CN108922616A - A kind of hearing aid is quickly from testing method of completing the square - Google Patents
A kind of hearing aid is quickly from testing method of completing the square Download PDFInfo
- Publication number
- CN108922616A CN108922616A CN201810667221.2A CN201810667221A CN108922616A CN 108922616 A CN108922616 A CN 108922616A CN 201810667221 A CN201810667221 A CN 201810667221A CN 108922616 A CN108922616 A CN 108922616A
- Authority
- CN
- China
- Prior art keywords
- parameter
- population
- probability
- chromosome
- user
- Prior art date
- Legal status (The legal status 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 status listed.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Abstract
The invention discloses a kind of hearing aids quickly from method of completing the square is tested, and step includes:1) initial population of chromosome is calculated:2) fitness of chromosome is calculated;3) judge and export;4) user preference the matching analysis;5) parameter selection;6) population diversity is analyzed;7) parameter is intersected;8) parameter variation;9) parameter is evolved.The present invention is to test with target with band gain, replaces parameter adaptation value with the subjective assessment of patient, and the update of parameter is carried out based on improved Interactive evolution computation.Method is improved:The aid decision-making system that can dynamically update is constructed to improve parameter iteration optimization efficiency;Decision system based on building is simultaneously tested come optimization algorithm with initial value and parameter iteration value in conjunction with patient's audiogram;From parameter coding, fitness selection, parameter cross and variation etc. improved adaptive GA-IAGA performance, the Searching efficiency of algorithm is improved.Experimental result shows, this, which is tested method of completing the square and can improve, listens barrier patient's phonetic recognization rate.
Description
Technical field
The present invention relates to a kind of methods of Audio Signal Processing, quickly test method of completing the square certainly more particularly to a kind of hearing aid.
Background technique
World Health Organization's statistics, the global barrier patient that listens is more than 2.7 hundred million.Dysaudia is to influence the third of human health
Class chronic disease.Moreover, global Aging Problem makes hearing rehabilitation face more stern challenge.Long-time dysaudia is not only
Will affect patient and normally talk ability, understandability and articulation ability, and will lead to patient shrink back, solitarily, it is irascible, seriously
There is mental handicape in person, or even develops to senile dementia, to bring a negative impact to family and society.Currently, wearing hearing aid
Device is one of most effective hearing rehabilitation means.
But even if listen barrier patient using the ratio of hearing aid also only to 1/5 in developed country.And it is limited by medical condition
The health care consciousness of system and the people influence, and the ratio of undeveloped country wants much lower.Interfere with the development middle national hearing aid utilization rate
One of factor is just a lack of the hearing talent of profession.The number of qualified audiologist is all in developing country and developed country
Wretched insufficiency.But the outflow of the hearing talent is so that the hearing aid related service of developing country is in more backward position
It sets, and then constrains the development of hearing aid, also improve the fringe cost of the hearing aid of developing country indirectly.
Aiming at the problem that hearing talent shortage, long-range audiology is considered as a solution.But due to doctor and
Patient is likely located at different countries, therefore brings many problems, such as working qualification, the identification of responsibility, reimbursement and quality control
Deng.In addition, the foundation and maintenance of long-range hearing network still need the personage of profession to assist patient, information is obtained, is led through
Journey.The development speed that these problems have seriously affected U.S.'s hearing tele-medicine is significantly slack-off.It is extensive with smart phone
Infiltration, the technology future can become the access point of the service of offer, and service content includes information/education, screening and possible diagnosis
And intervention.
From design concept, being tested certainly with hearing aid is the optimal selection for solving the problems, such as hearing talent shortage.It is helped from testing to match
Listening the essence of device is the assembly of hearing aid, tests and match, all completed by user using management, and whole process hardly needs professional person
It participates in, does not also need the equipment of profession.Early in 1984,It is mentioned in the patent of Wiener et al. using inside hearing aid
Pure tone generator tonal signal generated measurement user Hearing Threshold, then design a transmission function and meet to generate
Defined hearing aid configuration.This is to test the basic conception with hearing aid certainly, until the equipment that today integrates these design concepts
It is not implemented.The function of user-programmable hearing aid is also relatively easy at present, Main be current environment is judged by patient, then
Artificially changed by configuring switch.If patient is according to environmental difference, controls four kinds of parameters and realize hearing aid allocation optimum, or
Person adjusts algorithm gain, digital noise reduction, microphone modes and spectral enhancement by four independent memories.This method and from
It is identical for testing the design concept matched, and is all to lead to hearing aid output quality decline due to environmental change, thus by user-driven
Carry out the allotment of hearing aid parameter.Except that number of parameters and numerical value that user-programmable hearing aid can be deployed change journey
Degree is fixed, and tests and place one's entire reliance upon the subjective feeling of patient with algorithm certainly, and the range of the variation of parameter is big and actual effect
More preferably, but it is more complicated, realize that difficulty is big.In terms of being tested certainly with parameter more new algorithm, Takagi et al. (2007) is using interaction
Formula evolutionary computation method tentatively realizes the Parametric optimization problem of loudness compensation algorithm;2015, Gauss progress algorithm was also applied to
In hearing aid gain Compensation Research.In addition to this, relevant report is had no both at home and abroad.The research of Takagi et al. is to test certainly with hearing aid
The algorithm research of device provides a feasible thinking, but needs to further investigate there are still many problems, for example tests with effect
Rate is tested with process control etc..
Summary of the invention
It is an object of the invention to overcome deficiency in the prior art, a kind of hearing aid is provided quickly from method of completing the square is tested, is solved
The certainly cumbersome inefficient technical problem of hearing aid fitting in the prior art.
In order to solve the above technical problems, the technical scheme adopted by the invention is that a kind of hearing aid quickly from testing method of completing the square,
Include the following steps:
1) initial population of chromosome is calculated:According to the age of patient, gender, Downtime and audiogram are listened, is calculated and special
Algorithm parameter corresponding to the M sample of immediate (Euclidean distance is minimum) of patient information is as initial population in family's system
Body;
2) fitness of chromosome is calculated:Fitness is divided into 5 grades, and chooses and fits according to the method for hypergeometry operator
Answer angle value, the corresponding genetic probability of each grade isThe value of γ isQ is 0.5;
3) judge and export:If current chromosome is judged as " optimal solution " or reaches maximum number of iterations, terminate
Process;Otherwise optimizing is carried out.When the maximum number of iterations is reached, the solution of output is the optimal solution in iteration history.
4) user preference the matching analysis:The parameter most like with active user's preference is found out, and the parameter is added to now
Have in Advanced group species, and gets rid of the parameter that fitness is worst in current population;
5) parameter selection:The higher individual of fitness in current group according to probability PnOne disk is divided into M parts;So
Afterwards, when being selected, rotating circular disc selects individual i if certain point is fallen in the i-th sector.
6) population diversity is analyzed:Assess the evolution degree of current population, undated parameter crossover probability PcAnd parameter variation
Probability Pm。
7) parameter is intersected:With probability PcTwo individual chromosome dyads are exchanged, two new individuals are obtained.
8) parameter variation:By each genic value in each individual with probability PmIt makes a variation, mutation probability formula isWherein, PmmaxFor the mutation probability upper limit.
9) parameter is evolved:By selection, intersection and mutation operation, a new population is obtained.Above-mentioned steps are by given
Cycle-index after, optimization process terminate.
Further, the user preference the matching analysis includes the following steps:
1) allele unit is calculatedFitnessxi={ gi1,gi2...giNIt is population
In i-th of parameter, N is number of parameters, gikIndicate k-th of gain parameter of i-th of chromosome in population,Indicate population
In i-th of chromosome k-th of gain parameter jth position,Patient is represented to current chromosome xiEvaluation of estimate;
2) confidence level function is calculatedSimulate the credibility of people's subjective assessment, a is
Confidence level coefficient, NsFor parameter threshold;
3) adjustment user is to equipotential genetic unitPreference
4) current patents u is calculated1With the user u in knowledge base2To genetic unit gkPreference similarity
To calculate user u1With match user u2Similarity
5) according to σ (u1,u2) select and user u1The higher matching user u of similarity2。
Further, population diversity analysis includes the following steps:
1) diversity of population is calculatedL is the digit of each parameter,It is k-th
Channel i-th bit gene takes 1 probability, and value is
2) when diversity d is greater than threshold value D, crossover probability increases by 5%, the upper limit 0.95;Mutation probability increases by 0.5%,
The upper limit is 0.3;Otherwise, crossover probability and mutation probability remain unchanged.
Compared with prior art, the beneficial effects obtained by the present invention are as follows being:1, building is based on interactive calculating and expert system
The hearing aid fitting model of system, improves efficiency of algorithm;2, based on expert system to user preference matching and population diversity
It is analyzed, to optimize initial parameter and update iteration median, accelerates algorithm the convergence speed.
In conclusion hearing aid of the invention is quickly fast with convergence from method of completing the square is tested, the advantage good with effect is tested.It has
There is above-mentioned many and practical value, and there are no similar design in congenic method and publish or use and really belong to
Innovation has biggish improvement, technically has large improvement, there is the extensive utility value of industry, really for it is one novel, into
Step, practical new design.
Detailed description of the invention
Fig. 1 is of the present invention adaptive confirmed with general frame figure.
Fig. 2 is optimization algorithm flow chart of the present invention.
Fig. 3 is hearing aid expert system structure figure of the present invention.
Fig. 4 is phonetic recognization rate of the present invention comparison.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.Following embodiment is only used for clearly illustrating the present invention
Technical solution, and not intended to limit the protection scope of the present invention.
As shown in Figure 1, being tested certainly with hearing aid whole system includes three elements:It hearing aid test system, patient and tests certainly
Match system.Groundwork process is:Hearing aid test system handles input sound according to algorithm parameter, and exports treated sound
Sound is to user;User receives the sound of hearing aid output and evaluates according to its subjective criterion;The result of assessment is fed back to certainly
It tests match system and carries out optimizing and revising for parameter, and return to hearing aid test system.Wherein, match system is tested certainly by three part groups
At:Expert system, evolution module and user interface.
User interface is supplied to user's man-machine interaction mode, and the numerically sound after assessment processing.Evaluation content
It is evaluated comprising hearing evaluation and sound.1. hearing evaluation:It is quasi- to be considered as simplified pure tone threshold testing to complete to patient's
Damage is listened to assess.(a) basic test method is identical as general pure tone test, surveyed pure tone be divided into 11 frequency points 125,250,
500,750,1000,1500,2000,3000,4000,6000,8000Hz;(b) according to the grade of evaluation, dynamic is adjusted in proportion
The dB value of rank;(c) threshold value is listened at adjacent test frequency, is simulated using the method for curve matching.By listening damage to patient
It comments roughly, can be matched with the patient library in knowledge base, partially use for reference the parameter information of other patients, improve system effectiveness and can
By property;2. sound is evaluated:Sound after evaluation optimization, and carry out feedback allotment parameter.Evaluation criterion is from difference to being divided into 5 grades of marks well
It is quasi-.
Evolution module is the adaptive confirmed nucleus module matched.Assessment of the evolution module based on hearing aid user to output sound
As a result, updating algorithm parameter.Parameter is provided with the gain of different frequency range, and hearing aid, which forms new voice according to the parameter of update, to be believed
Number, it is evaluated and tested again to user.
The effect of expert system is to analyze the preference of user, calculates the diversity of population, so as to form corresponding " knowledge "
Evolution module is fed back to, to achieve the purpose that reduce human fatigue degree.According to the initial testing of patient, to evolution module and to help
It listens device parameter to be initialized or limited, to improve optimization efficiency, increases reliability;Test match during, with evolution module into
Row interaction, acceleration parameter convergence.
As shown in Fig. 2, the workflow of optimization algorithm is:
1) chromosome population is initialized:The data of solution space are subjected to binary coding, show as the gene in hereditary space
The structured data (i.e. chromosome) of type.According to the age of patient, gender, Downtime and audiogram are listened, calculate and is suffered from knowledge base
Algorithm parameter corresponding to the M sample of immediate (Euclidean distance is minimum) of person's information is as initial population.Algorithm by 0~
The voice signal of 8kHz is divided into 11 frequency band, respectively corresponds 11 frequency points of audiogram.Algorithm needs to mend by each frequency band
Yield value is repaid as gene, and is divided into four groups according to the inflection point of input-output curve, i.e., every group of parameter there are 44.All parameter groups
Parameter chromosome is formed at an array.
2) fitness of chromosome is calculated:Fitness is divided into 5 grades, and chooses and fits according to the method for hypergeometry operator
Answer angle value, the corresponding genetic probability of each grade isThe value of γ isQ is 0.5;
3) judge and export:If current chromosome is judged as " optimal solution " or reaches maximum number of iterations, terminate
Process;Otherwise optimizing is carried out.When the maximum number of iterations is reached, the solution of output is the optimal solution in iteration history.
4) user preference the matching analysis:The parameter most like with active user's preference is found out, and the parameter is added to now
Have in Advanced group species, and gets rid of the parameter that fitness is worst in current population.Calculating step is:
(1) allele unit is calculatedFitnessxi={ gi1,gi2...giNIt is population
In i-th of parameter, N is number of parameters, gikIndicate k-th of gain parameter of i-th of chromosome in population,It indicates in population
The jth position of k-th of gain parameter of i-th of chromosome,Patient is represented to current chromosome xiEvaluation of estimate;
(2) confidence level function is calculatedTo simulate the credibility of people's subjective assessment, a
For confidence level coefficient, NsFor parameter threshold;
(3) adjustment user is to equipotential genetic unitPreference
(4) current patents u is calculated1With the user u in knowledge base2To genetic unit gkPreference similarity
To calculate user u1With match user u2Similarity
(5) according to σ (u1,u2) select and user u1The higher matching user u of similarity2。
5) parameter selection:The higher individual of fitness in current group according to probability PnOne disk is divided into M parts;So
Afterwards, when being selected, rotating circular disc selects individual i if certain point is fallen in the i-th sector.
6) population diversity is analyzed:Assess the evolution degree of current population, undated parameter crossover probability PcAnd parameter variation
Probability Pm.Calculating step is:
(1) diversity of population is calculatedL is the digit of each parameter,It is k-th
Channel i-th bit gene takes 1 probability, and value is
(2) when diversity d is greater than threshold value D, crossover probability increases by 5%, the upper limit 0.95;Mutation probability increases by 0.5%,
The upper limit is 0.3;Otherwise, crossover probability and mutation probability remain unchanged.
7) parameter is intersected:With probability PcTwo individual chromosome dyads are exchanged, two new individuals are obtained.
8) parameter variation:By each genic value in each individual with probability PmIt makes a variation, mutation probability formula isWherein, PmmaxFor the mutation probability upper limit.
9) parameter is evolved:By selection, intersection and mutation operation, a new population is obtained.Above-mentioned steps are by given
Cycle-index after, optimization process terminate.
As shown in figure 3, the effect of expert system is to accelerate convergence, step-size in search is reduced, the accuracy of search is improved.In addition to
According to personal patient information, improve outside the selection of initial value, expert system is also able to achieve user preference matching and population diversity
Analysis.User preference matching is found in knowledge base while analyze user preference and currently using existing knowledge base
The most like chromosome of user preference replaces the chromosome that fitness is worst in current population with the chromosome.It is each in knowledge base
Chromosome is that the chromosome for testing and matching and finally searching is completed in different user.The purpose of population diversity analysis is to prevent precocity existing
As the every progress generation search of algorithm will carry out diversity analysis, improve intersection, variation if current population diversity is too low
Probability increases step-size in search, to enrich current species information for further evolutional operation.
Effect of the invention can be further illustrated by experiment.
Speech understanding degree test result as shown in Figure 4 it is found that in terms of tone testing, propose method test with effect compared with
Good, average recognition rate reaches 78.3%, improves 11.1% compared to Interactive evolution computation (IEC), mentions compared to traditional algorithm
It is high by 12.3%.Wherein, the discrimination highest of patient S8, it is minimum to reach 87.6%, S1 discrimination, reaches 67.0%.Comparison tradition
Algorithm and Interactive evolution computation, average recognition rate are not significantly improved.It can be seen that S2, S4 and S7 it is interactive into
The performance for changing algorithm is also less than tradition and tests with algorithm.Wherein, S4 reduction by 11.6%, it is little so as to cause two kinds of algorithm difference.
In conclusion a kind of hearing aid of the present invention is quickly to test with target from method of completing the square is tested with band gain, with patient's
Subjective assessment replaces parameter adaptation value, and the update of parameter is carried out based on improved Interactive evolution computation.Method is improved:
1) the building aid decision-making system that can dynamically update improves parameter iteration optimization efficiency;2) decision system and knot based on building
Conjunction patient's audiogram carrys out optimization algorithm and tests with initial value and parameter iteration value;3) intersect from parameter coding, fitness selection, parameter
Variation etc. improved adaptive GA-IAGA performance, improves the Searching efficiency of algorithm.Experimental result shows, this, which is tested method of completing the square and can improve, listens
Hinder patient's phonetic recognization rate.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformations can also be made, these improvement and deformations
Also it should be regarded as protection scope of the present invention.
Claims (3)
1. a kind of hearing aid quickly from method of completing the square is tested, includes the following steps:
1) initial population of chromosome is calculated:According to the age of patient, gender, Downtime and audiogram are listened, is calculated and expert system
Algorithm parameter corresponding to the immediate M sample of patient information is as initial population in system;
2) fitness of chromosome is calculated:Fitness is divided into 5 grades, and chooses fitness according to the method for hypergeometry operator
Value, the corresponding genetic probability of each grade areThe value of γ isQ is 0.5;
3) judge and export:If current chromosome is judged as " optimal solution " or reaches maximum number of iterations, terminate process;
Otherwise optimizing is carried out;When the maximum number of iterations is reached, the solution of output is the optimal solution in iteration history;
4) user preference the matching analysis:Find out the parameter most like with active user's preference, and by the parameter be added to it is existing into
Change in population, and removes the parameter that fitness is worst in current population;
5) parameter selection:The higher individual of fitness in current group according to probability PnOne disk is divided into M parts;Then, exist
When being selected, rotating circular disc selects individual i if certain point is fallen in the i-th sector;
6) population diversity is analyzed:Assess the evolution degree of current population, undated parameter crossover probability PcWith parameter variation probability
Pm;
7) parameter is intersected:With probability PcTwo individual chromosome dyads are exchanged, two new individuals are obtained;
8) parameter variation:By each genic value in each individual with probability PmIt makes a variation, mutation probability is
Wherein, PmmaxFor the mutation probability upper limit;
9) parameter is evolved:By selection, intersection and mutation operation, a new population is obtained;Above-mentioned steps are followed by given
After ring number, optimization process is terminated.
2. a kind of hearing aid according to claim 1 quickly tests method of completing the square certainly, which is characterized in that the user preference
Include the following steps with analysis:
1) allele unit is calculatedFitnessxi={ gi1,gi2...giNIt is in population
I-th of parameter, N are number of parameters, gikIndicate k-th of gain parameter of i-th of chromosome in population,It indicates i-th in population
The jth position of k-th of gain parameter of a chromosome,Patient is represented to current chromosome xiEvaluation of estimate;
2) confidence level function is calculatedThe credibility of people's subjective assessment is simulated, a is confidence level
Coefficient, NsFor parameter threshold;
3) adjustment user is to equipotential genetic unitPreference
4) current patents u is calculated1With the user u in knowledge base2To genetic unit gkPreference similarity
To calculate user u1With match user u2Similarity
5) according to σ (u1,u2) select and user u1The higher matching user u of similarity2。
3. a kind of hearing aid according to claim 1 quickly tests method of completing the square certainly, which is characterized in that the population diversity
Analysis includes the following steps:
1) diversity of population is calculatedL is the digit of each parameter,For k-th of channel
I-th bit gene takes 1 probability, and value is
2) when diversity d is greater than threshold value D, crossover probability increases by 5%, the upper limit 0.95;Mutation probability increases by 0.5%, the upper limit
It is 0.3;Otherwise, crossover probability and mutation probability remain unchanged.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810667221.2A CN108922616A (en) | 2018-06-26 | 2018-06-26 | A kind of hearing aid is quickly from testing method of completing the square |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810667221.2A CN108922616A (en) | 2018-06-26 | 2018-06-26 | A kind of hearing aid is quickly from testing method of completing the square |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108922616A true CN108922616A (en) | 2018-11-30 |
Family
ID=64423093
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810667221.2A Withdrawn CN108922616A (en) | 2018-06-26 | 2018-06-26 | A kind of hearing aid is quickly from testing method of completing the square |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108922616A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114339564A (en) * | 2021-12-23 | 2022-04-12 | 清华大学深圳国际研究生院 | User self-adaptive hearing aid self-fitting method based on neural network |
CN114938487A (en) * | 2022-05-13 | 2022-08-23 | 东南大学 | Hearing aid self-fitting method based on sound scene discrimination |
-
2018
- 2018-06-26 CN CN201810667221.2A patent/CN108922616A/en not_active Withdrawn
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114339564A (en) * | 2021-12-23 | 2022-04-12 | 清华大学深圳国际研究生院 | User self-adaptive hearing aid self-fitting method based on neural network |
CN114938487A (en) * | 2022-05-13 | 2022-08-23 | 东南大学 | Hearing aid self-fitting method based on sound scene discrimination |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109859767B (en) | Environment self-adaptive neural network noise reduction method, system and storage medium for digital hearing aid | |
Goehring et al. | Using recurrent neural networks to improve the perception of speech in non-stationary noise by people with cochlear implants | |
CN101933085B (en) | Objective measurement of audio quality | |
CN106531190A (en) | Speech quality evaluation method and device | |
CN101716108B (en) | System for processing electronic cochlear signal and coding electric stimulation | |
US9319812B2 (en) | System and methods of subject classification based on assessed hearing capabilities | |
CN106063295A (en) | A method of fitting a hearing aid system and a hearing aid fitting system | |
CN106303874B (en) | A kind of adaptive confirmed method of completing the square of digital deaf-aid | |
Alexandre et al. | Feature selection for sound classification in hearing aids through restricted search driven by genetic algorithms | |
CN112653980B (en) | Interactive self-checking and matching method for intelligent hearing aid | |
CN111951824A (en) | Detection method for distinguishing depression based on sound | |
CN109151692A (en) | Hearing aid based on deep learning network tests method of completing the square certainly | |
CN108922616A (en) | A kind of hearing aid is quickly from testing method of completing the square | |
WO2020024807A1 (en) | Artificial cochlea ambient sound sensing method and system | |
Berisha et al. | Modeling pathological speech perception from data with similarity labels | |
Hüwel et al. | Hearing aid research data set for acoustic environment recognition | |
CN111968677A (en) | Voice quality self-evaluation method for fitting-free hearing aid | |
Shoba et al. | Improving speech intelligibility in monaural segregation system by fusing voiced and unvoiced speech segments | |
CN111554325B (en) | Voice recognition method and system | |
Gonzalez et al. | On batching variable size inputs for training end-to-end speech enhancement systems | |
CN108919962B (en) | Auxiliary piano training method based on brain-computer data centralized processing | |
Lopez et al. | A clinical test battery for Better hEAring Rehabilitation (BEAR): Towards the prediction of individual auditory deficits and hearing-aid benefit | |
CN114938487A (en) | Hearing aid self-fitting method based on sound scene discrimination | |
Mogridge et al. | Non-Intrusive Speech Intelligibility Prediction for Hearing-Impaired Users using Intermediate ASR Features and Human Memory Models | |
Vecchi et al. | Hearing-impaired sound perception: What can we learn from a biophysical model of the human auditory periphery |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181130 |