3 A flowchart of another hearing aid fitting method is provided. Figure 3 As shown, the hearing aid intelligent fitting system is composed of two large language models, the first language model and the second language model. The two large models are respectively given different random parameters, and can reason about the same problem and give different adjustment schemes to obtain the first audio and the second audio. Figure 4 A schematic diagram of language model connection is provided, such as Figure 4 As shown in the figure, the input of the first language model is the user input, and the input of the second language model is the user input and the output of the first language model. This setting is intended to make the second language model generate different answers from the first language model, increasing the diversity of the outputs of the two language models. The answers of the two large models will be sent together to the subsequent other selection processing parts.

[0078] Continue to refer to Figure 3, the outputs of the second language model and the first language model are used to indicate the adjustment direction of the hearing aid, so the momentum parameter used to indicate the audio adjustment can be obtained according to the outputs of the second language model and the first language model. If the momentum parameter belongs to the momentum parameter outputted by the first iteration of this fitting, audio activation is performed based on the momentum parameter. Specifically, the frequency spectrum of the preset audio in the audio library is adjusted based on the momentum parameter outputted by the first language model, and the first audio is outputted. The frequency spectrum of the preset audio in the audio library is adjusted based on the momentum parameter outputted by the second language model, and the second audio is outputted. The user can independently choose the fitting adjustment direction of the hearing aid according to his or her own hearing experience of the two audio segments. In this way, not only can the problem of easily falling into the local optimal solution during the iteration of the language model be improved to a certain extent, but also the process of comparing the past audio and the current audio in the traditional fitting is transformed into the process of comparing the current two audios, so that the user can more intuitively feel the amplitude and direction of different audio adjustments.

[0079] In the case of multiple rounds of fitting, starting from the second round of iterative fitting, momentum is obtained by combining the hearing aid parameter adjustment strategy selected by the user in the historical iteration process. Taking the first language model as an example, starting from the second round of iterative fitting, after the first language model recognizes the user input and obtains the momentum parameter each time, the momentum parameter obtained in this round is combined with the momentum parameter selected in the user's historical iteration process, and inertia is used to make the answer result have a chance to break out of the local optimal solution and find the global optimal solution. The combination method is as follows:

[0080]

[0081] Among them, Action i is the momentum parameter selected in the current i-th dialogue, α is the weight parameter, ActionLLM i is the momentum parameter generated by the language model based on the input in the i-th dialogue, and the final strategy taken in the i-th step is Action i The weight parameter is selected in the range of 0.7 to 0.9, and the weight parameter can also be modified according to application requirements. The method of obtaining momentum by combining the second language model with the hearing aid parameter adjustment strategy selected by the user in the historical iteration process is similar.

[0082] After the user makes a selection based on the first audio and the second audio, the hearing aid is adjusted according to the momentum parameters corresponding to the audio selected by the user. Furthermore, the conversation data including user input, model output, and user selection will be stored in the database as historical conversations. In the next round of iterations, when the first language model and the second language model answer new questions, the conversation data of the previous round will be added to the input of the large language model through the prompt engineering method, so that the two language models can use the historical conversation as a reference for the next round of understanding and reasoning. The method of dynamically generating and adding prompt engineering also makes the entire system more flexible, and also has the function of self-optimizing the reasoning results to a certain extent, making the answers more accurate in multiple rounds of conversations.

[0083] Furthermore, in addition to optimizing the dialogue results by adding historical dialogue records through the prompting project, a reflection system can also be introduced into the hearing aid intelligent fitting system. The reflection system includes a preset thinking chain, which can infer and review the data of the language model based on the historical output data of the two language models and the corresponding user selection results, and act on the subsequent answers to achieve the result of reflecting and optimizing the past dialogue results. By using two feedback methods, the prompting project and the reflection system, in the hearing aid intelligent fitting system, the model has a certain ability of self-learning and correction, and can optimize future dialogue answers based on the content of past answers to increase credibility and accuracy.

[0084] In this embodiment, a large language model is used to replace a professional hearing aid fitter, and the fitting process is completed together with the user based on the semantic understanding and reasoning ability of the large language model. Compared with manual fitting, the fitting efficiency is high, the labor cost is low, and the use difficulty of intelligent fitting is low. At the same time, this embodiment takes into account that the text input by the user into the language model is often the text of a discrete single question, so the large model can only perform a single-step reasoning solution based on the current input. It may stop at a local optimal solution after processing several questions, and end the fitting in this state. In this case, the shorter model step size and the accumulation of some small directional errors will cause the model to miss the global optimal solution in another direction. In order to find the global optimal solution and improve the optimization fitting results, the multi-agent system introduces momentum parameters in combination with the historical answers and the output of the language model, so that the answer finally given by the large model jumps out of the local optimal solution as much as possible and finds the global optimal solution. Thereby achieving efficient, low-cost and high-accuracy hearing aid fitting.

[0085] It should be understood that, although the steps in the flowcharts involved in the above embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above embodiments may include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0086] Based on the same inventive concept, the embodiment of the present application also provides a hearing aid fitting device for implementing the hearing aid fitting method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more hearing aid fitting device embodiments provided below can refer to the limitations of the hearing aid fitting method above, and will not be repeated here.

