CN109246515B - A kind of intelligent earphone and method promoting personalized sound quality function - Google Patents
A kind of intelligent earphone and method promoting personalized sound quality function Download PDFInfo
- Publication number
- CN109246515B CN109246515B CN201811171346.2A CN201811171346A CN109246515B CN 109246515 B CN109246515 B CN 109246515B CN 201811171346 A CN201811171346 A CN 201811171346A CN 109246515 B CN109246515 B CN 109246515B
- Authority
- CN
- China
- Prior art keywords
- sound
- sound quality
- parameter sets
- user
- hearing
- 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R1/00—Details of transducers, loudspeakers or microphones
- H04R1/10—Earpieces; Attachments therefor ; Earphones; Monophonic headphones
- H04R1/1083—Reduction of ambient noise
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04R—LOUDSPEAKERS, MICROPHONES, GRAMOPHONE PICK-UPS OR LIKE ACOUSTIC ELECTROMECHANICAL TRANSDUCERS; DEAF-AID SETS; PUBLIC ADDRESS SYSTEMS
- H04R2201/00—Details of transducers, loudspeakers or microphones covered by H04R1/00 but not provided for in any of its subgroups
- H04R2201/10—Details of earpieces, attachments therefor, earphones or monophonic headphones covered by H04R1/10 but not provided for in any of its subgroups
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Acoustics & Sound (AREA)
- Signal Processing (AREA)
- Telephone Function (AREA)
Abstract
The invention discloses intelligent earphones and method that one kind can promote personalized sound quality function, a kind of intelligent earphone promoting personalized sound quality function, further include hearing compensation module, sound quality beautification module and memory module including sound output module: the memory module is used to store the hearing compensation parameter sets and sound quality beautification parameter sets of the user individual obtained according to the hearing test situation of user;The hearing compensation module carries out hearing compensation according to the hearing compensation parameter sets of the user individual sound to be output to earphone;The sound quality beautification module beautifies parameter sets according to the sound quality of user individual and carries out sound quality beautification to compensated sound to be output;Sound by sound quality beautification is played by sound output module.Clarity, naturalness and the comfort level that output sound can be improved, make user obtain good audio experience.
Description
Technical field
The present invention relates to intelligent earphones and method that one kind can promote personalized sound quality function.
Background technique
China has crowd's ratio of different degrees of dysaudia very high, even if in the normal crowd of hearing, everyone
Hearing characteristics be also to have nothing in common with each other.Hearing aid is the special earphone of the most common one kind of dysaudia personage, be may be implemented
Personalized hearing compensation function, but it tests method of completing the square complexity, needs the dispenser of profession to operate and completes, limits many patients
Use.Moreover, hearing aid is only used for hearing correction, the general equipment such as mobile phone connection of getting along well carries out music or voice is logical
Letter.
Common headphones cooperate mobile phone or other playback equipments, may be implemented to make a phone call, play the functions such as music, but cannot
For user personalized hearing characteristics carry out acoustic processing, cause dysaudia personage be unable to normal use earphone make a phone call or
It listens to music, and Subjects With Normal Hearing scholar can not obtain best sound effect.For example, patent No. CN200710168143.2 is disclosed
A kind of earphone with hearing-aid function, patent No. CN201510173183.1 disclose a kind of hearing assistance system and hearing aid earphone,
The two discloses the hearing aid earphone based on mobile phone, but this two patents only highlight mobile phone with hearing compensation,
Telephone relation function does not relate to it and carries out testing method of completing the square and sound quality method for improving automatically necessary to personalized compensation.
Summary of the invention
In view of the above-mentioned problems, the present invention provides a kind of intelligent earphone and method that can promote personalized sound quality function, not only
Suitable for dysaudia personage, it is also applied for the normal person with different hearing characteristics, user, which can independently be tested, to be matched, and is not required to
It wants the dispenser of profession to help, clarity, naturalness and the comfort level of output sound can be improved, user is made to obtain good listen
Feel experience.
To realize above-mentioned technical purpose and the technique effect, the invention is realized by the following technical scheme:
A kind of intelligent earphone promoting personalized sound quality function, including sound output module, further include hearing compensation mould
Block, sound quality beautification module and memory module:
The memory module is used to store the hearing compensation of the user individual obtained according to the hearing test situation of user
Parameter sets and sound quality beautify parameter sets;
The hearing compensation module according to the hearing compensation parameter sets of the user individual sound to be output to earphone into
Row hearing compensation;
The sound quality beautification module beautifies parameter sets to compensated sound to be output according to the sound quality of user individual
Sound carries out sound quality beautification;
Sound by sound quality beautification is played by sound output module.
