A programmable hearing aid detection method based on machine learning

Abstract

The present invention discloses a programmable hearing aid detection method based on machine learning, comprising the following steps: obtaining a relationship matrix, wherein the relationship matrix comprises access weights of different hearing aids; sorting a plurality of chip debugging interfaces according to the relationship matrix; and calling the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful. The programmable hearing aid detection method based on machine learning provided by the present invention solves the problem of slow automatic detection and reading speed for various models of hearing aids during hearing aid debugging, saves time, and can automatically optimize and improve the connection program according to the connection situation.

Description

Technical Field

[0001] The present invention relates to a programmable hearing aid detection method based on machine learning. Background Art

[0002] Current programmable digital hearing aids can be connected to a computer to configure the electro-acoustic amplification parameters to meet the needs of users with different hearing losses. Hearing aids need to be used in conjunction with computer programming software and programming tools.

[0003] When reading hearing aids through the programming software on the computer, since various brands and styles of hearing aids involve multiple chips, if the fitter is not clear about the hearing aid model, the programming software needs to traverse and call the debugging interface of each chip to automatically detect the hearing aid, which wastes time every time to read and debug the hearing aid. At the same time, for hearing aid manufacturers, the production of hearing aid models is often affected by market policies and product sales. It is also a waste of time to traverse and call the debugging interface of each chip to automatically detect the hearing aid. If the interface is accurately selected according to the production model each time, the operation steps and the risk of error are increased. Summary of the invention

[0004] The present invention provides a programmable hearing aid detection method based on machine learning to solve the above-mentioned technical problems, and specifically adopts the following technical solutions:

[0005] A method for detecting a programmable hearing aid based on machine learning comprises the following steps:

[0006] Acquire a relationship matrix, wherein the relationship matrix includes access weights of different hearing aids;

[0007] Sorting multiple chip debugging interfaces according to the relationship matrix;

[0008] The chip debugging interface is called according to the sorting order to establish a connection with the hearing aid until the connection is successful.

[0009] Furthermore, after calling the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful, the programmable hearing aid detection method based on machine learning further comprises:

[0010] A first correction is performed on the relationship matrix.

[0011] Furthermore, the specific method for performing the first correction on the relationship matrix is:

[0012] The value of the access weight corresponding to the hearing aid successfully connected this time in the relationship matrix is ​​increased by 1.

[0013] Furthermore, before performing the first correction on the relationship matrix, it is determined whether a special correction condition is met, and if not, the first correction is performed, and if satisfied, the special correction is performed on the relationship matrix.

[0014] Furthermore, the specific method for determining whether the special correction is currently met is:

[0015] When the same hearing aid is successfully connected continuously, it is determined whether the number of continuous connections a reaches a predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special modified condition is satisfied.

[0016] Furthermore, the specific method for determining whether the special correction is currently met is:

[0017] When the same hearing aid is successfully connected continuously within the first preset time, it is determined whether the number of continuous connections a reaches the predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special modified condition is satisfied.

[0018] Furthermore, the specific method for determining the predetermined number of times b is:

[0019] Taking the current hearing aid connection time as the starting point, the total connection times of all hearing aids within the second predetermined time are counted, and the predetermined number b is determined according to the total connection times.

[0020] Furthermore, the specific method for determining the value of the predetermined number of times b according to the total number of connections is:

[0021] If the total number of connections is less than or equal to the first preset value and greater than or equal to the second preset value, the predetermined number b is 3;

[0022] If the total number of connections is greater than the first preset value, the predetermined number b is set to 2;

[0023] If the total number of connections is less than the second preset value, the predetermined number b is 4.

[0024] Furthermore, the specific method for performing special correction on the relationship matrix is:

[0025] The value of the access weight corresponding to the hearing aid successfully connected this time in the relationship matrix is ​​increased by 2a.

[0026] Furthermore, before obtaining the relationship matrix, the programmable hearing aid detection method based on machine learning further comprises:

[0027] Before the first connection, the initial relationship matrix is ​​configured.

[0028] The benefit of the present invention lies in the provided programmable hearing aid detection method based on machine learning, which solves the problem of slow automatic detection and reading speed for various types of hearing aids during hearing aid debugging, saves time, and can automatically optimize and improve the connection program according to the connection situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0030] Figure 1 It is a schematic diagram of a programmable hearing aid detection method based on machine learning of the present invention. DETAILED DESCRIPTION

[0031] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0032] like Figure 1 The present invention shows a method for detecting a programmable hearing aid based on machine learning, which includes the following steps: S1: obtaining a relationship matrix, which includes access weights of different hearing aids. S2: sorting multiple chip debugging interfaces according to the relationship matrix. S3: calling the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful. The above steps are described in detail below.

[0033] For step S1: obtaining a relationship matrix, the relationship matrix includes access weights of different hearing aids.

[0034] When it is detected that a hearing aid is connected to the device, the relationship matrix is ​​first obtained.

[0035] Hearing aid type A = {A1, A2, A3, A4, ..., An-1, An}, chip type B = {B1, B2, B3, B4, ..., Bm-1, Bm}. Each hearing aid type An may be equipped with a different chip type Bm, so that when combined in pairs, m×n different types of hearing aids are formed. Each type of hearing aid has its own specific chip debugging interface, thus forming a two-dimensional relationship matrix T, which is:

[0036]

[0037] Among them, ω 11 Indicates using chip B 1 A 1 Access weight of hearing aid, ω n3 Indicates using chip B 3 A n Access weight of hearing aid. The so-called access weight reflects the probability that an unknown hearing aid is a specific model of hearing aid.

[0038] Therefore, before obtaining the relationship matrix, the programmable hearing aid detection method based on machine learning also includes:

[0039] Before the first connection, configure the initial relationship matrix.

[0040] The method of configuring the initial relationship matrix may be to set a balanced initial value, such as taking all ω values ​​as the same number, such as 10, so as to obtain a relationship matrix T in which each element is initially 10.

[0041] It is understandable that in order to make the initial relationship matrix T more scientific, a relationship matrix T that has been optimized on other devices can be imported into the current device. In this way, the initial relationship matrix T already reflects a more reasonable access weight of each hearing aid.

[0042] For step S2: sort the multiple chip debugging interfaces according to the relationship matrix.

[0043] Different ω values ​​in the relationship matrix T can reflect the market share of the corresponding hearing aids. This share largely reflects which hearing aid is currently connected to the programming device. In this way, according to the value of each element in the relationship matrix, the chip debugging interfaces corresponding to these elements are sorted from large to small.

[0044] For step S3: call the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful.

[0045] Then, according to the sorting, the chip debugging interface is called one by one in order to establish a connection with the hearing aid, and attempts are made continuously until the connection is successful, thus completing the hearing aid connection process.

[0046] As a preferred implementation, after the chip debugging interface is called according to the sorting order to establish a connection with the hearing aid until the connection is successful, the programmable hearing aid detection method based on machine learning further includes:

[0047] Perform a first revision of the relationship matrix.

[0048] In order to enable the relationship matrix T to be self-updated according to the actual connection situation, each time a hearing aid is connected, the relationship matrix is ​​first corrected according to the specific model of the currently connected hearing aid. The specific method for performing the first correction on the relationship matrix is:

[0049] Add 1 to the access weight of the hearing aid that is successfully connected in the relationship matrix.

[0050] As a preferred implementation, before performing the first correction on the relationship matrix, it is determined whether the conditions for the special correction are met. If not, the first correction is performed; if satisfied, the relationship matrix is ​​subjected to the special correction.

[0051] It is understandable that in order to make the relationship matrix T take shape more quickly, the distance between each other can be quickly pulled apart according to the number of accesses, so as to facilitate sorting. In the present application, after each access to the hearing aid, in addition to the more conventional first correction of the relationship matrix T, when the characteristic conditions are met, the relationship matrix T is also specially corrected.

[0052] In this application, the specific method for determining whether the special modification is currently met is:

[0053] When the same hearing aid is successfully connected continuously, it is determined whether the number of continuous connections a reaches a predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special correction condition is met.

[0054] The specific method for making special corrections to the relationship matrix is:

[0055] Add 2a to the access weight of the hearing aid that is successfully connected in the relationship matrix.

[0056] Distance description, assuming the predetermined number of times is 3, if a certain model of hearing aid is connected for 2 consecutive times, no special correction is performed. If a certain model of hearing aid is connected for 3 consecutive times, the access weight corresponding to the hearing aid model is increased by 6. If a certain model of hearing aid is connected for 4 consecutive times, the access weight corresponding to the hearing aid model is increased by 8, and so on.

[0057] As a more preferred implementation, the specific method for determining whether the special correction is currently satisfied is:

[0058] When the same hearing aid is successfully connected continuously within the first preset time, it is determined whether the number of continuous connections a reaches the predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special correction condition is met. Compared with the previous embodiment, in this embodiment, the statistical time of continuous access is limited. Only continuous access completed within the first predetermined time can be regarded as valid access. The first predetermined time can be set as needed. In this application, the first predetermined time is set to 24h. Starting from the current access time, the number of continuous connections of the hearing aid within 24h is counted forward.

[0059] It is understandable that the predetermined number of times b can be set according to specific circumstances. In this application, the specific method for determining the predetermined number of times b is:

[0060] Taking the current hearing aid connection time as the starting point, the total connection times of all hearing aids within the second predetermined time are counted, and the predetermined number b is determined according to the total connection times.

[0061] Furthermore, a specific method for determining the value of the predetermined number of times b according to the total number of connections is:

[0062] If the total number of connections is less than or equal to the first preset value and greater than or equal to the second preset value, the predetermined number b is 3.

[0063] If the total number of connections is greater than the first preset value, the predetermined number b is set to 2.

[0064] If the total number of connections is less than the second preset value, the predetermined number b is set to 4.

[0065] Specifically, in the present application, the total number of connections of all hearing aids within 72 hours is counted forward. When the total number of connections is between 50 and 100, the predetermined number b is 3. When the total number of connections is greater than 100, it indicates that the number of hearing aids currently connected to the device is large and the types are numerous, and it is more difficult to obtain continuous access opportunities. Therefore, the predetermined number b is taken to be slightly smaller. When the total number of connections is less than 20, it indicates that the number of hearing aids currently connected to the device is small and the types may be even fewer, and it is relatively easier to obtain continuous access opportunities. Therefore, the predetermined number b is taken to be slightly larger. It can be understood that the first preset value, the second preset value and the second predetermined time can be adjusted according to the specific circumstances, and are not limited here.

[0066] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A method for detecting programmable hearing aids based on machine learning, characterized in that: The following steps are involved: Acquire a relationship matrix, wherein the relationship matrix includes access weights of different hearing aids; Sorting multiple chip debugging interfaces according to the relationship matrix; Calling the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful; After calling the chip debugging interface to establish a connection with the hearing aid according to the sorting order until the connection is successful, the programmable hearing aid detection method based on machine learning further includes: Performing a first correction on the relationship matrix; The specific method for performing the first correction on the relationship matrix is: Add 1 to the access weight of the hearing aid that is successfully connected in the relationship matrix; Before performing the first correction on the relationship matrix, determining whether a special correction condition is met, if not, performing the first correction, if met, performing the special correction on the relationship matrix; The specific method for judging whether the special correction is currently met is: When the same hearing aid is successfully connected continuously, it is determined whether the number of continuous connections a reaches a predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special modified condition is satisfied.

2. The method for detecting a programmable hearing aid based on machine learning according to claim 1, characterized in that: The specific method for judging whether the special correction is currently met is: When the same hearing aid is successfully connected continuously within the first preset time, it is determined whether the number of continuous connections a reaches the predetermined number of times b. If the number of continuous connections a reaches the predetermined number of times b, it is determined that the special modified condition is satisfied.

3. The method for detecting a programmable hearing aid based on machine learning according to claim 2, characterized in that: The specific method for determining the predetermined number b is: Taking the current hearing aid connection time as the starting point, the total connection times of all hearing aids within the second predetermined time are counted, and the predetermined number b is determined according to the total connection times.

4. The method for detecting a programmable hearing aid based on machine learning according to claim 3, characterized in that: The specific method for determining the value of the predetermined number of times b according to the total number of connections is: If the total number of connections is less than or equal to the first preset value and greater than or equal to the second preset value, the predetermined number b is 3; If the total number of connections is greater than the first preset value, the predetermined number b is set to 2; If the total number of connections is less than the second preset value, the predetermined number b is 4.

5. The method for detecting a programmable hearing aid based on machine learning according to any one of claims 1 to 4, characterized in that: The specific method for making special corrections to the relationship matrix is: The value of the access weight corresponding to the hearing aid successfully connected this time in the relationship matrix is ​​increased by 2a.

6. The method for detecting a programmable hearing aid based on machine learning according to claim 1, characterized in that: Before obtaining the relationship matrix, the programmable hearing aid detection method based on machine learning further comprises: Before the first connection, the initial relationship matrix is ​​configured.

Images