[0087] In one embodiment, Figure 5 As shown, a hearing aid fitting device is provided, comprising: a first output module, a second output module, a generation module and an adjustment module, wherein:

[0088] A first output module, used for inputting a fitting text into a first language model and outputting first adjustment data of the hearing aid;

[0089] A second output module, used for inputting the fitting text into the second language model, and outputting second adjustment data of the hearing aid;

[0090] A generating module, configured to generate a first audio based on the first adjustment data, and generate a second audio based on the second adjustment data;

[0091] The adjustment module is used to adjust the hearing aid parameters according to the first selection result of the user on the first audio and the second audio.

[0092] In one embodiment, after adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, the first output module can also first determine whether the hearing aid meets the preset conditions; if not, update the fitting text, and repeat the steps of inputting the fitting text to the first language model, outputting the first adjustment data of the hearing aid; inputting the fitting text to the second language model, outputting the second adjustment data of the hearing aid; generating the first audio based on the first adjustment data, and generating the second audio based on the second adjustment data; adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, until the hearing aid meets the preset conditions.

[0093] Optionally, after the first output module inputs the fitting text into the first language model and outputs the first adjustment data of the hearing aid, the method includes: obtaining the second selection result of the user for the first audio and the second audio generated in the historical iteration process; merging the first adjustment data and the second selection result to obtain new first adjustment data. Optionally, the first output module merges the first adjustment data and the second selection result to obtain new first adjustment data, including: obtaining the third adjustment data generated in the historical iteration process according to the audio corresponding to the second selection result; merging the first adjustment data and the third adjustment data under the preset weight to obtain new first adjustment data.

[0094] In one embodiment, the generating module generates the first audio based on the first adjustment data, including: adjusting the audio spectrum in the audio library based on the first adjustment data to obtain the first audio.

[0095] In one embodiment, the first output module inputs the fitting text into the first language model and outputs the first adjustment data of the hearing aid, including: inputting the fitting text and the prompt language of the first language model into the first language model and outputting the first adjustment data. Optionally, after the first output module inputs the fitting text into the first language model and outputs the first adjustment data of the hearing aid, the method further includes: inputting the first adjustment data, the second adjustment data and the first selection result into a preset thinking chain, and outputting an evaluation result of the first adjustment data based on the thinking chain; if the evaluation result does not meet the predetermined target, the first language model adjusts the output first adjustment data according to the fitting text and the evaluation result.

[0096] Each module in the above-mentioned hearing aid fitting device can be implemented in whole or in part by software, hardware or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each module.

[0097] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store hearing aid fitting parameters and other data involved in executing a hearing aid fitting method. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a hearing aid fitting method is implemented.

[0098] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0099] In one embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0100] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0101] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0102] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0103] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0104] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A hearing aid fitting method, characterized in that: The method comprises: Inputting the fitting text into the first language model, and outputting the first adjustment data of the hearing aid; Inputting the fitting text into a second language model, and outputting second adjustment data of the hearing aid; generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data; The hearing aid parameters are adjusted according to the first selection result of the user between the first audio and the second audio.

2. The hearing aid fitting method according to claim 1, characterized in that: After adjusting the hearing aid parameters according to the first selection result of the user between the first audio and the second audio, the method further includes: Determining whether the hearing aid meets a preset condition; If not, update the fitting text, and repeat the steps of inputting the fitting text into the first language model, outputting the first adjustment data of the hearing aid; inputting the fitting text into the second language model, outputting the second adjustment data of the hearing aid; generating a first audio based on the first adjustment data, and generating a second audio based on the second adjustment data; and adjusting the hearing aid parameters according to the user's first selection result of the first audio and the second audio, until the hearing aid meets the preset conditions.

3. The hearing aid fitting method according to claim 2, characterized in that: After inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method includes: Obtain a second selection result of the user on the first audio and the second audio generated in the historical iteration process; The first adjustment data and the second selection result are combined to obtain new first adjustment data.

4. The method according to claim 3, characterized in that The combining the first adjustment data and the second selection result to obtain new first adjustment data includes: Obtaining third adjustment data generated in the historical iteration process according to the audio corresponding to the second selection result; The first adjustment data and the third adjustment data under preset weights are combined to obtain the new first adjustment data.

5. The method according to claim 1, characterized in that The generating the first audio based on the first adjustment data includes: The audio spectrum in the audio library is adjusted based on the first adjustment data to obtain the first audio.

6. The method according to any one of claim 1, characterized in that: The step of inputting the fitting text to the first language model and outputting first adjustment data of the hearing aid includes: The fitting text and the prompt language of the first language model are input into the first language model, and the first adjustment data is output.

7. The method according to claim 1 or 6, characterized in that: After inputting the fitting text into the first language model and outputting the first adjustment data of the hearing aid, the method further includes: inputting the first adjustment data, the second adjustment data and the first selection result into a preset thinking chain, and outputting an evaluation result of the first adjustment data based on the thinking chain; In the case that the evaluation result does not meet the predetermined target, the first language model adjusts the output first adjustment data according to the fitting text and the evaluation result.

8. A hearing aid fitting device, characterized in that: The device comprises: A first output module, used for inputting a fitting text into a first language model and outputting first adjustment data of the hearing aid; A second output module, used for inputting the fitting text into a second language model, and outputting second adjustment data of the hearing aid; A generating module, configured to generate a first audio based on the first adjustment data, and generate a second audio based on the second adjustment data; The adjustment module is used to adjust the hearing aid parameters according to the first selection result of the user on the first audio and the second audio.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images