It is preferred that further including voice pickup module:
The voice pickup module is for picking up external sound and by the sound of pickup through hearing compensation module and sound quality beauty
After changing resume module, played by sound output module;
Alternatively, the speech concurrent that the voice pickup module is used to pick up user gives the intelligent terminal of connection.
It is preferred that the intelligent earphone is wireless headset, the wireless headset further includes bluetooth communication, and the bluetooth is logical
Letter module is connected with voice pickup module and hearing compensation module respectively.
It is preferred that the intelligent earphone includes three kinds of operating modes: hearing aid mode, music playing mode and call mode,
Wherein:
Hearing aid mode: bluetooth communication is closed, and voice pickup module picks up external sound, through hearing compensation module and
After sound quality beautifies resume module, played by the sound output module of earphone;
Music playing mode: voice pickup module is closed, and bluetooth communication receives the music that intelligent terminal plays, through listening
After force compensating module and sound quality beautification resume module, played by the sound output module of earphone;
Call mode: the voice of intelligent terminal transmission is received by bluetooth communication, through hearing compensation module and sound quality
After beautifying resume module, played by the sound output module of earphone, meanwhile, voice pickup module picks up voice and passes through bluetooth
Communication module is sent to intelligent terminal.
A method of personalized sound quality function is promoted based on intelligent earphone, is included the following steps:
Step 1, the hearing compensation parameter sets for obtaining user individual and sound quality beautify parameter sets;
Step 2 carries out hearing benefit to the sound that intelligent earphone exports according to the hearing compensation parameter sets of user individual
It repays;
Step 3 beautifies parameter sets according to the sound quality of user individual to intelligent earphone compensated sound progress sound quality
Beautification;
Step 4, the sound by sound quality beautification are played by the output module of intelligent earphone.
It is preferred that obtaining user personality by audiometery, speech hearing test and noise perception test in step 1
The hearing compensation parameter sets and sound quality of change beautify parameter sets.
It is preferred that audiometery includes the following steps:
Step 101 generates the different tonal signals to sound intensity degree by intelligent terminal, and passes through the Intellgent Ear machine sowing of connection
It puts;
Step 102 feeds back the tonal signal that whether can hear broadcasting by user on intelligent terminal, obtains user's
Pure tone audiometry curve;It is fed back on intelligent terminal by user whether because uncomfortable too loudly, obtains not relaxing for user
Suitable threshold curve;
Step 103, intelligent terminal generate user according to obtained pure tone audiometry curve, thresholds of uncomfortable level curve and prescriptive formula
Hearing compensation parameter sets initial value.
It is preferred that speech hearing test includes the following steps:
Step 201, intelligent terminal use test according to the hearing compensation parameter sets initial value that audiometery obtains
Monosyllabic predicative sound signal carry out hearing compensation, compensated first monosyllable by intelligent earphone broadcasting;
Step 202 feeds back heard monosyllable by user on intelligent terminal, obtains user to the monosyllable
Currently to the identification record under sound intensity degree;
Step 203 calculates user to the speech discrimination score for currently testing the consonant and vowel that are included with monosyllable, meter
Calculating formula isWherein, RijFor speech discrimination score, i=1,2 ..., 32, j
=1,2 ..., J, J are the word number currently tested, wherein the speech discrimination score of the consonant and vowel do not tested keeps original
Value;
Step 204 calculates the monosyllabic FFT spectrum { X of current testk, k=1,2 ..., K }, K is FFT frequency point number;
Step 205, by current speech discrimination score { Rij, i=1,2 ..., 32 } and FFT spectrum { Xk, k=1,2 ..., N }
The first depth convolutional neural networks are inputted, update is iterated, export hearing compensation parameter sets, including { multichannel loudness is mended
Repay each channel inflection point Pij;Each channel compressions ratio CRij, i=1,2 ..., M;J=1,2 ..., N }, wherein N is point in each channel
Number of segment, M indicate port number;
Step 206, intelligent terminal are according to the updated hearing compensation parameter sets of iteration to next test single syllable
Word compensates, compensated monosyllabic by intelligent earphone broadcasting, and enters step 202, until single syllable vocabulary is tested
It finishes, obtains optimizing hearing compensation parameter sets.
It is preferred that noise perception test includes the following steps:
Step 301, intelligent terminal play the dialogic voice under different noise scenarios and are played by intelligent earphone;
Step 302, for the dialogic voice under variant noise scenarios, intelligent terminal proposes that one group of guided bone is asked to user
Topic, user feed back the above problem in intelligent terminal, calculate satisfaction scoring according to feedback result;
Step 303 is higher than threshold value T when satisfaction scores, and output optimizes sound quality and beautifies parameter sets, exits noise perception
Test, otherwise, enters step 304;
Step 304, feed back input the second depth convolutional neural networks by above-mentioned leading question, are iterated update,
It exports sound quality and beautifies parameter sets, including { each subband noise reduction parameters NRl, l=1,2 ..., L, each subband parametric equalizer EQl, l
=1,2 ..., L, each subband chauvent's criterion parameter HWl, l=1,2 ..., L }, L is the sub-band number of sound quality beautifying functions, into step
Rapid 305;
Step 305, intelligent terminal beautify parameter sets according to updated sound quality and generate new test dialogic voice, lead to
Intelligent earphone broadcasting is crossed, and enters step 302.
It is preferred that in step 101, difference is generated from 125Hz to 8000Hz in range to the pure of sound intensity degree by intelligent terminal
Sound signal, it be 8, N value is 3 that L, which takes 32, K to take 256, M value, and the intelligent terminal is mobile phone;
The network structure of first depth convolutional neural networks and the second depth convolutional neural networks is all by 3 convolutional Neural nets
Network module composition, each module include a convolutional layer, a pond layer and a discarding layer, and the output of three modules passes through flat
Leveling is by multidimensional characteristic one-dimensional, then by cascading layers by multiple convolutional layer characteristic aggregations, finally by two layers of full articulamentum
Export final argument.
The beneficial effects of the present invention are:
(1) it not only includes the parameter sets for hearing compensation module that the present invention, which tests the parameter matched and obtained, further includes being used for
Sound quality beautifies the parameter sets of module, both can be adapted for dysaudia personage, improves lalognogis degree, is readily applicable to listen
The normal personage of power improves the clarity and comfort level of sound, and versatility is stronger, applied widely;
(2) it is tested by audiometery, speech hearing test and the noise perception to user, knows user individual
Hearing characteristics, and promoted for hearing compensation and sound quality, user can be made to obtain optimal sound experience, specific aim is more
By force, experience effect is more preferable;
(3) by independently testing method of completing the square based on depth convolutional neural networks, simplification is tested with process, and user is existed
It independently tests and matches in family, without the help of professional dispenser, convenient for promoting.
Detailed description of the invention
Fig. 1 is the structural block diagram for the intelligent earphone that one kind of the present invention can promote personalized sound quality function;
Fig. 2 is the network structure of the first depth convolutional neural networks and the second depth convolutional neural networks of the invention.
Specific embodiment
Technical solution of the present invention is described in further detail with specific embodiment with reference to the accompanying drawing, so that ability
The technical staff in domain can better understand the present invention and can be practiced, but illustrated embodiment is not as to limit of the invention
It is fixed.
A kind of intelligent earphone promoting personalized sound quality function, as shown in Figure 1, including sound output module, hearing benefit
Module, sound quality beautification module and memory module, each module is repaid to be described below:
The memory module is used to store the hearing compensation of the user individual obtained according to the hearing test situation of user
Parameter sets and sound quality beautify parameter sets.Before the intelligent earphone works normally, user is autonomous by smart phone Terminal Type
Progress hearing, which is tested, matches, and measures hearing loss.The hearing loss of user is tested, the present invention preferably passes through audiometery, speech
Hearing test and noise perception test obtain the hearing compensation parameter sets of user individual and sound quality beautifies parameter sets.
The hearing compensation module according to the hearing compensation parameter sets of the user individual sound to be output to earphone into
Row hearing compensation;
The sound quality beautification module beautifies parameter sets to compensated sound to be output according to the sound quality of user individual
Sound carries out sound quality beautification;
Sound by sound quality beautification is played by sound output module.
It is preferred that further including voice pickup module, voice pickup module is for picking up external sound, in different modes,
It is different that information transmits object: when hearing aid mode, voice pickup module is for picking up external sound and by the sound of pickup through listening
After force compensating module and sound quality beautification resume module, played by sound output module;When call mode, voice pickup module is used
The intelligent terminal (intelligent terminal that earphone can be connected) of connection is given in the speech concurrent for picking up user, in general, intelligent terminal
The mostly mobile terminals such as mobile phone, pad.
Intelligent earphone can select wired earphone, can also select wireless headset, a kind of wireless headset involved in Fig. 1, nothing
Line earphone embedded with bluetooth communication module, the bluetooth communication are connected with voice pickup module and hearing compensation module respectively,
Wireless headset carries out information exchange by bluetooth communication and intelligent terminal, i.e. wireless headset will be picked up by bluetooth communication
The sound taken is sent to intelligent terminal, and the sound of broadcasting is sent to hearing compensation mould by bluetooth communication by intelligent terminal
Block carries out subsequent processing.Optimization hearing compensation parameter sets and sound quality the beautification parameter sets of acquisition can be logical by bluetooth
Letter module is stored in the memory module of intelligent earphone, the personalizing parameters as hearing compensation module and sound quality beautification module.
The settable multiple-working mode of intelligent earphone of the invention, most common includes three kinds of operating modes: hearing aid mould
Formula, music playing mode and call mode, now with wireless headset for example:
Hearing aid mode: bluetooth communication is closed, and voice pickup module picks up external sound, through hearing compensation module and
After sound quality beautifies resume module, played by the sound output module of earphone.
Music playing mode: voice pickup module is closed, and bluetooth communication receives the music that intelligent terminal plays, through listening
After force compensating module and sound quality beautification resume module, played by the sound output module of earphone.
Call mode: receiving the far-end speech of intelligent terminal transmission by bluetooth communication, through hearing compensation module and
After sound quality beautifies resume module, played by the sound output module of earphone, meanwhile, voice pickup module picks up near-end speech simultaneously
Intelligent terminal is sent to by bluetooth communication.
It, can since the output sound of these three operating modes all have passed through personalized hearing compensation and sound quality beautification
To effectively improve the clarity, naturalness and comfort level of sound.Corresponding, one kind promoting personalized sound quality function based on intelligent earphone
The method of energy, includes the following steps:
Step 1, the hearing compensation parameter sets for obtaining user individual and sound quality beautify parameter sets;
Step 2 carries out hearing benefit to the sound that intelligent earphone exports according to the hearing compensation parameter sets of user individual
It repays;
Step 3 beautifies parameter sets according to the sound quality of user individual to intelligent earphone compensated sound progress sound quality
Beautification;
Step 4, the sound by sound quality beautification are played by the output module of intelligent earphone.
It is preferred that obtaining user personality by audiometery, speech hearing test and noise perception test in step 1
The hearing compensation parameter sets and sound quality of change beautify parameter sets, and each testing procedure is described as follows.
It is preferred that audiometery includes the following steps:
Step 101 generates the different tonal signals to sound intensity degree by intelligent terminal, and passes through the Intellgent Ear machine sowing of connection
It puts, for example, generating from 125Hz to 8000Hz tonal signal of the difference to sound intensity degree in range by mobile phone.
Step 102 feeds back the tonal signal that whether can hear broadcasting by user on intelligent terminal, obtains user's
Pure tone audiometry curve;It is fed back on intelligent terminal by user whether because uncomfortable too loudly, obtains not relaxing for user
Suitable threshold curve.
Step 103, intelligent terminal generate user according to obtained pure tone audiometry curve, thresholds of uncomfortable level curve and prescriptive formula
Hearing compensation parameter sets initial value, without loss of generality, NAL-NL2 may be selected in prescriptive formula.
It is preferred that speech hearing test includes the following steps:
Step 201, intelligent terminal use test according to the hearing compensation parameter sets initial value that audiometery obtains
Monosyllabic predicative sound signal carry out hearing compensation, compensated first monosyllable by intelligent earphone broadcasting.
Step 202 feeds back heard monosyllable by user on intelligent terminal, obtains user to the monosyllable
It is correctly 1 currently to the identification record under sound intensity degree, mistake 0.
Step 203 calculates user to the speech discrimination score for currently testing the consonant and vowel that are included with monosyllable, meter
Calculating formula isWherein, RijFor speech discrimination score, i=1,2 ..., 32, be
Refer to 22 consonants of mandarin (/b/ ,/p/ ,/m/ ,/f/ ,/z/ ,/c/ ,/s/ ,/d/ ,/t/ ,/n/ ,/l/ ,/zh/ ,/ch/ ,/sh/ ,/
R/ ,/ng/ ,/g/ ,/k/ ,/h/ ,/j/ ,/q/ ,/x/) and 10 vowels (/a/ ,/o/ ,/e/ ,/i/ ,/u/ ,/ü/,/ê/, before the tip of the tongue
Vowel/- i/, the tip of the tongue back vowel/- i/ ,/er/), j=1,2 ..., J, J are the word number currently tested, wherein are not tested
The speech discrimination score of consonant and vowel keeps original value.
Step 204 calculates the monosyllabic FFT spectrum { X of current testk, k=1,2 ..., K }, K is FFT (quick Fu
Family name's transformation) frequency point number, without loss of generality, can use K is 256.
Step 205, by current speech discrimination score { Rij, i=1,2 ..., 32 } and FFT spectrum { Xk, k=1,2 ..., N }
The first depth convolutional neural networks are inputted, update is iterated, export hearing compensation parameter sets, including { multichannel loudness is mended
Repay each channel inflection point Pij;Each channel compressions ratio CRij, i=1,2 ..., M;J=1,2 ..., N }, wherein N is point in each channel
Number of segment, without loss of generality, N can value be 3, M indicate port number, without loss of generality, M can value be 8.
Step 206, intelligent terminal are according to the updated hearing compensation parameter sets of iteration to next test single syllable
Word compensates, compensated monosyllabic by intelligent earphone broadcasting, and enters step 202, until single syllable vocabulary is tested
It finishes, obtains optimizing hearing compensation parameter sets.
It is preferred that noise perception test includes the following steps:
Step 301, intelligent terminal play the dialogic voice under different noise scenarios and are played by intelligent earphone, play not
With the dialogic voice under noise scenarios, noise scenarios include but is not limited to: downtown area scene, Driving Scene, meeting room scene, family
Scene, airport scene etc..
Step 302, for the dialogic voice under variant noise scenarios, intelligent terminal proposes that one group of guided bone is asked to user
Topic, comprising:
User to high-frequency noise, mid-frequency noise, low-frequency noise acceptance level (three-level score, respectively represent it is noisy, can connect
By, without influencing);To downtown area scene, Driving Scene, meeting room scene, home scenarios, airport scene the dialogue degree of understanding
(every kind of scene three-level scoring, respectively represent major part and do not hear, can almost not hear, not hear completely);Whether sound timbre is ear-piercing
(three-level scoring, respectively represent ear-piercing, sometimes ear-piercing, not ear-piercing);Whether sound timbre is stuffy, and (three-level scoring, respectively represents hair
It is bored, sometimes stuffy, not stuffy);Whether sound timbre is empty (three-level scoring, respectively represent empty, sometimes empty, not empty);
Sound timbre it is whether fuzzy (three-level scoring, respectively represent it is fuzzy, obscure, do not obscure sometimes);Whether sound, which utters long and high-pitched sounds, (is maked a whistling sound always
It cries, utter long and high-pitched sounds once in a while, nothing is uttered long and high-pitched sounds).
User feeds back the above problem in intelligent terminal, calculates satisfaction scoring according to feedback result, score formula
Following formula can be used:Wherein siFor the score of above-mentioned each problem, S is total score best result of all the problems.
Step 303 is higher than threshold value T when satisfaction scores, and output optimizes sound quality and beautifies parameter sets, exits noise perception
Test, otherwise, enters step 304, threshold value T can be arranged according to the actual situation, such as be set as the 75% of full marks.
Step 304, feed back input the second depth convolutional neural networks by above-mentioned leading question, are iterated update,
It exports sound quality and beautifies parameter sets, including { each subband noise reduction parameters NRl, l=1,2 ..., L, each subband parametric equalizer EQl, l
=1,2 ..., L, each subband chauvent's criterion parameter HWl, l=1,2 ..., L }, L is the sub-band number of sound quality beautifying functions, does not lose one
As property, L desirable 32 enters step 305.
Step 305, intelligent terminal beautify parameter sets according to updated sound quality and generate new test dialogic voice, lead to
Intelligent earphone broadcasting is crossed, and enters step 302, user is tested again and scores the satisfaction of leading question, until scoring is high
The upper limit is reached in threshold value T or testing time, without loss of generality, which may be configured as 3~6 times.
Wherein, the first depth convolutional neural networks and the second depth convolutional neural networks used by smart phone Terminal Type
Network structure as shown in Fig. 2, being all made of 3 convolutional neural networks modules Block1, Block2 and Block3, each module
Including a convolutional layer (CONV1), a pond layer (MaxPooling) and a discarding layer (Dropout), three modules
Output by multidimensional characteristic one-dimensional, then passes through cascading layers (Concatenate) for multiple convolution by flat layer (Flatten)
Layer characteristic aggregation exports final argument finally by two layers of full articulamentum (Dense).After the completion of above-mentioned test, user is obtained
Optimization hearing compensation parameter sets and sound quality beautify parameter sets.
Intelligent earphone of the invention cooperate it is a set of run on cell phone type intelligent terminal based on depth convolutional neural networks
Test software, can independently test with obtain user hearing compensation parameter sets and sound quality beautification parameter sets, pass through bluetooth
Communications are to intelligent earphone and store, and intelligent earphone is improved defeated by personalized hearing compensation module and sound quality beautification module
The clarity, naturalness and comfort level of sound out.
The above is only a preferred embodiment of the present invention, is not intended to limit the scope of the invention, all to utilize this hair
Equivalent structure made by bright specification and accompanying drawing content perhaps equivalent process transformation or be directly or indirectly used in other correlation
Technical field, be included within the scope of the present invention.
Claims (2)
1. a kind of method for promoting personalized sound quality function based on intelligent earphone, which comprises the steps of:
Step 1, the hearing compensation parameter sets for obtaining user individual and sound quality beautify parameter sets;
Step 2 carries out hearing compensation to the sound that intelligent earphone exports according to the hearing compensation parameter sets of user individual;
Step 3 beautifies parameter sets according to the sound quality of user individual to the progress sound quality beautification of intelligent earphone compensated sound;
Step 4, the sound by sound quality beautification are played by the output module of intelligent earphone;
By audiometery, speech hearing test and noise perception test in step 1, the hearing for obtaining user individual is mended
Repay parameter sets and sound quality beautification parameter sets;
Audiometery includes the following steps:
Step 101 generates the different tonal signals to sound intensity degree by intelligent terminal, and is played by the intelligent earphone of connection;
Step 102 feeds back the tonal signal that whether can hear broadcasting by user on intelligent terminal, obtains the pure tone of user
Threshold audiogram;It is fed back on intelligent terminal by user whether because uncomfortable too loudly, obtains the thresholds of uncomfortable level of user
Curve;
Step 103, intelligent terminal generate hearing user according to obtained pure tone audiometry curve, thresholds of uncomfortable level curve and prescriptive formula
Compensating parameter set initial value;
Speech hearing test includes the following steps:
List of the hearing compensation parameter sets initial value that step 201, intelligent terminal are obtained according to audiometery to test
Syllable vocabulary voice signal carries out hearing compensation, and compensated first monosyllable is played by intelligent earphone;
Step 202 feeds back heard monosyllable by user on intelligent terminal, obtains user and is working as to the monosyllable
Identification record under the preceding degree to the sound intensity;
Step 203 calculates user to the speech discrimination score for currently testing the consonant and vowel that are included with monosyllable, calculates public
Formula isWherein, RijFor speech discrimination score, i=1,2 ..., 32, j=1,
2 ..., J, J are the word number currently tested, wherein the speech discrimination score of the consonant and vowel do not tested keeps original value;
Step 204 calculates the monosyllabic FFT spectrum { X of current testk, k=1,2 ..., K }, K is FFT frequency point number;
Step 205, by current speech discrimination score { Rij, i=1,2 ..., 32 } and FFT spectrum { Xk, k=1,2 ..., N } input
First depth convolutional neural networks are iterated update, export hearing compensation parameter sets, including { multichannel loudness compensation is each
Channel inflection point Pij;Each channel compressions ratio CRij, i=1,2 ..., M;J=1,2 ..., N }, wherein N is the segmentation in each channel
Number, M indicate port number;
Step 206, intelligent terminal according to the updated hearing compensation parameter sets of iteration to next test with it is monosyllabic into
Row compensation, it is compensated monosyllabic by intelligent earphone broadcasting, and 202 are entered step, until single syllable vocabulary is completed,
It obtains optimizing hearing compensation parameter sets;
Noise perception test includes the following steps:
Step 301, intelligent terminal play the dialogic voice under different noise scenarios and are played by intelligent earphone;
Step 302, for the dialogic voice under variant noise scenarios, intelligent terminal proposes one group of leading question to user,
User feeds back the above problem in intelligent terminal, calculates satisfaction scoring according to feedback result;
Step 303 is higher than threshold value T when satisfaction scores, and output optimizes sound quality and beautifies parameter sets, exits noise perception and surveys
Examination, otherwise, enters step 304;
Step 304, feed back input the second depth convolutional neural networks by above-mentioned leading question, are iterated update, output
Sound quality beautifies parameter sets, including { each subband noise reduction parameters NRl, l=1,2 ..., L, each subband parametric equalizer EQl, l=1,
2 ..., L, each subband chauvent's criterion parameter HWl, l=1,2 ..., L }, L is the sub-band number of sound quality beautifying functions, is entered step
305;
Step 305, intelligent terminal beautify parameter sets according to updated sound quality and generate new test dialogic voice, pass through intelligence
Energy earphone plays, and enters step 302.
2. a kind of method for promoting personalized sound quality function based on intelligent earphone according to claim 1, which is characterized in that
In step 101, generate that for difference to the tonal signal of sound intensity degree, L takes 32, K in range from 125Hz to 8000Hz by intelligent terminal
It is 3 that take 256, M value, which be 8, N value, and the intelligent terminal is mobile phone;
The network structure of first depth convolutional neural networks and the second depth convolutional neural networks is all by 3 convolutional neural networks moulds
Block is constituted, and each module includes that a convolutional layer, a pond layer and a discarding layer, the output of three modules pass through flat layer
By multidimensional characteristic one-dimensional, then multiple convolutional layer characteristic aggregations are exported finally by two layers of full articulamentum by cascading layers
Final argument.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811171346.2A CN109246515B (en) | 2018-10-09 | 2018-10-09 | A kind of intelligent earphone and method promoting personalized sound quality function |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811171346.2A CN109246515B (en) | 2018-10-09 | 2018-10-09 | A kind of intelligent earphone and method promoting personalized sound quality function |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109246515A CN109246515A (en) | 2019-01-18 |
CN109246515B true CN109246515B (en) | 2019-10-29 |
Family
ID=65054372
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811171346.2A Active CN109246515B (en) | 2018-10-09 | 2018-10-09 | A kind of intelligent earphone and method promoting personalized sound quality function |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109246515B (en) |
Families Citing this family (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109982231B (en) * | 2019-04-04 | 2021-04-30 | 腾讯音乐娱乐科技(深圳)有限公司 | Information processing method, device and storage medium |
CN112306446A (en) * | 2019-07-26 | 2021-02-02 | 北京小米移动软件有限公司 | Hearing compensation method and device and electronic equipment |
WO2021127842A1 (en) * | 2019-12-23 | 2021-07-01 | 深圳市易优斯科技有限公司 | Equalizer setting method, apparatus and device, and computer readable storage medium |
CN111800696B (en) * | 2020-06-30 | 2022-06-21 | 歌尔科技有限公司 | Hearing assistance method, earphone, and computer-readable storage medium |
CN112969115B (en) * | 2021-01-22 | 2022-12-02 | 深圳市睿耳电子有限公司 | Tone quality optimization method of wireless earphone and related equipment |
CN112887877B (en) * | 2021-01-28 | 2023-09-08 | 歌尔科技有限公司 | Audio parameter setting method and device, electronic equipment and storage medium |
CN113015059B (en) * | 2021-02-23 | 2022-10-18 | 歌尔科技有限公司 | Audio optimization method, device, equipment and readable storage medium |
CN113286243A (en) * | 2021-04-29 | 2021-08-20 | 佛山博智医疗科技有限公司 | Error correction system and method for self-testing speech recognition |
WO2023272631A1 (en) | 2021-06-30 | 2023-01-05 | 深圳市汇顶科技股份有限公司 | Noise cancellation method for audio signal, audio signal processing apparatus, and electronic device |
CN114938487B (en) * | 2022-05-13 | 2023-05-30 | 东南大学 | Hearing aid self-checking method based on sound field scene discrimination |
CN116347284B (en) * | 2022-12-29 | 2024-04-09 | 荣耀终端有限公司 | Earphone sound effect compensation method and earphone sound effect compensation device |
CN118555518A (en) * | 2024-07-29 | 2024-08-27 | 深圳市美恩微电子有限公司 | Audio control method based on intelligent earphone |
Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101437190A (en) * | 2007-11-13 | 2009-05-20 | 美律实业股份有限公司 | Earphone with hearing-aid function |
CN103491491A (en) * | 2013-09-22 | 2014-01-01 | 江苏贝泰福医疗科技有限公司 | Full-digital hearing aid and non-traditional hearing aid fitting method |
CN104053112A (en) * | 2014-06-26 | 2014-09-17 | 南京工程学院 | Hearing aid self-fitting method |
CN104144374A (en) * | 2013-05-06 | 2014-11-12 | 展讯通信(上海)有限公司 | Listening assisting method and system based on mobile device |
CN105390142A (en) * | 2015-12-17 | 2016-03-09 | 广州大学 | Digital hearing aid voice noise elimination method |
CN106454642A (en) * | 2016-09-23 | 2017-02-22 | 佛山科学技术学院 | Adaptive sub-band audio feedback suppression method |
CN107374644A (en) * | 2017-08-29 | 2017-11-24 | 苏州妙听智能科技有限公司 | A kind of hearing test device |
CN107569236A (en) * | 2017-09-06 | 2018-01-12 | 杨捷 | A kind of Multifunctional listening test and hearing assistance system and its hearing test method |
CN107864441A (en) * | 2017-12-11 | 2018-03-30 | 北京康易聆医疗科技有限公司 | A kind of Bluetooth chip with hearing-aid function, bluetooth earphone and radio hearing aid system |
CN107911528A (en) * | 2017-12-15 | 2018-04-13 | 刘方辉 | A kind of hearing compensation system based on smart mobile phone and its self-service test method of completing the square |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101155439A (en) * | 2006-09-30 | 2008-04-02 | 安富科技股份有限公司 | Method for intensifying sound effect processing function of Bluetooth stereo device |
CN104811864B (en) * | 2015-04-20 | 2018-11-13 | 深圳市冠旭电子股份有限公司 | A kind of method and system of automatic adjusument audio |
-
2018
- 2018-10-09 CN CN201811171346.2A patent/CN109246515B/en active Active
Patent Citations (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101437190A (en) * | 2007-11-13 | 2009-05-20 | 美律实业股份有限公司 | Earphone with hearing-aid function |
CN104144374A (en) * | 2013-05-06 | 2014-11-12 | 展讯通信(上海)有限公司 | Listening assisting method and system based on mobile device |
CN103491491A (en) * | 2013-09-22 | 2014-01-01 | 江苏贝泰福医疗科技有限公司 | Full-digital hearing aid and non-traditional hearing aid fitting method |
CN104053112A (en) * | 2014-06-26 | 2014-09-17 | 南京工程学院 | Hearing aid self-fitting method |
CN105390142A (en) * | 2015-12-17 | 2016-03-09 | 广州大学 | Digital hearing aid voice noise elimination method |
CN106454642A (en) * | 2016-09-23 | 2017-02-22 | 佛山科学技术学院 | Adaptive sub-band audio feedback suppression method |
CN107374644A (en) * | 2017-08-29 | 2017-11-24 | 苏州妙听智能科技有限公司 | A kind of hearing test device |
CN107569236A (en) * | 2017-09-06 | 2018-01-12 | 杨捷 | A kind of Multifunctional listening test and hearing assistance system and its hearing test method |
CN107864441A (en) * | 2017-12-11 | 2018-03-30 | 北京康易聆医疗科技有限公司 | A kind of Bluetooth chip with hearing-aid function, bluetooth earphone and radio hearing aid system |
CN107911528A (en) * | 2017-12-15 | 2018-04-13 | 刘方辉 | A kind of hearing compensation system based on smart mobile phone and its self-service test method of completing the square |
Non-Patent Citations (2)
Title |
---|
基于移动终端的听力测试及补偿系统设计与实现;章勤杰;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20180415;全文 * |
数字助听器的自适应验配技术的研究;冯超;《中国优秀硕士学位论文全文数据库信息科技辑》;20170315;全文 * |
Also Published As
Publication number | Publication date |
---|---|
CN109246515A (en) | 2019-01-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109246515B (en) | A kind of intelligent earphone and method promoting personalized sound quality function | |
US8369549B2 (en) | Hearing aid system adapted to selectively amplify audio signals | |
US7936888B2 (en) | Equalization apparatus and method based on audiogram | |
US20140193008A1 (en) | System and method for fitting of a hearing device | |
CN107147981B (en) | Single ear intrusion speech intelligibility prediction unit, hearing aid and binaural hearing aid system | |
CN105376668B (en) | A kind of earphone noise-reduction method and device | |
CN107615651A (en) | System and method for improved audio perception | |
US11564048B2 (en) | Signal processing in a hearing device | |
CN104980870A (en) | Self-calibration of multi-microphone noise reduction system for hearing assistance devices using an auxiliary device | |
US11095995B2 (en) | Hearing assist device fitting method, system, algorithm, software, performance testing and training | |
CN109121057A (en) | A kind of method and its system of intelligence hearing aid | |
US20070286350A1 (en) | Speech-based optimization of digital hearing devices | |
CN107113523A (en) | Perceived to determine head related transfer function data according to user's sounding | |
US20240080609A1 (en) | Systems, apparatus, and methods for acoustic transparency | |
CN107564538A (en) | The definition enhancing method and system of a kind of real-time speech communicating | |
US20200069224A1 (en) | Hearing Assist Device Fitting Method And Software | |
Woods et al. | Predicting the effect of hearing loss and audibility on amplified speech reception in a multi-talker listening scenario | |
CN105596008A (en) | Method of fitting a hearing device to a user, a fitting system for a hearing device and a hearing device | |
CN107569236A (en) | A kind of Multifunctional listening test and hearing assistance system and its hearing test method | |
US20140205117A1 (en) | System and method for fitting hearing devices | |
CN107454537A (en) | Hearing devices including wave filter group and start detector | |
EP3864862A1 (en) | Hearing assist device fitting method, system, algorithm, software, performance testing and training | |
CN107320109B (en) | Frequency identification test method | |
Skinner et al. | Identification of speech by cochlear implant recipients with the Multipeak (MPEAK) and Spectral Peak (SPEAK) speech coding strategies I. Vowels | |
WO2020077348A1 (en) | Hearing assist device fitting method, system, algorithm, software, performance testing and training |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |