WO2023060745A1 - 一种辅听耳机及其增益处理方法、装置及可读存储介质 - Google Patents

一种辅听耳机及其增益处理方法、装置及可读存储介质 Download PDF

Info

Publication number
WO2023060745A1
WO2023060745A1 PCT/CN2021/137867 CN2021137867W WO2023060745A1 WO 2023060745 A1 WO2023060745 A1 WO 2023060745A1 CN 2021137867 W CN2021137867 W CN 2021137867W WO 2023060745 A1 WO2023060745 A1 WO 2023060745A1
Authority
WO
WIPO (PCT)
Prior art keywords
gain
data
user data
user
listening
Prior art date
Application number
PCT/CN2021/137867
Other languages
English (en)
French (fr)
Inventor
童紫薇
刘春燕
Original Assignee
歌尔股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 歌尔股份有限公司 filed Critical 歌尔股份有限公司
Publication of WO2023060745A1 publication Critical patent/WO2023060745A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R25/00Deaf-aid sets, i.e. electro-acoustic or electro-mechanical hearing aids; Electric tinnitus maskers providing an auditory perception
    • H04R25/50Customised settings for obtaining desired overall acoustical characteristics
    • H04R25/505Customised settings for obtaining desired overall acoustical characteristics using digital signal processing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04RLOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
    • H04R2225/00Details of deaf aids covered by H04R25/00, not provided for in any of its subgroups
    • H04R2225/41Detection or adaptation of hearing aid parameters or programs to listening situation, e.g. pub, forest

Definitions

  • the present invention relates to the technical field of hearing aids, in particular to a hearing aid earphone and a gain processing method, device and system thereof.
  • the intensity range of speech is generally 50-100dBSPL.
  • the speech intensity is within the dynamic range.
  • the moderate speech intensity is the optimum threshold.
  • the speech intensity is different. comfort zone, and need to use auxiliary listening earphones to hear the sound of the outside world.
  • the gain of ordinary linear amplified listening earphones is fixed, that is, the gain provided for all sound intensities is the same, and only when the sound intensity is above the discomfort threshold can the peak clipping technology be used for processing.
  • the WDRC parameters of auxiliary listening earphones are fitted by professional fitters according to the preset calculation formula, which cannot be personalized according to the specific situation of the user, resulting in poor configuration accuracy and unable to provide more user-friendly
  • the gain of demand affects the user experience.
  • the purpose of the embodiments of the present invention is to provide a hearing aid earphone and its gain processing method, device and computer-readable storage medium, which can more accurately obtain the gain that meets the user's needs during use, and is conducive to improving user experience.
  • an embodiment of the present invention provides a gain processing method for a listening earphone, including:
  • the raw user data including initial hearing data
  • the gain data includes gain values corresponding to different frequencies and different loudnesses; wherein the gain model is based on multiple pre-collected histories Created from sample user data;
  • the establishment process of the gain model is:
  • KCCA KCCA algorithm to analyze multiple historical sample user data in the sample library to obtain a correlation coefficient corresponding to each of the historical sample data
  • a gain model is established according to the maximum correlation coefficient.
  • the pre-established gain model is used to analyze the processed user data to obtain gain data.
  • the process of preprocessing the initial hearing data in the original user data is:
  • the user data also includes age, gender, and place of origin.
  • An embodiment of the present invention also provides a gain processing device for a listening earphone, including:
  • An acquisition module configured to acquire the original user data of the user, the original user data including hearing data
  • An analysis module configured to analyze the original user data using a pre-established gain model to obtain gain data, the gain data including gain values corresponding to different loudnesses at different frequencies; wherein the gain model is based on pre-collected Created from multiple historical sample user data;
  • a setting module configured to set the auxiliary listening gain of the auxiliary listening earphone according to the gain data.
  • the embodiment of the present invention also provides an auxiliary listening earphone, including a left earphone, a right earphone, a memory and a processor, wherein:
  • the memory is used to store computer programs
  • the processor is configured to implement the steps of the above-mentioned gain processing method for a hearing aid earphone when executing the computer program.
  • An embodiment of the present invention also provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the above-mentioned method for processing the gain of an auxiliary listening earphone is implemented. step.
  • Embodiments of the present invention provide a gain processing method, device, system, and computer-readable storage medium for a listening earphone.
  • the method includes: acquiring original user data of a user, where the original user data includes initial hearing data; using a pre-established gain
  • the model analyzes the original user data to obtain gain data, which includes gain values corresponding to different frequencies and different loudnesses; the gain model is established based on multiple historical sample user data collected in advance.
  • the gain data corresponding to the user can be obtained, and the gain data includes gain values corresponding to different frequencies and different loudnesses respectively, according to
  • the gain data sets the auxiliary listening gain of the auxiliary listening earphone, and the earphone can output a sound that is more in line with the user's needs when working under the auxiliary listening gain.
  • FIG. 1 is a schematic flowchart of a gain processing method for a listening earphone provided by an embodiment of the present invention
  • FIG. 2 is a schematic diagram of a data transmission structure provided by an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a gain model architecture provided by an embodiment of the present invention.
  • FIG. 4 is a schematic diagram of gain processing of a gain model provided by an embodiment of the present invention.
  • Fig. 5 is a graph of output data corresponding to Fig. 4.
  • FIG. 6 is a schematic diagram of gain processing of another gain model provided by an embodiment of the present invention.
  • Fig. 7 is a graph corresponding to the output data of Fig. 6;
  • Fig. 8 is a schematic structural diagram of a gain processing device for a listening earphone provided by an embodiment of the present invention.
  • FIG. 1 is a schematic flowchart of a gain processing method for a hearing aid headphone provided by an embodiment of the present invention. The method includes:
  • S110 Acquire the original user data of the user, where the original user data includes initial hearing data
  • the user data of a large number of users are collected in advance, these user data are stored in the sample library as historical sample user data, and a gain model is established based on these historical sample user data, wherein the historical sample user data It includes hearing data and gain data corresponding to the user.
  • the historical sample user data in the embodiment of the present invention may also include user information in addition to hearing data and gain data, The user information may specifically include information such as age, gender, and place of origin, and of course may also include user location information.
  • the user's original user data can be obtained first, wherein the original user data It includes the initial hearing data of the user, and the initial hearing data may be obtained by performing a hearing test on the user by using auxiliary hearing earphones.
  • the original user data in the corresponding embodiment of the present invention may include information such as the user's age, gender, and place of origin in addition to the initial hearing data.
  • the gain data includes gain values corresponding to different frequencies and different loudnesses; the gain model is based on multiple historical sample users in the sample library in advance created from data;
  • the original user data is input into a pre-established gain model, and the original user data is analyzed through the gain model to obtain corresponding gain data.
  • the gain data in the embodiment of the present invention corresponds to the user's personal situation.
  • a better auxiliary listening effect can be obtained, so that the output of the auxiliary listening earphones The sound better matches the user's own needs.
  • the method may further include:
  • the original user data transmitted by the client can be accepted first.
  • the data can be received according to the data format shown in Figure 2.
  • the data can also be received by sampling XML or a custom format, and in During the transmission of the original user data, it can be transmitted in an encrypted form to prevent data leakage.
  • symmetric encryption or asymmetric encryption is used.
  • it can also be transmitted in a non-encrypted form. The details can be determined according to the actual situation. It needs to be determined, and this embodiment of the present invention makes no special limitation.
  • preprocessing can be performed on the original user data in order to eliminate abnormal data and obtain processed user data, so that the pre-established gain model of subsequent sampling can be used for the processed user data.
  • the data is analyzed to obtain the gain data corresponding to the user, so that the obtained gain data is more accurate and better meets the user's personal needs.
  • the preprocessing process can specifically be:
  • the initial hearing data in the original user data is converted into listening data (the loudness that can be heard at different frequencies), and the abnormal data in the listening data that is not within the preset range is eliminated to obtain the preprocessed user data.
  • the hearing threshold data is less than -10, the value is -10, and if the hearing threshold data is greater than 100, then the value is 100, so as to obtain effective two-dimensional data.
  • the historical sample user data is also preprocessed User data, this preprocessing is to process the original user data into the same dimensions and parameters as the historical sample user data.
  • the method may also include:
  • the gain data can also be displayed to the user through the terminal data device.
  • different Gain values of different loudness such as 20db, 35db, 50db, 65db, 80db, 95db, etc.
  • frequencies such as 125, 250, 500, 1000, 2000, 4000, 8000, etc.
  • Adjust the displayed gain data for example, adjust the gain value corresponding to a certain loudness at a certain frequency, so that the user can gain better and more Comfortably hear the corresponding sound clearly.
  • the gain data can be displayed through the mobile terminal (mobile phone) APP interface or the computer display interface, etc., and the user can adjust the gain data through the mobile terminal (mobile phone) APP interface or the computer display interface, etc., to obtain the most suitable for the user. gain value.
  • the method may also include:
  • the adjusted gain data is combined with the corresponding user data as a whole and added to the sample library as a new historical sample user data, so that by continuously updating the sample library Improvements are made so that better deep learning gain modules can be obtained when model training is performed using historical sample user data in the sample library.
  • the establishment process of the above-mentioned gain model may specifically be:
  • Adopt KCCA nuclear canonical correlation analysis
  • a gain model is established according to the maximum correlation coefficient.
  • each historical sample user data includes user sample data, hearing sample data, and gain sample data.
  • the user sample data and hearing sample data are used as input data, and the corresponding gain data is used as output data.
  • Groups of input data and output data are analyzed to obtain the correlation coefficient of each group of input data and the corresponding output data, so as to obtain multiple sets of correlation coefficients, and then find the maximum correlation coefficient from these multiple sets of correlation coefficients, and use the maximum correlation coefficient Build a gain model.
  • a deep learning gain module is established according to the maximum correlation coefficient, and as shown in Figure 3, the user's original user data is input into the gain model to obtain output data, which is the gain data corresponding to the user.
  • user sample data and hearing sample data can be used as input data.
  • user sample data mainly refers to the user's environmental noise, age, gender, hometown and other data, which is equivalent to the user's basic information
  • hearing sample data refers to What is the user’s gain data, that is, different users may have different gain data in different environments, the input data is taken as sample X, the corresponding gain data is taken as output data, and the output data is taken as sample Y.
  • the samples X, Y and the kernel function K(x, z) to define and calculate the kernel matrix K X and K Y , where the kernel function is:
  • x, z ⁇ X, X belongs to the R(n) space
  • n is the space dimension
  • x and z represent any two elements in the input sample
  • represents the hyperparameter of the RBF kernel.
  • ⁇ and ⁇ respectively represent the projection of the high-dimensional feature space on the vector, where ⁇ i is the i-th element of the vector ⁇ , ⁇ i is the i-th element of the vector ⁇ , and N' represents the vector ⁇ and ⁇ .
  • the correlation coefficient ⁇ ( ⁇ , v) is further calculated, where:
  • ⁇ i represents the i-th element of the vector ⁇
  • vi represents the i-th element of the vector v.
  • the above calculation process is used to calculate the corresponding correlation coefficient ⁇ ( ⁇ , v), and then the maximum correlation coefficient is selected from these correlation coefficients.
  • the gain model After calculating the correlation coefficient, the gain model can be constructed as follows
  • the correlation formula of the maximum correlation coefficient ⁇ is as follows (including but not limited to this linear method), and the gain data vector Y can be obtained
  • Y ⁇ 1*X1+ ⁇ 2*X2+...+ ⁇ n*Xn; where ⁇ 1 is the correlation coefficient between X1 and Y (gain data), and ⁇ n is the correlation coefficient between Xn and Y (gain data).
  • the gain data corresponding to the user can be obtained, and the gain data includes gain values corresponding to different frequencies and different loudnesses respectively, according to
  • the gain data sets the auxiliary listening gain of the auxiliary listening earphone, and the earphone can output a sound that is more in line with the user's needs when working under the auxiliary listening gain.
  • the embodiment of the present invention also provides a gain processing device for a listening earphone, please refer to Fig. 8 for details, the device includes:
  • the obtaining module 21 is used to obtain the original user data of the user, and the original user data includes hearing data;
  • the analysis module 22 is used to analyze the original user data by using a pre-established gain model to obtain gain data, and the gain data includes gain values corresponding to different loudnesses at different frequencies; wherein, the gain model is based on multiple historical samples collected in advance Created from user data;
  • the setting module 23 is configured to set the auxiliary listening gain of the auxiliary listening earphone according to the gain data.
  • the gain processing device for the earphones in the embodiment of the present invention has the same beneficial effect as the gain processing method for the earphones provided in the above-mentioned embodiments, and it has the same beneficial effect on the
  • the gain processing method of the auxiliary listening earphone please refer to the above-mentioned embodiments, and the present invention will not be repeated here.
  • an embodiment of the present invention also provides a listening earphone, including a left earphone, a right earphone, a memory, and a processor, wherein:
  • the processor is configured to implement the steps of the above-mentioned gain processing method for the earphone when executing the computer program.
  • the processor in the embodiment of the present invention is specifically used to obtain the original user data of the user, the original user data includes the initial hearing data; the pre-established gain model is used to analyze the original user data to obtain the gain data, the gain data includes different Gain values corresponding to frequencies and different loudnesses; the gain model is established based on multiple historical sample user data collected in advance; the auxiliary listening gain of the auxiliary listening earphone is set according to the gain data.
  • an embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the gain processing of the above-mentioned auxiliary listening earphone is realized. method steps.
  • the computer-readable storage medium may include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk, etc., which can store program codes. medium.
  • each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other.
  • the description is relatively simple, and for the related information, please refer to the description of the method part.
  • RAM random access memory
  • ROM read-only memory
  • EEPROM electrically programmable ROM
  • EEPROM electrically erasable programmable ROM
  • registers hard disk, removable disk, CD-ROM, or any other Any other known storage medium.

Landscapes

  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Neurosurgery (AREA)
  • Otolaryngology (AREA)
  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Acoustics & Sound (AREA)
  • Signal Processing (AREA)
  • Circuit For Audible Band Transducer (AREA)
  • Signal Processing For Digital Recording And Reproducing (AREA)

Abstract

本申请一些实施例公开了一种辅听耳机的增益处理方法、装置、系统及计算机可读存储介质,该方法包括:获取用户的原始用户数据,原始用户数据包括初始听力数据;采用预先建立的增益模型对原始用户数据进行分析,得到增益数据,增益数据包括不同频率下、不同响度分别对应的增益值;其中,增益模型为根据预先采集的多个历史样本用户数据建立而成的;根据增益数据对辅听耳机的辅听增益进行设置;本发明在使用过程中能够更加准确的得到满足用户需求的增益,有利于提高用户使用体验。

Description

一种辅听耳机及其增益处理方法、装置及可读存储介质
本申请要求于2021年10月15日提交中国专利局、申请号为202111204755.X、发明名称为“一种辅听耳机及其增益处理方法、装置及可读存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及辅听技术领域,特别是涉及一种辅听耳机及其增益处理方法、装置及系统。
背景技术
语音的强度范围一般为50~100dBSPL,对于正常听力者来说语音强度均在动态范围内,其中中等语音强度为最适阈,但是对于不同听阈感音神经性耳聋患者,其对语音强度有不同的舒适区,并且需要借助辅听耳机来听清外界的声音。
普通线性放大辅听耳机的增益是固定的,也即对所有的声强所提供的增益均相同,只有在声强在不适阈以上才采用削峰技术进行处理。通常辅听耳机的WDRC参数是由专业的验配师根据预先设置的计算公式进行验配而成的,不能够根据用户的具体情况进行个性化配置,导致配置精度差,不能够提供更符合用户需求的增益,影响用户使用体验。
鉴于此,如何提供一种能够提升用户使用体验的辅听耳机及其增益处理方法、装置及计算机可读存储介质成为本领域技术人员需要解决的问题。
发明内容
本发明实施例的目的是提供一种辅听耳机及其增益处理方法、装置及计算机可读存储介质,在使用过程中能够更加准确的得到满足用户需求的增益,有利于提高用户使用体验。
为解决上述技术问题,本发明实施例提供了一种辅听耳机的增益处理方法,包括:
获取用户的原始用户数据,所述原始用户数据包括初始听力数据;
采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据,所述增益数据包括不同频率下、不同响度分别对应的增益值;其中,所述增益模型为根据预先采集的多个历史样本用户数据建立而成的;
根据所述增益数据对辅听耳机的辅听增益进行设置。
可选的,所述增益模型的建立过程为:
采用KCCA算法对样本库中的多个历史样本用户数据进行分析,得到与每个所述历史样本数据对应的相关系数;
从各个所述相关系数中选择出最大相关系数;
依据所述最大相关系数建立增益模型。
可选的,在所述获取用户的原始用户数据之后,还包括:
将所述原始用户数据中的异常听力数据删除,得到处理后的用户数据;
则,所述采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据的过程为:
采用预先建立的增益模型对所述处理后的用户数据进行分析,得到增益数据。
可选的,所述对所述原始用户数据中的初始听力数据进行预处理的过程为:
将所述原始用户数据中的初始听力数据转换为听阀数据;
将所述听阀数据中不在预设范围内的异常数据剔除,得到预处理后的用户数据。
可选的,还包括:
将所述增益数据通过终端设备进行展示,以便用户对所述增益数据进行调节;
接收所述用户输入的调节指令,依据所述调节指令确定出目标响度及目标增益值;
将与所述目标响度对应的增益值调节至所述目标增益值。
可选的,还包括:
将调节后的增益数据及与所述用户对应的用户数据作为新的历史样本用户数据;
将所述新的历史样本用户数据添加至存储所述历史样本用户数据的样本库中,并对所述样本库进行更新。
可选的,所述用户数据还包括年龄、性别、籍贯。
本发明实施例还提供了一种辅听耳机的增益处理装置,包括:
获取模块,用于获取用户的原始用户数据,所述原始用户数据包括听力数据;
分析模块,用于采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据,所述增益数据包括不同频率下不同响度分别对应的增益值;其中,所述增益模型为根据预先采集的多个历史样本用户数据建立而成的;
设置模块,用于根据所述增益数据对辅听耳机的辅听增益进行设置。
本发明实施例还提供了一种辅听耳机,包括左耳机、右耳机、存储器和处理器,其中:
所述存储器,用于存储计算机程序;
所述处理器,用于执行所述计算机程序时实现如上述所述辅听耳机的增益处理方法的步骤。
本发明实施例还提供了一种计算机可读存储介质,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如上述所述辅听耳机的增益处理方法的步骤。
本发明实施例提供了一种辅听耳机的增益处理方法、装置、系统及计算机可读存储介质,该方法包括:获取用户的原始用户数据,原始用户数据包括初始听力数据;采用预先建立的增益模型对原始用户数据进行分析,得到增益数据,增益数据包括不同频率下、不同响度分别对应的增益值;其中,增益模型为根据预先采集的多个历史样本用户数据建立而成的。
可见,本发明实施例通过对预先建立的增益模型对获取到的原始用户数据进行分析,即可得到与用户对应的增益数据,该增益数据包括不同频率下、不同响度分别对应的增益值,根据该增益数据对辅听耳机的辅听增益进行设置,耳机在该辅听增益下工作能够输出更加符合用户需求的声音,本发明在使用过程中能够更加准确的得到满足用户需求的增益,有利于提高用户使用体验。
附图说明
为了更清楚地说明本申请实施例或现有技术中的技术方案,下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本申请的一部分附图,对于本领域普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据提供的附图获得其他的附图。
图1为本发明实施例提供的一种辅听耳机的增益处理方法的流程示意图;
图2为本发明实施例提供的一种数据传输结构示意图;
图3为本发明实施例提供的一种增益模型架构示意图;
图4为本发明实施例提供的一种增益模型的增益处理示意图;
图5为与图4对应的输出数据的曲线图;
图6为本发明实施例提供的另一种增益模型的增益处理示意图;
图7为与图6对应的输出数据的曲线图;
图8为本发明实施例提供的一种辅听耳机的增益处理装置的结构示意图。
具体实施方式
下面将结合本申请实施例中的附图,对本申请实施例中的技术方案进行描述,显然,所描述的实施例仅仅是本申请一部分实施例,而不是全部的实施例。基于本申请中的实施例,本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例,都属于本申请保护的范围。
请参照图1,图1为本发明实施例提供的一种辅听耳机的增益处理方法的流程示意图。该方法包括:
S110:获取用户的原始用户数据,原始用户数据包括初始听力数据;
需要说明的是,本发明实施例中预先采集大量用户的用户数据,将这些用户数据作为历史样本用户数据存储至样本库中,并根据这些历史样本用户数据建立增益模型,其中,历史样本用户数据包括对应用户的听力数据及增益数据。另外,为了使所建立的增益模型更加优化,以便通过该增益模型能够得到更优的增益数据,本发明实施例中历史样本用户数据除了包括听力数据及增益数据之外,还可以包括用户信息,该用户信息具体可以包括年龄、性别、籍贯等信息,当然还可以包括用户位置信息。
具体的,在针对需要使用辅听耳机的用户来说,在用户正常使用辅听耳机之前,需要确定出使用用户需求的辅听增益,具体可以先获取用户的原始用户数据,其中该原始用户数据包括用户的初始听力数据,该初始听力数据可以采用辅听耳机对用户进行听力测试得到的。当然,相应的本发明实施例中的原始用户数据除了包括初始听力数据之外,还可以包括用户的年龄、性别、籍贯等信息。
S120:采用预先建立的增益模型对原始用户数据进行分析,得到增益数据,增益数据包括不同频率下、不同响度分别对应的增益值;其中,增益模型为预先根据样本库中的多个历史样本用户数据建立而成的;
需要说明的是,在得到用户的原始用户数据后,将该原始用户数据输入至预先建立的增益模型中,通过该增益模型对该原始用户数据进行分析,得到对应的增益数据。
S130:根据增益数据对辅听耳机的辅听增益进行设置。
具体的,本发明实施例中的增益数据是与用户的本人情况对应的,通过该增益数据对辅听耳机进行辅听增益调节后,能够获得较好的辅听效果,使辅听耳机输出的声音更好符合该用户的自身需求。
进一步的,在上述S110获取用户的原始用户数据之后,该方法还可以包括:
将原始用户数据中的异常听力数据删除,得到处理后的用户数据;
需要说明的是,在实际应用中可以先接受客户端传输的原始用户数据,具体可以按照图2所示的数据格式进行数据的接收,当然还可以采样XML或自定义格式进行数据接收,并且在对原始用户数据传输过程中可以以加密的形式进行传输,以防止数据的泄漏,例如此案有对称加密或非对称加密的方式进行,当然,也可以采用非加密的形式传输,具体可以根据实际需要进行确定,本发明实施例不作特殊限定。
具体的,在接收到用户的原始用户数据后,可以对原始用户数据进行预处理,以便剔除不正常的数据,得到处理后的用户数据,以便后续采样预先建立的增益模型对该处理后的用户数据进行分析,得到与该用户对应的增益数据,从而使得到的增益数据更加准确,更好的满足该用户的个人需求。
其中,预处理的过程具体可以为:
将原始用户数据中的初始听力数据转换为听阀数据(不同频率下能够听到的响度),并将听阀数据中不在预设范围内的异常数据剔除,得到预处理后的用户数据。其中,如果听阈数据小于-10,则取值为-10,如果听阈数据大于100,则取值为100,从而获得有效的二维数据,需要说明的是,历史样本用户数据也是经过预处理的用户数据,本次预处理是将原始用户数据处理为与历史样本用户数据相同的维度和参数。
更进一步的,该方法还可以包括:
将增益数据通过终端设备进行展示,以便用户对增益数据进行调节;
接收用户输入的调节指令,依据调节指令确定出目标响度及目标增益值;
将与目标响度对应的增益值调节至目标增益值。
需要说明的是,为了进一步满足用户的个性化需求,以更好的满足用户需求,本发明实施例中在得到增益数据后,还可以通过终端数设备将增益数据展示给用户,具体的将不同频率(例如125、250、500、1000、2000、4000、8000等)下不同响度(例如20db、35db、50db、65db、80db、95db等)的增益值均展示给用户,用户可以根据自身需要对所展示的增益数据进行调节,例如对某个频率下某一个响度对应的增益值进行调节,以使在该频率下按照调节后的增益值对对应的响度进行增益处理后用户能够更好、更舒适地听清楚对应的声音。具体的,可以通过移动终端(手机)APP界面或电脑显示界面等对增益数据进行展示,并使用户通过移动终端(手机)APP界面或电脑显示界面等对增益数据进行调节,以获得最适合用户的增益值。
进一步的,该方法还可以包括:
将调节后的增益数据及与用户对应的用户数据作为新的历史样本用户数据;
将新的历史样本用户数据添加至样本库,并对样本库进行更新。
需要说明的是,本发明实施例中在用户对增益数据调节后,将调节后的增益数据结合对应的用户数据整体作为一个新的历史样本用户数据添加至样本库中,从而通过不断对样本库进行完善,以便通过样本库中的历史样本用户数据进行模型训练时,得到更优的深度学习增益模块。
进一步的,上述增益模型的建立过程,具体可以为:
采用KCCA(核典型相关分析)算法对样本库中的多个历史样本用户数 据进行分析,得到与每个所述历史样本数据对应的相关系数;
从各个所述相关系数中选择出最大相关系数;
依据所述最大相关系数建立增益模型。
需要说明的是,每个历史样本用户数据包括用户样本数据、听力样本数据及增益样本数据,将用户样本数据和听力样本数据作为输入数据,将对应的增益数据作为输出数据,采用KCCA算法对每组输入数据和输出数据进行分析,得到每组输入数据和对应的输出数据的相关系数,从而得到多组相关系数,然后从这多组相关系数中找出最大相关系数,并采用该最大相关系数建立增益模型。
具体的,根据该最大相关系数建立深度学习增益模块,并如图3所示,将用户的原始用户数据输入至增益模型中得到输出数据,该输出数据即为与该用户对应的增益数据。
具体的,可以将用户样本数据和听力样本数据作为输入数据,其中,用户样本数据:主要是指用户的环境噪音、年龄、性别、籍贯等数据,相当于用户的基本信息;听力样本数据:指的是用户的增益数据,即不同用户在不同环境下可能有不同的增益数据,将输入数据作为样本X,将对应的增益数据作为输出数据,将输出数据作为样本Y。并采用样本X、Y以及核函数K(x,z)定义计算核矩阵K X和K Y,其中核函数为:
Figure PCTCN2021137867-appb-000001
x,z∈X,X属于R(n)空间,n为空间维度,x和z表示输入样本中的任意两个元素,σ表示RBF核的超参数。
根据关系式
Figure PCTCN2021137867-appb-000002
计算出M、L和N,其中,J=I-λλ T,λ=(1,...,1) T,I表示单位向量,η表示常数因子,通过该约束,使在高维数特征空间内产生有意义的典型变量,K X=ΦX′ΦX,ΦX是输入数据的向量形式。
根据关系式L -1MN -1M Tα=λ 2α和N -1M TL -1Mβ=λ 2β,进一步计算出向量α和β,其中,λ表示单位向量。
根据关系式
Figure PCTCN2021137867-appb-000003
Figure PCTCN2021137867-appb-000004
计算出μ和ν分别表示高维数特征空间在向量上的投影,其中,α i为向量α的第i个元素,β i为向量β的第i个元素,N'表示向量α和β的元素总数量,K x(X i,X)和K Y(Y i,Y)均表示核矩阵,X i表示向量X的第i个元素,Y i表示向量Y的第i个元素。根据μ和ν 进一步计算出相关系数ρ(μ,v),其中:
Figure PCTCN2021137867-appb-000005
其中,μ i表示向量μ的第i个元素,vi表示向量v的第i个元素。
针对每一组输入数据和输出数据采用上述计算过程计算出对应的相关系数ρ(μ,v),然后再从这些相关系数中选择出最大相关系数。
计算出相关系数后,增益模型的构建即可如下
通过输入一个用户样本数据X,经过最大相关系数ρ的相关的一个公式如下(包括但不限于这种线性方式),即可得到增益数据向量Y
Y=ρ1*X1+ρ2*X2+...+ρn*Xn;其中ρ1为X1与Y(增益数据)之间的相关系数,ρn为Xn与Y(增益数据)之间的相关系数。
另外,如图4至图7所示,针对某用户在40db噪声下得到的增益数据,其中,图4中将用户的听阀数据作为输入数据,得到的输出数据如图5和表1所示,图6中将用户的用户数据(例如年龄、性别、籍贯等)和听阀数据作为输入数据,得到的输出数据如图7和表2所示。可见,在更多训练输入条件下,得出的增益数据对患者有更好的增益。其中:
表1
  125 250 500 1000 2000 4000
原始 42 63 56 45 44 37
35db 37 55 49 40 39 32
50db 45 67 60 48 47 40
65db 29 43 38 32 31 24
80db 23 28 28 26 25 18
95db 15 20 20 18 17 10
表2
  125 250 500 1000 2000 4000
原始 42 63 56 45 44 37
35db 40 58 52 43 42 35
50db 47 69 62 50 49 42
65db 31 44 41 35 32 25
80db 24 29 29 27 26 19
95db 15 20 20 18 17 10
可见,本发明实施例通过对预先建立的增益模型对获取到的原始用户数 据进行分析,即可得到与用户对应的增益数据,该增益数据包括不同频率下、不同响度分别对应的增益值,根据该增益数据对辅听耳机的辅听增益进行设置,耳机在该辅听增益下工作能够输出更加符合用户需求的声音,本发明在使用过程中能够更加准确的得到满足用户需求的增益,有利于提高用户使用体验。
在上述实施例的基础上,本发明实施例还提供了一种辅听耳机的增益处理装置,具体请参照图8,该装置包括:
获取模块21,用于获取用户的原始用户数据,原始用户数据包括听力数据;
分析模块22,用于采用预先建立的增益模型对原始用户数据进行分析,得到增益数据,增益数据包括不同频率下不同响度分别对应的增益值;其中,增益模型为根据预先采集的多个历史样本用户数据建立而成的;
设置模块23,用于根据增益数据对辅听耳机的辅听增益进行设置。
需要说明的是,本发明实施例中的辅听耳机的增益处理装置具有与上述实施例中所提供的辅听耳机的增益处理方法相同的有益效果,并且对于本发明实施例中所涉及到的辅听耳机的增益处理方法的具体介绍请参照上述实施例,本发明在此不再赘述。
在上述实施例的基础上,本发明实施例还提供了一种辅听耳机,包括左耳机、右耳机、存储器和处理器,其中:
存储器,用于存储计算机程序;
处理器,用于执行计算机程序时实现如上述辅听耳机的增益处理方法的步骤。
例如,本发明实施例中的处理器具体用于实现获取用户的原始用户数据,原始用户数据包括初始听力数据;采用预先建立的增益模型对原始用户数据进行分析,得到增益数据,增益数据包括不同频率下、不同响度分别对应的增益值;其中,增益模型为根据预先采集的多个历史样本用户数据建立而成的;根据增益数据对辅听耳机的辅听增益进行设置。
在上述实施例的基础上,本发明实施例还提供了一种计算机可读存储介 质,计算机可读存储介质上存储有计算机程序,计算机程序被处理器执行时实现如上述辅听耳机的增益处理方法的步骤。
该计算机可读存储介质可以包括:U盘、移动硬盘、只读存储器(Read-Only Memory,ROM)、随机存取存储器(Random Access Memory,RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
本说明书中各个实施例采用递进的方式描述,每个实施例重点说明的都是与其他实施例的不同之处,各个实施例之间相同相似部分互相参见即可。对于实施例公开的装置而言,由于其与实施例公开的方法相对应,所以描述的比较简单,相关之处参见方法部分说明即可。
本领域普通技术人员还可以理解,结合本文中所公开的实施例描述的各示例的单元及算法步骤,能够以电子硬件、计算机软件或者二者的结合来实现,为了清楚地说明硬件和软件的可互换性,在上述说明中已经按照功能一般性地描述了各示例的组成及步骤。这些功能究竟以硬件还是软件方式来执行,取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为超出本申请的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以直接用硬件、处理器执行的软件模块,或者二者的结合来实施。软件模块可以置于随机存储器(RAM)、内存、只读存储器(ROM)、电可编程ROM、电可擦除可编程ROM、寄存器、硬盘、可移动磁盘、CD-ROM、或技术领域内所公知的任意其它形式的存储介质中。
还需要说明的是,在本文中,诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来,而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外 的相同要素。

Claims (10)

  1. 一种辅听耳机的增益处理方法,其特征在于,包括:
    获取用户的原始用户数据,所述原始用户数据包括初始听力数据;
    采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据,所述增益数据包括不同频率下、不同响度分别对应的增益值;其中,所述增益模型为根据预先采集的多个历史样本用户数据建立而成的;
    根据所述增益数据对辅听耳机的辅听增益进行设置。
  2. 根据权利要求1所述的辅听耳机的增益处理方法,其特征在于,所述增益模型的建立过程为:
    采用KCCA算法对样本库中的多个历史样本用户数据进行分析,得到与每个所述历史样本数据对应的相关系数;
    从各个所述相关系数中选择出最大相关系数;
    依据所述最大相关系数建立增益模型。
  3. 根据权利要求1所述的辅听耳机的增益处理方法,其特征在于,在所述获取用户的原始用户数据之后,还包括:
    将所述原始用户数据中的异常听力数据删除,得到处理后的用户数据;则,所述采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据的过程为:
    采用预先建立的增益模型对所述处理后的用户数据进行分析,得到增益数据。
  4. 根据权利要求3所述的辅听耳机的增益处理方法,其特征在于,所述将所述原始用户数据中的异常听力数据删除的过程为:
    将所述原始用户数据中的初始听力数据转换为听阀数据;
    将所述听阀数据中不在预设范围内的异常数据剔除,得到预处理后的用户数据。
  5. 根据权利要求1至4任意一项所述的辅听耳机的增益处理方法,其特征在于,还包括:
    将所述增益数据通过终端设备进行展示,以便用户对所述增益数据进行调节;
    接收所述用户输入的调节指令,依据所述调节指令确定出目标响度及目 标增益值;
    将与所述目标响度对应的增益值调节至所述目标增益值。
  6. 根据权利要求5所述的辅听耳机的增益处理方法,其特征在于,还包括:
    将调节后的增益数据及与所述用户对应的用户数据作为新的历史样本用户数据;
    将所述新的历史样本用户数据添加至存储所述历史样本用户数据的样本库中,并对所述样本库进行更新。
  7. 根据权利要求1所述的辅听耳机的增益处理方法,其特征在于,所述用户数据还包括年龄、性别、籍贯。
  8. 一种辅听耳机的增益处理装置,其特征在于,包括:
    获取模块,用于获取用户的原始用户数据,所述原始用户数据包括听力数据;
    分析模块,用于采用预先建立的增益模型对所述原始用户数据进行分析,得到增益数据,所述增益数据包括不同频率下不同响度分别对应的增益值;其中,所述增益模型为根据预先采集的多个历史样本用户数据建立而成的;
    设置模块,用于根据所述增益数据对辅听耳机的辅听增益进行设置。
  9. 一种辅听耳机,其特征在于,包括左耳机、右耳机、存储器和处理器,其中:
    所述存储器,用于存储计算机程序;
    所述处理器,用于执行所述计算机程序时实现如权利要求1至7任一项所述辅听耳机的增益处理方法的步骤。
  10. 一种计算机可读存储介质,其特征在于,所述计算机可读存储介质上存储有计算机程序,所述计算机程序被处理器执行时实现如权利要求1至7任一项所述辅听耳机的增益处理方法的步骤。
PCT/CN2021/137867 2021-10-15 2021-12-14 一种辅听耳机及其增益处理方法、装置及可读存储介质 WO2023060745A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111204755.XA CN113873416B (zh) 2021-10-15 2021-10-15 一种辅听耳机及其增益处理方法、装置及可读存储介质
CN202111204755.X 2021-10-15

Publications (1)

Publication Number Publication Date
WO2023060745A1 true WO2023060745A1 (zh) 2023-04-20

Family

ID=78999835

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/137867 WO2023060745A1 (zh) 2021-10-15 2021-12-14 一种辅听耳机及其增益处理方法、装置及可读存储介质

Country Status (2)

Country Link
CN (1) CN113873416B (zh)
WO (1) WO2023060745A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110106508A1 (en) * 2007-08-29 2011-05-05 Phonak Ag Fitting procedure for hearing devices and corresponding hearing device
CN104937954A (zh) * 2013-01-09 2015-09-23 听优企业 用于自管理声音增强的方法和系统
CN110708652A (zh) * 2019-11-06 2020-01-17 佛山博智医疗科技有限公司 一种利用自身语音信号调节助听设备的系统及方法
CN113194395A (zh) * 2021-04-23 2021-07-30 歌尔股份有限公司 辅听设备的参数调整方法、装置、系统及可读存储介质
US20220078564A1 (en) * 2019-01-09 2022-03-10 The Trustees Of Indiana University System and method for individualized hearing air prescription

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2019024083A1 (en) * 2017-08-04 2019-02-07 Nokia Technologies Oy ARTIFICIAL NEURONAL NETWORK

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110106508A1 (en) * 2007-08-29 2011-05-05 Phonak Ag Fitting procedure for hearing devices and corresponding hearing device
CN104937954A (zh) * 2013-01-09 2015-09-23 听优企业 用于自管理声音增强的方法和系统
US20220078564A1 (en) * 2019-01-09 2022-03-10 The Trustees Of Indiana University System and method for individualized hearing air prescription
CN110708652A (zh) * 2019-11-06 2020-01-17 佛山博智医疗科技有限公司 一种利用自身语音信号调节助听设备的系统及方法
CN113194395A (zh) * 2021-04-23 2021-07-30 歌尔股份有限公司 辅听设备的参数调整方法、装置、系统及可读存储介质

Also Published As

Publication number Publication date
CN113873416A (zh) 2021-12-31
CN113873416B (zh) 2022-09-13

Similar Documents

Publication Publication Date Title
US11122374B2 (en) Systems and methods for providing personalized audio replay on a plurality of consumer devices
Glasberg et al. Prediction of absolute thresholds and equal-loudness contours using a modified loudness model
Nielsen et al. Perception-based personalization of hearing aids using Gaussian processes and active learning
Oetting et al. Restoring perceived loudness for listeners with hearing loss
EP2031900A2 (en) Hearing aid fitting procedure and processing based on subjective space representation
Wu et al. The effect of hearing aid signal-processing schemes on acceptable noise levels: perception and prediction
Vaerenberg et al. A uniform graphical representation of intensity coding in current-generation cochlear implant systems
CN106233754A (zh) 听力辅助设备控制
US20170195811A1 (en) Audio Monitoring and Adaptation Using Headset Microphones Inside User's Ear Canal
Rasetshwane et al. Categorical loudness scaling and equal-loudness contours in listeners with normal hearing and hearing loss
Vroegop et al. How to optimally fit a hearing aid for bimodal cochlear implant users: A systematic review
CN114007177B (zh) 助听控制方法、装置、助听设备和存储介质
Whilby et al. Monaural and binaural loudness of 5-and 200-ms tones in normal and impaired hearing
WO2020214482A1 (en) Self-fit hearing instruments with self-reported measures of hearing loss and listening
CN107534820A (zh) 用于驱动动态压缩器的装置和方法以及用于确定动态压缩器的放大值的方法
Salehi et al. Learning-based reference-free speech quality measures for hearing aid applications
English et al. Fitting recommendations and clinical benefit associated with use of the NAL-NL2 hearing-aid prescription in Nucleus cochlear implant recipients
Anderson et al. The use of self-report measures to examine changes in perception in response to fittings using different signal processing parameters
WO2023060745A1 (zh) 一种辅听耳机及其增益处理方法、装置及可读存储介质
Hodgetts et al. Technology-limited and patient-derived versus audibility-derived fittings in bone-anchored hearing aid users: a validation study
Portelli et al. Functional outcomes for speech-in-noise intelligibility of NAL-NL2 and DSL v. 5 prescriptive fitting rules in hearing aid users
US20210326099A1 (en) Systems and methods for providing content-specific, personalized audio replay on consumer devices
Kokkinakis et al. Optimized gain functions in ideal time-frequency masks and their application to dereverberation for cochlear implants
Seewald et al. A comparison of manufacturer-specific prescriptive procedures for infants
Aazh et al. Real ear measurement methods for open fit hearing aids: Modified pressure concurrent equalization (MPCE) versus modified pressure stored equalization (MPSE)

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21960470

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 18695220

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE