CN109729448A - Neck wears the voice control optimization method and device of formula interactive voice earphone - Google Patents
Neck wears the voice control optimization method and device of formula interactive voice earphone Download PDFInfo
- Publication number
- CN109729448A CN109729448A CN201711024112.0A CN201711024112A CN109729448A CN 109729448 A CN109729448 A CN 109729448A CN 201711024112 A CN201711024112 A CN 201711024112A CN 109729448 A CN109729448 A CN 109729448A
- Authority
- CN
- China
- Prior art keywords
- voice
- audio
- acoustic
- digital
- asr
- 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.)
- Pending
Links
- 230000002452 interceptive effect Effects 0.000 title claims abstract description 29
- 238000005457 optimization Methods 0.000 title claims abstract description 28
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000001914 filtration Methods 0.000 claims abstract description 37
- 230000005236 sound signal Effects 0.000 claims abstract description 30
- 238000006243 chemical reaction Methods 0.000 claims abstract description 17
- 230000006870 function Effects 0.000 claims description 31
- 210000000613 ear canal Anatomy 0.000 claims description 30
- 238000012545 processing Methods 0.000 claims description 24
- 238000010276 construction Methods 0.000 claims description 20
- 238000012546 transfer Methods 0.000 claims description 19
- 238000003058 natural language processing Methods 0.000 claims description 12
- 238000007493 shaping process Methods 0.000 claims description 11
- 230000007246 mechanism Effects 0.000 claims description 9
- 230000006978 adaptation Effects 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 8
- 238000005259 measurement Methods 0.000 claims description 6
- 238000011946 reduction process Methods 0.000 claims description 6
- 230000001629 suppression Effects 0.000 claims description 6
- 230000003595 spectral effect Effects 0.000 claims description 4
- 230000005764 inhibitory process Effects 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 3
- 210000000988 bone and bone Anatomy 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000012937 correction Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 208000003443 Unconsciousness Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000000919 ceramic Substances 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Abstract
The present invention provides voice control optimization methods and device that a kind of neck wears formula interactive voice earphone, the voice control optimization method that neck wears formula interactive voice earphone includes: reception audio signal, before audio signal enters speech recognition system (ASR), acoustic filtering is carried out to audio signal, obtains preprocessed audio；Preprocessed audio is acquired by microphone, generates digital audio by analog-to-digital conversion after analog circuit filters；Digital audio is treated as efficient voice instruction via speech recognition system (ASR)；According to the corresponding function of efficient voice instruction triggers smart machine.The present invention is directed to the application scenarios of pleasant microphone, proposes series of optimum method, then effectively inhibition noise contribution is pushed to relevant back-end services module to obtain purer voice data, to ensure the offer of high-quality service.
Description
Technical field
The present invention relates to intelligent wearable device technical fields, wear formula interactive voice earphone in particular to a kind of neck
Voice control optimization method and device, a kind of earphone integrating In-Ear microphone and a kind of wearable device.
Background technique
It listens and says, be the essential attribute of communication possessed by people.Correspondingly, intelligent wearable device be also required to it is basic
Communication ability can be said, can also be listened.The listening and speaking ability of intelligent wearable device is passed by the acoustics of two kinds of input and output
Sensor is completed.The former, acquires sound by microphone input sensor；The latter is broadcast by loudspeaker output transducer
Sound reproduction sound.It in general will not only include a kind of simple sound by the collected sound of common microphone, but voice,
The mixture of musical sound and various ambient sounds (noise).And many service (voices provided by intelligent wearable device
Wake up, speech recognition, voice communication, music etc.) completion, often only depend on the vocal sections of mixed sound, and
It is the voice of equipment wearer, ambient sound in addition to this, the presence of the voice of non-wearer and other noise contributions, instead
It can hinder the normal operation of certain services.Therefore, intelligent wearable device needs an acoustics front end subsystem, to complete externally
The acquisition and relevant treatment of boundary's sound, provide effective audio data to the greatest extent, to guarantee the various function services moulds in rear end
The normal operation of block.
Summary of the invention
The present invention is directed to solve at least one of the technical problems existing in the prior art or related technologies.Propose a kind of neck
Wear the voice control optimization method, device and the earphone and the wearable device that integrate In-Ear microphone of formula interactive voice earphone
The acquisition and relevant treatment to external sound are completed, provides effective audio data, to the greatest extent to guarantee the various function in rear end
The normal operation of energy service module.
For this purpose, an object of the present invention is to provide the voice control optimization sides that a kind of neck wears formula interactive voice earphone
Method.
It is another object of the present invention to a kind of corresponding voice control optimizations for providing neck and wearing formula interactive voice earphone
Device.
It is yet a further object of the present invention to provide a kind of earphones for integrating In-Ear microphone.
It is yet a further object of the present invention to provide a kind of wearable devices.
To achieve the above object, the first aspect of the present invention provides the voice control that a kind of neck wears formula interactive voice earphone
Optimization method handles the collected voice signal of audio signal reception device by speech recognition system (ASR) and obtains command signal, according to institute
Command signal realization is stated to the voice control of smart machine, comprising:
Audio signal is received, before the audio signal enters the speech recognition system (ASR), the audio is believed
Number carry out acoustic filtering, obtain preprocessed audio；
The preprocessed audio is acquired by microphone, generates digital sound by analog-to-digital conversion after analog circuit filters
Frequently；
The digital audio is treated as efficient voice instruction via speech recognition system (ASR)；
According to the corresponding function of the efficient voice instruction triggers smart machine.
Preferably, the acoustic filtering be by mechanical mechanism and/or acoustic construction to the voice signal of spatial domain into
Row noise reduction process.
In any of the above-described scheme preferably, the digital audio is effective language by speech recognition system (ASR) processing
Sound instruction, specifically includes: handling the number by the ear canal adaptation algorithm of DSP module, dereverberation algorithm and noise suppression algorithms
Word tone frequency, to the speech recognition system (ASR), the speech recognition system (ASR) will connect output treated digital signal
The digital signal received is converted into text information, and the text information is treated as institute by natural language processing system (NLP)
State efficient voice instruction.
In any of the above-described scheme preferably, the microphone is bone-conduction microphone and/or In-Ear microphone.
In any of the above-described scheme preferably, the preprocessed audio is believed according to the ssystem transfer function of acoustic construction
Number be filtered shaping, the ssystem transfer function of the acoustic construction by helmholtz resonance centre frequency fcIt determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
In any of the above-described scheme preferably, described that digital sound is generated by analog-to-digital conversion after analog circuit filters
Frequently, it specifically includes: the filtering is carried out according to the ssystem transfer function of analogue filter circuit.
In any of the above-described scheme preferably, it is treated as in the digital audio via speech recognition system (ASR)
Before efficient voice instruction, further includes: carry out digital filter to the digital audio according to the ssystem transfer function of digital filter
Wave.
In any of the above-described scheme preferably, the ear canal adaptation algorithm specifically includes: according to measurement, wearer is read aloud
The spectral response of fixed voice sequence obtain ear canal system relative to acoustic microphones system transmitting letter to the digital audio
It carries out ear canal and adapts to calibration.
Second part of the invention is related to the voice control optimization device that a kind of neck wears formula interactive voice earphone, passes through voice
Identifying system (ASR) the processing collected voice signal of audio signal reception device obtains command signal, according to the realization pair of described instruction signal
The voice control of smart machine characterized by comprising
Acoustic filtering unit, receive audio signal, the audio signal enter the speech recognition system (ASR) it
Before, acoustic filtering is carried out to the audio signal, obtains preprocessed audio；
AD conversion unit, the preprocessed audio are acquired by microphone, are turned after analog circuit filters by modulus
Change generation digital audio；
Voice recognition unit, the digital audio are treated as efficient voice via speech recognition system (ASR) and instruct；
Unit is executed instruction, according to the corresponding function of the efficient voice instruction triggers smart machine.
Preferably, the acoustic filtering be by mechanical mechanism and/or acoustic construction to the voice signal of spatial domain into
Row noise reduction process.
In any of the above-described scheme preferably, the voice recognition unit specifically includes: digital processing subelement, leads to
Ear canal adaptation algorithm, dereverberation algorithm and the noise suppression algorithms for crossing DSP module handle the digital audio, and treated for output
Digital signal to the speech recognition system (ASR), text conversion subelement, the speech recognition system (ASR) will receive
Digital signal be converted into text information, instruction generation unit, the text information passes through natural language processing system (NLP) quilt
Processing is that the efficient voice instructs.
In any of the above-described scheme preferably, the microphone is bone-conduction microphone and/or In-Ear microphone.
In any of the above-described scheme preferably, further includes: shaping filter unit transmits letter according to the system of acoustic construction
Several to be filtered shaping to the preprocessed audio signal, the ssystem transfer function of the acoustic construction is by helmholtz resonance
Centre frequency fcIt determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
In any of the above-described scheme preferably, the AD conversion unit specifically includes: analog filtering subelement, root
The filtering is carried out according to the ssystem transfer function of analogue filter circuit.
In any of the above-described scheme preferably, further includes: digital filtering subelement is passed according to the system of digital filter
Delivery function carries out digital filtering to the digital audio.
In any of the above-described scheme preferably, the ear canal adaptation algorithm specifically includes: according to measurement, wearer is read aloud
The spectral response of fixed voice sequence obtain ear canal system relative to acoustic microphones system transmitting letter to the digital audio
It carries out ear canal and adapts to calibration.
Part III of the invention is related to a kind of earphone for integrating In-Ear microphone, including the processing of shell, analogue audio frequency
Module and Digital audio process module, further includes:
Neck described in any of the above embodiments wears the voice control optimization device of formula interactive voice earphone；And Acoustic treatment mould
Block carries out acoustics filter according to mechanical mechanism and/or acoustic construction as the preposition processing module of the analogue audio frequency processing module
Wave；
Radio reception pipe in ear is connected to the Acoustic treatment module and the analogue audio frequency processing module, is arranged described
Enclosure interior is located in ear canal and carries out sound wave collection.
The fourth aspect of the present invention is related to a kind of wearable device, comprising: it is as described in the above technical scheme integrate it is In-Ear
The earphone of microphone.
The beneficial effect that the present invention obtains is: passing through circuit module, the numeric field of acoustic module, analog domain to spatial domain
DSP module related algorithm collaboration processing, complete to extract pendant to filtering/shaping/noise restraint of effective sound element
The voice and other noise contributions of the voice of wearer, filtering environmental noise and non-wearer, improve smart machine and refer to voice
The discrimination of order.
Additional aspect and advantage of the invention will be set forth in part in the description, and will partially become from the following description
Obviously, or practice through the invention is recognized.
Detailed description of the invention
Above-mentioned and/or additional aspect of the invention and advantage will become from the description of the embodiment in conjunction with the following figures
Obviously and it is readily appreciated that, in which:
Fig. 1 is a preferred embodiment of the voice control optimization method that neck according to the invention wears formula interactive voice earphone
Flow chart；
Fig. 2 is a preferred embodiment of the voice control optimization device that neck according to the invention wears formula interactive voice earphone
Schematic block diagram；
Fig. 3 is the schematic block diagram of a preferred embodiment of the earphone according to the invention for integrating In-Ear microphone；
Fig. 4 is a kind of schematic block diagram of a preferred embodiment of wearable device according to the invention；
Fig. 5 is the hardware system schematic diagram according to a preferred embodiment of wearable device of the present invention；
Fig. 6 is a preferred embodiment of the voice control optimization method that neck according to the invention wears formula interactive voice earphone
Signal processing flow schematic diagram.
Specific embodiment
To better understand the objects, features and advantages of the present invention, with reference to the accompanying drawing and specific real
Applying mode, the present invention is further described in detail.It should be noted that in the absence of conflict, the implementation of the application
Feature in example and embodiment can be combined with each other.
In the following description, numerous specific details are set forth in order to facilitate a full understanding of the present invention, still, the present invention may be used also
To be implemented using other than the one described here other modes, therefore, protection scope of the present invention is not by described below
Specific embodiment limitation.
Fig. 1 is a preferred embodiment of the voice control optimization method that neck according to the invention wears formula interactive voice earphone
Flow chart；
As shown in Figure 1, a kind of neck wears the voice control optimization method of formula interactive voice earphone, comprising: step 102, receive
Audio signal carries out acoustic filtering to audio signal before audio signal enters speech recognition system (ASR), obtains pre- place
Manage audio；Step 104, preprocessed audio is acquired by microphone, generates number by analog-to-digital conversion after analog circuit filters
Audio；Step 106, digital audio is treated as efficient voice instruction via speech recognition system (ASR)；Step 108, according to having
Imitate the corresponding function of phonetic order triggering smart machine.
According to the abovementioned embodiments of the present invention, it is preferable that acoustic filtering is by mechanical mechanism and/or acoustic construction to sky
Between domain voice signal carry out noise reduction process.
According to the abovementioned embodiments of the present invention, it is preferable that digital audio is effective language by speech recognition system (ASR) processing
Sound instruction, specifically includes: handling digital sound by the ear canal adaptation algorithm of DSP module, dereverberation algorithm and noise suppression algorithms
Frequently, output treated digital signal is to speech recognition system (ASR), the number letter that speech recognition system (ASR) will receive
It number is converted into text information, text information passes through natural language processing system (NLP, Natural Language
Processing) it is treated as efficient voice instruction.
According to the abovementioned embodiments of the present invention, it is preferable that microphone is bone-conduction microphone and/or In-Ear microphone.
According to the abovementioned embodiments of the present invention, it is preferable that further include: according to the ssystem transfer function of acoustic construction to pre- place
Reason audio signal is filtered shaping, the ssystem transfer function of acoustic construction by helmholtz resonance centre frequency fcIt determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
According to the abovementioned embodiments of the present invention, it is preferable that number is generated by analog-to-digital conversion after analog circuit filters
Audio specifically includes: being filtered according to the ssystem transfer function of analogue filter circuit.
According to the abovementioned embodiments of the present invention, it is preferable that be treated as in digital audio via speech recognition system (ASR)
Before efficient voice instruction, further includes: carry out digital filtering to digital audio according to the ssystem transfer function of digital filter.
According to the abovementioned embodiments of the present invention, it is preferable that ear canal adaptation algorithm specifically includes: according to measurement, wearer is read aloud
The spectral response of fixed voice sequence obtains ear canal system and carries out relative to the transmitting letter of acoustic microphones system to digital audio
Ear canal adapts to calibration.
As shown in Fig. 2, the second aspect of the present invention proposes the voice control optimization that a kind of neck wears formula interactive voice earphone
Device 200 handles the collected voice signal of audio signal reception device by speech recognition system (ASR) and obtains command signal, according to finger
Signal is enabled to realize the voice control to smart machine, further includes: acoustic filtering unit 202 receives audio signal, in audio signal
Into before speech recognition system (ASR), acoustic filtering is carried out to audio signal, obtains preprocessed audio；AD conversion unit
204, preprocessed audio is acquired by microphone, generates digital audio by analog-to-digital conversion after analog circuit filters；Voice is known
Other unit 206, digital audio are treated as efficient voice via speech recognition system (ASR) and instruct；Unit 208 is executed instruction,
According to the corresponding function of efficient voice instruction triggers smart machine.
According to the abovementioned embodiments of the present invention, it is preferable that acoustic filtering is by mechanical mechanism and/or acoustic construction to sky
Between domain voice signal carry out noise reduction process.
According to the abovementioned embodiments of the present invention, it is preferable that voice recognition unit 206 specifically includes: digital processing subelement
2062, digital audio, output processing are handled by the ear canal adaptation algorithm of DSP module, dereverberation algorithm and noise suppression algorithms
Digital signal afterwards to speech recognition system (ASR), text conversion subelement 2064, speech recognition system (ASR) will receive
Digital signal be converted into text information, instruction generation unit 2066, text information passes through natural language processing system (NLP) quilt
Processing is that efficient voice instructs.
According to the abovementioned embodiments of the present invention, it is preferable that microphone is bone-conduction microphone and/or In-Ear microphone.
According to the abovementioned embodiments of the present invention, it is preferable that further include: shaping filter unit 212, what it is according to acoustic construction is
System transmission function is filtered shaping to preprocessed audio signal, and the ssystem transfer function of acoustic construction is by helmholtz resonance
Centre frequency fcIt determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
According to the abovementioned embodiments of the present invention, it is preferable that AD conversion unit 204 specifically includes: analog filtering subelement
2042, it is filtered according to the ssystem transfer function of analogue filter circuit.
According to the abovementioned embodiments of the present invention, it is preferable that further include: digital filtering subelement 210, according to digital filter
Ssystem transfer function to digital audio carry out digital filtering.
As shown in figure 3, integrating the earphone 300 of In-Ear microphone, including shell 302,304 and of analogue audio frequency processing module
Digital audio process module 306, further includes: as the neck of above-mentioned any embodiment wears the voice control optimization of formula interactive voice earphone
Device 200；And Acoustic treatment module 308, as the preposition processing module of analogue audio frequency processing module, according to mechanical mechanism
And/or acoustic construction carries out acoustic filtering；Radio reception pipe 310 in ear is connected to Acoustic treatment module and analogue audio frequency processing mould
Block, setting inside housings, are located in ear canal and carry out sound wave collection.
A kind of wearable device 400 as shown in Figure 4, comprising: as described in above-described embodiment integrate In-Ear microphone
Earphone 300.
As shown in figure 5, acoustics front terminal system sounds signal processing flow, as described below:
Acoustical vibration when people speaks everywhere and in ear canal, then passes through Acoustic treatment mould to head by bone conduction
Block (mechanical mechanism/acoustic construction) carries out acoustic filtering and relevant treatment；
The vibration/sound signal that acoustic module is handled is converted into analogue audio frequency electricity by osteoacusis/pleasant microphone
Signal；
Analog audio signal carries out analog filtering and relevant treatment by analog circuit (analogue audio frequency processing module)；
The electric signal that analog module is handled is converted into storable digital signal by A/D converter；
Digital signal is converted into and is directed to by the complex process of the related algorithm loaded on digital signal processor (DSP)
Effective audio data of all kinds of back-end services；
Treated audio data activates relevant back-end services module, and by cloud, push/acquisition/presentation is all kinds of
The specific service of terminaloriented user；
To sum up, acoustics front end subsystem, effective to activate by the relevant treatment to equipment wearer human voice signal
Rear module, to provide the service of high quality.
Another embodiment according to the present invention, the performance of acoustics front end subsystem, is highly dependent on through sensor (Mike
Wind) metadata that obtains, and different sensors can provide the audio metadata of different characteristics:
Bone-conduction microphone (piezoelectric ceramics Mike) can collect the vibration of trunnion when people speaks/bone, physically every
Non-contacting ambient noise/non-wearer voice absolutely, however due to sound conduction path/medium difference and microphone
The acoustic characteristic of itself, compared with through the common microphone of air transmitted, collected voice has had great change,
Need to do corresponding shaping correction, could provide significantly more efficient data is used in rear end.
Pleasant microphone (electret mike) is located in ear canal, after the bone conduct vibrations to ear canal when people speaks, pushes
Air vibration in ear canal, by the sound conduction of wearer to microphone.And the passive Noise Reduction that earplug is fabulous, it equally can be with
Effectively isolated environment noise/non-wearer voice.However, similarly there is pleasant microphone sound conduction path/medium to ask
Data distortion caused by inscribing is asked；Moreover, ear canal closed by earplug after black-out effect, can low frequency part amplitude to voice generate
Up to 20~30dB influences；In addition, also will affect the clarity of voice data there are also the reverberation problem in ear canal.Above situation,
Relevant shaping correction is required, to provide significantly more efficient data as used in rear end.Therefore, acoustics front end subsystem, needs
The collaboration of the related algorithm of the acoustic module of spatial domain, the DSP module of the circuit module of analog domain, numeric field is handled, completion pair
Filtering/shaping/noise restraint of effective sound element.Relevant signal processing flow block diagram, as shown in Figure 6:
Since the related algorithm model of rear module (speech recognition/voice wakes up) is built upon standard acoustic Mike's elegance
On the data of collection, and the data that current system obtains are based on osteoacusis/pleasant microphone, and therefore, it is necessary to pass through test
Obtain ssystem transfer function H of the osteoacusis/pleasant microphone relative to acoustic microphonesRBE(n):
HRBE(n)=HBE(n)/HRef(n)
The ssystem transfer function H of standard acoustic microphoneRef(n) wherein, by the amplitude frequency of the standard microphone of actual measurement
Rate response and phase-frequency response accordingly determine；And osteoacusis/pleasant microphone ssystem transfer function HBE(n) by actual measurement
Osteoacusis/pleasant microphone amplitude-frequency response and phase-frequency response accordingly determine.
When carrying out relevant filter shape to osteoacusis/pleasant microphone data, need using HRBE(n) inverse function
HIRB(n), so that osteoacusis/pleasant microphone data and standard microphone data are more consistent after processing:
HIRB(n)=[HRBE(n)]-1
Again because of acoustic filtering module and analog filter block that osteoacusis/pleasant microphone data flows through, by acoustics
Various limitations of structure design and hardware simulation circuit design, therefore, HIRBE(n) specific implementation, meeting cutting are HAcoustic
(n), HAnalog(n) and HDigital(n) three parts first meet acoustic construction and the limitation of analog hardware circuit step by step, recycle number
The flexibility of filter module reaches overall filter shape requirement.
HIRBE(n)=HAcoustic(n)HAnalog(n)HDigital(n)
HAcousticIt (n) is the relevant ssystem transfer function of acoustic construction, key property, by Helmholtz resonance
Frequency of heart formula determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
Party's formula assumes that resonator is the simple structure that the pipeline of a cavity and sliver transvers section equalization is connected to form.
HAnalogIt (n) is the relevant ssystem transfer function of analog hardware filter circuit, HDigitalIt (n) is digital filter
Ssystem transfer function.
Generally, analog filter block and digital filtering module can achieve approximate filter effect, still, simulation filter
Wave module can provide bigger dynamic range, and digital filtering module can provide better flexibility, and the combination of the two can be with
Higher validity and reliability is provided to system design.
HIRBE(n) it decomposes after determining, so that it may which related procedure according to fig. 2 carries out algorithm process step by step.
Audio data is through acoustic module according to HAcoustic(n) (sound is not by microphone at this time after correlation properties processing
How acquisition, handled), incoming osteoacusis/pleasant microphone is acquired data.Then, through hardware simulation module
According to HAnalog(n) after correlation properties processing, then after A/D converter is handled, it is stored as digital signal.
Digital signal is by digital filtering module according to HDigital(n) correlation properties processing after, basically reach osteoacusis/
The effect of pleasant microphone enhancing algorithm.
But due to individual wearing mode and the difference of ear canal structure, HIRBE(n) it can be deposited with different because of wearer
In certain individual difference, needing ear canal to adapt to module using characteristic is Hcomp(n) difference is adjusted.
Ear canal adapts to module, and there are both of which: calibration mode and modification model.
Hcomp(n).Under calibration mode, ear canal adapts to the frequency spectrum sound that module reads aloud fixed voice sequence by measuring wearer
It answers, obtains
Under modification model, ear canal adapts to module and directly uses Hcomp(n) voice data is handled, to match individual spy
Property.
Then, for the application scenarios of pleasant microphone, it is also necessary to general dereverberation mould and noise suppression module, into one
Step effectively inhibits relevant noise contribution, to obtain purer voice data, is then pushed to relevant rear end
Service module, to ensure the offer of high-quality service.
After being pre-processed by acoustic filtering to collected acoustic signals according to above-mentioned each embodiment present invention again
Subsequent filtering processing is carried out through analog circuit and/or digital circuit, collected noise contribution is effectively inhibited, is passed through
ASR/NLP system substantially increases the standard of speech recognition to effective phonetic order is identified in more pure voice signal
Exactness, to ensure the offer of high-quality service.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any to repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of neck wears the voice control optimization method of formula interactive voice earphone, radio reception is handled by speech recognition system (ASR)
The collected voice signal of device obtains command signal, realizes the voice control to smart machine according to described instruction signal,
It is characterized in that, comprising:
Receive audio signal, the audio signal enter the speech recognition system (ASR) before, to the audio signal into
Row acoustic filtering obtains preprocessed audio；
The preprocessed audio is acquired by microphone, generates digital audio by analog-to-digital conversion after analog circuit filters；
The digital audio is treated as efficient voice instruction via speech recognition system (ASR)；
According to the corresponding function of the efficient voice instruction triggers smart machine.
2. the voice control optimization method that neck according to claim 1 wears formula interactive voice earphone, which is characterized in that described
Acoustic filtering is to carry out noise reduction process to the voice signal of spatial domain by mechanical mechanism and/or acoustic construction.
3. the voice control optimization method that neck according to claim 1 wears formula interactive voice earphone, which is characterized in that described
Digital audio is that efficient voice instructs by speech recognition system (ASR) processing, specifically includes: being adapted to by the ear canal of DSP module
Algorithm, dereverberation algorithm and noise suppression algorithms handle the digital audio, output treated digital signal to the voice
The digital signal received is converted text information, the text by identifying system (ASR), the speech recognition system (ASR)
Information is treated as the efficient voice instruction by natural language processing system (NLP).
4. the voice control optimization method that neck according to claim 1 wears formula interactive voice earphone, which is characterized in that described
Microphone is bone-conduction microphone and/or In-Ear microphone.
5. the voice control optimization method that neck according to claim 4 wears formula interactive voice earphone, which is characterized in that also wrap
It includes: shaping is filtered to the preprocessed audio signal according to the ssystem transfer function of acoustic construction, the acoustic construction
Ssystem transfer function by helmholtz resonance centre frequency fcIt determines:
Wherein, wherein c is air velocity, AHIt is the cross-sectional area of acoustic aperture, LHIt is the length of acoustic aperture, VCIt is the volume of cavity.
6. the voice control optimization method that neck according to claim 4 wears formula interactive voice earphone, which is characterized in that described
Digital audio is generated by analog-to-digital conversion after analog circuit filters, is specifically included: being passed according to the system of analogue filter circuit
Delivery function carries out the filtering.
7. the voice control optimization method that neck according to claim 4 wears formula interactive voice earphone, which is characterized in that in institute
It states digital audio to be treated as before efficient voice instruction via speech recognition system (ASR), further includes: according to digital filter
Ssystem transfer function to the digital audio carry out digital filtering.
8. the voice control optimization method that neck according to claim 3 wears formula interactive voice earphone, which is characterized in that described
Ear canal adaptation algorithm specifically includes: it is opposite to obtain ear canal system according to the spectral response that measurement wearer reads aloud fixed voice sequence
Ear canal is carried out to the digital audio in the transmitting letter of acoustic microphones system and adapts to calibration.
9. a kind of neck wears the voice control optimization device of formula interactive voice earphone, radio reception is handled by speech recognition system (ASR)
The collected voice signal of device obtains command signal, realizes the voice control to smart machine, institute according to described instruction signal
It states audio signal reception device the neck is set and wear on the earplug of formula interactive voice earphone and/or on host；It is characterised by comprising:
Acoustic filtering unit receives audio signal, right before the audio signal enters the speech recognition system (ASR)
The audio signal carries out acoustic filtering, obtains preprocessed audio；
AD conversion unit, the preprocessed audio are acquired by microphone, raw by analog-to-digital conversion after analog circuit filters
At digital audio；
Voice recognition unit, the digital audio are treated as efficient voice via speech recognition system (ASR) and instruct；
Unit is executed instruction, according to the corresponding function of the efficient voice instruction triggers smart machine；
It the acoustic filtering unit, AD conversion unit, voice recognition unit and executes instruction unit and is integrated in control dress
It sets.
10. the voice control optimization device that neck according to claim 9 wears formula interactive voice earphone, which is characterized in that institute
Stating acoustic filtering is to carry out noise reduction process to the voice signal of spatial domain by mechanical mechanism and/or acoustic construction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711024112.0A CN109729448A (en) | 2017-10-27 | 2017-10-27 | Neck wears the voice control optimization method and device of formula interactive voice earphone |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711024112.0A CN109729448A (en) | 2017-10-27 | 2017-10-27 | Neck wears the voice control optimization method and device of formula interactive voice earphone |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109729448A true CN109729448A (en) | 2019-05-07 |
Family
ID=66292001
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711024112.0A Pending CN109729448A (en) | 2017-10-27 | 2017-10-27 | Neck wears the voice control optimization method and device of formula interactive voice earphone |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109729448A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110265036A (en) * | 2019-06-06 | 2019-09-20 | 湖南国声声学科技股份有限公司 | Voice awakening method, system, electronic equipment and computer readable storage medium |
| CN113794963A (en) * | 2021-09-14 | 2021-12-14 | 深圳大学 | Speech enhancement system based on low-cost wearable sensor |
| CN114268820A (en) * | 2021-12-20 | 2022-04-01 | 歌尔科技有限公司 | Barrage sending method based on Bluetooth headset, equipment and medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201118917Y (en) * | 2007-08-24 | 2008-09-17 | 富声（东莞）电器配件有限公司 | Noise-reduction microphone |
| US20080260180A1 (en) * | 2007-04-13 | 2008-10-23 | Personics Holdings Inc. | Method and device for voice operated control |
| CN102138337A (en) * | 2008-08-13 | 2011-07-27 | W·W·格雷林 | Wearable headset with self-contained vocal feedback and vocal command |
| US20130028440A1 (en) * | 2011-07-26 | 2013-01-31 | Akg Acoustics Gmbh | Noise reducing sound reproduction system |
| CN103533472A (en) * | 2013-10-12 | 2014-01-22 | 中山市天键电声有限公司 | Front cavity structure of micro-receiver with acoustic filter |
| CN204652616U (en) * | 2015-04-14 | 2015-09-16 | 江苏南大电子信息技术股份有限公司 | A kind of noise reduction module earphone |
| CN106210983A (en) * | 2016-07-11 | 2016-12-07 | 歌尔股份有限公司 | A kind of realize the method for Kara OK function, device and earphone by earphone |
-
2017
- 2017-10-27 CN CN201711024112.0A patent/CN109729448A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080260180A1 (en) * | 2007-04-13 | 2008-10-23 | Personics Holdings Inc. | Method and device for voice operated control |
| CN201118917Y (en) * | 2007-08-24 | 2008-09-17 | 富声（东莞）电器配件有限公司 | Noise-reduction microphone |
| CN102138337A (en) * | 2008-08-13 | 2011-07-27 | W·W·格雷林 | Wearable headset with self-contained vocal feedback and vocal command |
| US20130028440A1 (en) * | 2011-07-26 | 2013-01-31 | Akg Acoustics Gmbh | Noise reducing sound reproduction system |
| CN103533472A (en) * | 2013-10-12 | 2014-01-22 | 中山市天键电声有限公司 | Front cavity structure of micro-receiver with acoustic filter |
| CN204652616U (en) * | 2015-04-14 | 2015-09-16 | 江苏南大电子信息技术股份有限公司 | A kind of noise reduction module earphone |
| CN106210983A (en) * | 2016-07-11 | 2016-12-07 | 歌尔股份有限公司 | A kind of realize the method for Kara OK function, device and earphone by earphone |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110265036A (en) * | 2019-06-06 | 2019-09-20 | 湖南国声声学科技股份有限公司 | Voice awakening method, system, electronic equipment and computer readable storage medium |
| CN113794963A (en) * | 2021-09-14 | 2021-12-14 | 深圳大学 | Speech enhancement system based on low-cost wearable sensor |
| CN113794963B (en) * | 2021-09-14 | 2022-08-05 | 深圳大学 | Speech enhancement system based on low-cost wearable sensor |
| CN114268820A (en) * | 2021-12-20 | 2022-04-01 | 歌尔科技有限公司 | Barrage sending method based on Bluetooth headset, equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11671773B2 (en) | Hearing aid device for hands free communication | |
| CN108200492A (en) | Voice control optimization method, device and the earphone and wearable device that integrate In-Ear microphone | |
| US10803857B2 (en) | System and method for relative enhancement of vocal utterances in an acoustically cluttered environment | |
| US20200213776A1 (en) | External ear insert for hearing enhancement | |
| JP2022531363A (en) | Auditory device with bone conduction sensor | |
| CN109729448A (en) | Neck wears the voice control optimization method and device of formula interactive voice earphone | |
| EP2482566B1 (en) | Method for generating an audio signal | |
| JP2002125298A (en) | Microphone device and earphone microphone device | |
| CN112116918A (en) | Speech signal enhancement processing method and earphone | |
| US11589173B2 (en) | Hearing aid comprising a record and replay function | |
| CN207518797U (en) | Neck wears the voice control optimization device of formula interactive voice earphone | |
| CN207884862U (en) | Audio frequency apparatus based on human ear simulation architecture | |
| CN113194374A (en) | Earphone based on bone voiceprint recognition technology | |
| CN207995324U (en) | Neck wears formula interactive voice earphone | |
| EP4351171A1 (en) | A hearing aid comprising a speaker unit | |
| EP4297436A1 (en) | A hearing aid comprising an active occlusion cancellation system and corresponding method | |
| US20120250918A1 (en) | Method for improving the comprehensibility of speech with a hearing aid, together with a hearing aid | |
| CN115942190A (en) | ENC noise reduction method and system for simulating sound generation in auditory canal based on artificial head frame device | |
| TW202406367A (en) | Earphone device, compensation method thereof and computer program product | |
| CN116709116A (en) | Sound signal processing method and earphone device | |
| CN113038315A (en) | Voice signal processing method and device | |
| CN109961783A (en) | A kind of integrated microphone earphone | |
| EP3318071A1 (en) | External ear insert for hearing enhancement | |
| Jung et al. | A Study on Noise Reduction in In-Ear Microphone using Two Microphones |
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 |