CN110415691A - Control method, device and computer readable storage medium based on speech recognition - Google Patents
Control method, device and computer readable storage medium based on speech recognition Download PDFInfo
- Publication number
- CN110415691A CN110415691A CN201810404609.3A CN201810404609A CN110415691A CN 110415691 A CN110415691 A CN 110415691A CN 201810404609 A CN201810404609 A CN 201810404609A CN 110415691 A CN110415691 A CN 110415691A
- Authority
- CN
- China
- Prior art keywords
- information
- acoustic information
- wake
- acoustic
- exempt
- 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
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Abstract
The embodiment of the invention discloses a kind of control methods based on speech recognition, belong to technical field of intelligent equipment.This method comprises: obtain acoustic information, when determine acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize the function that acoustic information is characterized.Using the technical solution provided in the embodiment of the present invention, can determine whether acoustic information is to exempt to wake up acoustic information, and without using word is waken up, user can interact with smart machine, and user experience effect is good.A kind of control device and computer readable storage medium based on speech recognition is also disclosed in the embodiment of the present invention.
Description
Technical field
The present invention relates to technical field of intelligent equipment, in particular to a kind of control method based on speech recognition, device and
Computer readable storage medium.
Background technique
Intelligent far field voice is the function of having merged a change user experience of artificial intelligent voice search, without pressing
The voice key of remote controler says instruction against remote controler.And smart television or smart box can be allowed by only saying instruction
Son is made a response.
Major part television manufacturer is laid out far field voice field one after another at present.In existing far field phonetic function, often use
" wake up word+function " expression way, still, each interactive process require to wake up word and make interacting for user and smart machine
Journey becomes trouble, affect experience effect.
Summary of the invention
The embodiment of the invention provides a kind of control method based on speech recognition, can determine whether voice messaging is to exempt to call out
Awake information, without using word is waken up, user can interact with smart machine.
In order to which some aspects of the embodiment to disclosure have a basic understanding, simple summary is shown below.It should
Summarized section is not extensive overview, nor to determine key/critical component or describe the protection scope of these embodiments.
Its sole purpose is that some concepts are presented with simple form, in this, as the preamble of following detailed description.
According to a first aspect of the embodiments of the present invention, a kind of control method based on speech recognition is provided, comprising:
Obtain acoustic information；
When determine the acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize the acoustic information
The function of being characterized.
In some alternative embodiments, the determination acoustic information, which belongs to, exempts to wake up information, comprising:
Traversal is exempted to wake up information bank, matches to the acoustic information；
If successful match, it is determined that the acoustic information, which belongs to, exempts to wake up information.
In some alternative embodiments, the acquisition acoustic information, comprising:
Obtain the first acoustic information of equipment end；
Obtain the second sound information in environment；
First acoustic information is filtered out in the second sound information, obtains the acoustic information.
In some alternative embodiments, the acquisition acoustic information, comprising:
Obtain third acoustic information；
Means of chaotic signals is filtered out in the third acoustic information, obtains the acoustic information.
According to a second aspect of the embodiments of the present invention, a kind of control device based on speech recognition is provided, comprising:
First module, for obtaining acoustic information；
Second module, for when determine the acoustic information belong to exempt to wake up information when, it is real to control corresponding executing agency
The function that the existing acoustic information is characterized.
In some alternative embodiments, second module includes:
Matching unit is exempted to wake up information bank, be matched to the acoustic information for traversing；If successful match, really
The fixed acoustic information, which belongs to, exempts to wake up information.
In some alternative embodiments, first module includes:
First unit, for obtaining the first acoustic information of equipment end；
Second unit, for obtaining the second sound information in environment；
Filter element obtains the sound for filtering out first acoustic information in the second sound information
Information.
In some alternative embodiments, first module includes:
Third unit, for obtaining third acoustic information；
Noise reduction unit obtains the acoustic information for filtering out means of chaotic signals in the third acoustic information.
According to a third aspect of the embodiments of the present invention, a kind of intelligent apparatus is provided, comprising:
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to:
Obtain acoustic information；
When determine the acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize the acoustic information
The function of being characterized.
According to a fourth aspect of the embodiments of the present invention, a kind of computer readable storage medium is provided, calculating is stored thereon with
Machine program realizes the above-mentioned control method based on speech recognition when the computer program is executed by processor.
Using the technical solution provided in the embodiment of the present invention, can determine whether acoustic information is to exempt to wake up acoustic information,
Without using word is waken up, user can interact with smart machine, and user experience effect is good.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
It can the limitation present invention.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows and meets implementation of the invention
Example, and be used to explain the principle of the present invention together with specification.
Fig. 1 is a kind of flow diagram of control method based on speech recognition shown according to an exemplary embodiment；
Fig. 2 is a kind of flow diagram for obtaining acoustic information shown according to an exemplary embodiment；
Fig. 3 is a kind of flow diagram for obtaining acoustic information shown according to an exemplary embodiment；
Fig. 4 is a kind of flow diagram of control method based on speech recognition shown according to an exemplary embodiment；
Fig. 5 is that a kind of pair of acoustic information shown according to an exemplary embodiment carries out matched flow diagram；
Fig. 6 is that a kind of pair of acoustic information shown according to an exemplary embodiment carries out matched flow diagram；
Fig. 7 is that a kind of pair of acoustic information shown according to an exemplary embodiment carries out matched flow diagram；
Fig. 8 is a kind of block diagram of control device based on speech recognition shown according to an exemplary embodiment；
Fig. 9 is a kind of block diagram of control device based on speech recognition shown according to an exemplary embodiment；
Figure 10 is a kind of block diagram of control device based on speech recognition shown according to an exemplary embodiment；
Figure 11 is a kind of block diagram of control device based on speech recognition shown according to an exemplary embodiment.
Specific embodiment
The following description and drawings fully show specific embodiments of the present invention, to enable those skilled in the art to
Practice them.Embodiment only represents possible variation.Unless explicitly requested, otherwise individual components and functionality is optional, and
And the sequence of operation can change.The part of some embodiments and feature can be included in or replace other embodiments
Part and feature.The range of embodiment of the present invention includes the entire scope of claims and the institute of claims
There is obtainable equivalent.Herein, each embodiment can individually or generally be indicated that this is only with term " invention "
It is merely for convenience, and if in fact disclosing the invention more than one, it is not meant to automatically limit the range of the application
For any single invention or inventive concept.Herein, relational terms such as first and second and the like are used only for one
Entity, which is perhaps operated, to be distinguished and exists without requiring or implying between these entities or operation with another entity or operation
Any actual relationship or sequence.Moreover, the terms "include", "comprise" or its any other variant be intended to it is non-exclusive
Property include so that include a series of elements process, method or equipment not only include those elements, but also including
Other elements that are not explicitly listed.Each embodiment herein is described in a progressive manner, and each embodiment stresses
Be the difference from other embodiments, the same or similar parts in each embodiment may refer to each other.For implementing
For structure, product etc. disclosed in example, since it is corresponding with part disclosed in embodiment, so being described relatively simple, phase
Place is closed referring to method part illustration.
Wake-up word herein refers to pet name that artificial intelligence personalizes, such as little Bai, small T etc..
Equipment herein can be the terminal intelligents equipment such as smart television, smart phone, tablet computer and intelligent hand
The wearable smart machine such as table, Intelligent bracelet, intelligent glasses, and it is without being limited thereto.
According to a first aspect of the embodiments of the present invention, a kind of control method based on speech recognition is provided.
As shown in Figure 1, in some alternative embodiments, being somebody's turn to do the control method based on speech recognition, comprising:
S101, acoustic information is obtained；
About above sound information, in some alternative embodiments, acoustic information includes waveform sound information and language
Adopted acoustic information.Wherein, waveform sound information is the acoustic information of acquisition, which is showed in the form of waveform, is parsed
Waveform sound information can obtain semantic acoustic information.For example, waveform sound can be obtained after the acoustic information that user says is collected
Message breath, the waveform sound information can be semantic acoustic information by escape, and semantic acoustic information is showed in the form of text at this time,
Text therein is the semanteme that can characterize function.Optionally, pass through data science and research institute IDST
(InstituteofDataScience&Technologies) waveform sound information is parsed to obtain semantic acoustic information.By wave
Shape acoustic information escape is semantic acoustic information, convenient for obtaining the semanteme in acoustic information, is used to accurately obtain for characterizing
The control instruction of family demand.About user demand, for example, when user does not hear the sound of smart television, the need of user at this time
Seeking Truth increases volume.
S102, when determine acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize acoustic information institute
The function of characterization.
For example, acoustic information is " increase volume ", and the acoustic information that " should increase volume " belongs to and exempts to wake up information, this
When executing agency have volume adjusting function correlation module, to volume adjustment mechanism send increase volume instruction, thus
Realize the function of increasing volume.
In the prior art scheme, if the demand of user is to increase the volume of smart television, user needs to tell " so-and-so
(the wake-up word of the smart television, such as little Bai) increases volume ", when user frequently controls the smart television, need
Frequently say the wake-up word unrelated with regulatory function, interactive process will show cumbersome.
And in the control method based on speech recognition, can determine whether acoustic information is to exempt to wake up information, using this
Technical solution in embodiment, without using word is waken up, user can interact with smart machine, user experience
Effect is good.
During the determination of S102, however, it is determined that acoustic information is not belonging to exempt to wake up information, and when in determining voice messaging
When comprising waking up word, controlling corresponding executing agency and realizing the function that acoustic information is characterized.
In the technical scheme, user can unrestricted choice whether need to wake up word, strong flexibility, user experience effect is good.
The technical program is suitable for the application scenarios of far field speech recognition, therefore during S101 obtains sound, always
With echoing, the semantic acoustic information which will lead to the escape of waveform voice messaging institute is not accurate enough, thus cannot be quasi-
Really obtain the demand of user.
Based on the above issues, as shown in Fig. 2, in some alternative embodiments, obtaining acoustic information, comprising:
S201, the first acoustic information for obtaining equipment end；
Optionally, the first acoustic information of equipment end is obtained in a manner of multichannel synchronousing collection.
Optionally, the first acoustic information in S201 is waveform sound information.
Second sound information in S202, acquisition environment；
Environment in this step refers to the position far from smart machine.
Optionally, second sound information whole in current environment is obtained in a manner of multichannel synchronousing collection.Optionally,
Second sound information in S202 is waveform sound information.
S203, the first acoustic information is filtered out in second sound information, obtain acoustic information.
The acoustic information obtained in this step is the acoustic information for filtering out echo, herein should for convenience of describing
The acoustic information for filtering out echo is known as echoless acoustic information, accordingly, the acoustic information for not filtering out echo is known as having
Echo acoustic information.It optionally, is semantic acoustic information by echoless acoustic information escape after S203.
Echoless acoustic information is than user's actual need is more actually reflected, when echoless acoustic information belongs to waveform sound
When message ceases, which can parse the semantic acoustic information than more actually characterizing user demand, Jin Erke
It more accurately controls corresponding executing agency and realizes the function that voice messaging is characterized.
During obtaining user voice information, the signal for being loaded with user voice information is interfered, and will generate and make an uproar
Sound.For example, electromagnetic interference generates noise, power supply disturbance generates noise and earth-return circuit also generates noise.These noises
It is present in waveform sound information.It will lead to and be difficult to be accurate semantic acoustic information by waveform sound information escape, or even lead
It causes not being semantic acoustic information by waveform sound information escape.
As shown in figure 3, in some alternative embodiments, obtaining acoustic information, comprising:
S301, third acoustic information is obtained；
Third acoustic information in this step refers to the acoustic information comprising user speech, including echoless sound
Information and there is echo acoustic information.
Optionally, the third acoustic information in S301 is obtained in a manner of two channels or multichannel.
S302, noise signal is filtered out in third acoustic information, obtain acoustic information.
For ease of description, the acoustic information for filtering out noise signal in S302 and obtaining is known as noise reduction herein
Acoustic information.It optionally, is semantic acoustic information by noise reduction acoustic information escape after S302.
In a kind of optional mode classification, noise signal can be divided into: steady-state noise, nonstationary noise and impulsive noise
Deng.
Noise reduction acoustic information is than user's actual need is more actually reflected, when noise reduction acoustic information belongs to waveform sound letter
When breath, which can parse the semantic acoustic information than more actually characterizing user demand, and then than calibrated
The true corresponding executing agency of control realizes the function that voice messaging is characterized.
The standard of acoustic information can be improved in the above-mentioned echo cancellation process carried out to acoustic information and reduction noise processed
True property.In some alternative embodiments, echo cancellation process is carried out to acoustic information and reduces noise processed.Further mention
The accuracy of high sound information.
As shown in figure 4, the specific implementation step of the technical program is described in detail by taking the first processing mode as an example:
S401, the first acoustic information for obtaining equipment end；
Second sound information in S402, acquisition environment；
S403, the first acoustic information is filtered out in second sound information, obtain third acoustic information；
Third acoustic information in this step is echoless acoustic information；
S404, means of chaotic signals is filtered out in third acoustic information, obtain waveform sound information；
Acoustic information obtained is the sound letter for passing through echo cancellation process and reducing noise processed in this step
Breath；
S405, parsing waveform sound information are to obtain semantic acoustic information；
Optionally, waveform sound information is parsed to obtain semantic acoustic information by IDST；
S406, when determine semantic acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize semantic sound
Message ceases characterized function；
S407, include when determining that semantic acoustic information is not belonging to exempt to wake up information, and when confirming in semantic acoustic information
When waking up word, controls corresponding executing agency and realize the function that semantic acoustic information is characterized.
In some alternative embodiments, it determines that acoustic information belongs to exempt to wake up information, comprising:
Traversal is exempted to wake up information bank, matches to acoustic information；
If successful match, it is determined that the acoustic information, which belongs to, exempts to wake up information.
Wherein, exempting from wake up information bank in, comprising it is several it is preset exempt from wake up information.In some alternative embodiments,
It is preset to exempt to wake up information for characterizing the common function of user.
For example, the user for seeing that video, adjusting volume etc. belong on smart television is common when applying on smart television
Function, these functions all with it is corresponding exempt from wake up information association, and be preset at exempt from wake up information bank in.So, when user thinks
When seeing video, so that it may say " I will see certain video ", it is control when exempting to wake up information that smart television, which recognizes " I will see certain video ",
Fixture has the correlation module removal search video of search video capability, to realize characterization selection video in the acoustic information of user
Function.Above-mentioned " certain video " can be some film, some TV play, some cartoon etc..Further specifically, there is search
The correlation module of video capability is when searching for video according to the voice messaging of user, if only searching a video resource, directly
It connects and plays the video resource；If enumerating all video resources searched if searching cadre's associated video resource, wait stand-by
Family is further to be selected.Smart television is mute or cancels the function that mute, smart television increases volume or reduction volume, control
Process processed is similar with the above-mentioned realization search control flow of function of video, and those skilled in the art can search for according to above-mentioned realization
The control flow of the function of video designs the corresponding control flow for realizing other common functions, no longer repeats one by one here.
If it fails to match, the determination acoustic information is not belonging to exempt to wake up information, gives up the acoustic information.If matching is lost
It loses, then it represents that exempting to wake up in information bank does not have the preset acoustic information, about the acoustic information that it fails to match, there is following several feelings
Condition: it when identifying the acoustic information, needs to wake up word；The function that the acoustic information is characterized, which is not belonging to intelligent apparatus, to be executed
Function；The acoustic information cannot characterize functional semanteme.
In some alternative embodiments, when matching to acoustic information, which is semantic sound letter
Breath.Compared to waveform sound information, semantic acoustic information is more convenient for being matched.It is corresponding, it is preset to exempt to wake up institute in information bank
Exempt to wake up acoustic information as semantic acoustic information.
As shown in figure 5, in some alternative embodiments, traversal is exempted to wake up information bank, acoustic information is matched,
The following steps are included:
S501, N characteristic character included in semantic acoustic information is obtained.
In S501, N for the length less than the included character of semantic acoustic information nonnegative integer, alternatively, N be less than or
Equal to the positive integer of the length of the included character of semantic acoustic information；For example, semantic acoustic information is " reducing volume ", then N
For the nonnegative integer less than 4, i.e. N=0,1,2,3；Alternatively, N is positive integer less than or equal to 4, i.e. N=1,2,3,4.
S502, traversal N-1 temporarily exempt to wake up information bank, obtain and exempt from wake-up information comprising N characteristic character.
In S502, if N characteristic character is the first character of semantic acoustic information, such as the 0th characteristic character or
1st characteristic character temporarily exempts to wake up information bank as N-1 in order to avoid waking up information bank at this time, that is, exempts to wake up information bank to be interim
Exempt from the original state of wake-up information bank.
S503, when confirm successfully obtain comprising N characteristic character exempt from wake up information when, with this include N characteristic character
Exempt from wake up information as N temporarily exempt from wake up information bank.
In S503, confirms successfully to obtain and exempt from wake-up information comprising N characteristic character, refer to temporarily exempting from N-1
There are one or more in wake-up information bank exempts from wake-up information comprising N characteristic character.
S504, when confirmation exempt from wake up information bank first exempt from wake up information included character whole successful match after, really
Recognize the acoustic information belong to exempt from wake up information.
In S504, first, which exempts from wake-up information, refers to exempting from for any one exempted from wake up in information bank to wake up information.Confirmation
Exempt to wake up first in information bank to exempt to wake up the character whole successful match that information is included, refer in semantic acoustic information
In acquired several characteristic characters, exempt to wake up all characters in information comprising first.Such as first exempt from wake up information be " return
Return desktop ", then in several characteristic characters acquired in S501, including " returning ", " returning ", " table ", " face " this four words, then
It can confirm that this first exempts to wake up the information character whole successful match that is included.
As shown in fig. 6, optionally, S501 may be selected to be embodied as following steps to S504:
S601, the i-th=0 characteristic character for obtaining semantic acoustic information；
S602, traversal are exempted to wake up information bank；
S603, matching comprising the i-th characteristic character x (x=0,1,2,3 ...) article exempt from wake up information Y (Y=y1, y2,
Y3 ...) a character；
In S603, x articles of matching exempts to wake up the Y character of information, refers to that xth article exempts to wake up yx in information
Character matches with the i-th=0 characteristic character.For example, illustrating that being matched to 2 exempts to wake up information, if this two are exempted from as x=2
Waking up information is " television mute " and " increasing volume ", and the i-th=0 characteristic character of semantic acoustic information is " sound ", then, at this time
Y1=3, y2=2.In S603, each of Y element is started counting from 0.
S604, judge whether to be successfully matched to and exempt to wake up information: if so, executing S605；Otherwise S613 is executed；
In S604, when confirming x is positive integer, that is, indicates to be successfully matched to and exempt to wake up information, can confirm and successfully obtain
It takes and exempts from wake-up information comprising the i-th characteristic character；
S605, it empties and temporarily exempts to wake up information bank；
S606, x item is exempted to wake up information and be added temporarily to exempt to wake up information bank；
S607, successively judge that the whether small information of exempting to wake up corresponding to ym of every ym for exempting to wake up in information is included
The number of character；If it exists one be unsatisfactory for above-mentioned Rule of judgment exempt from wake up information, then execute S614；If all x items exempt to call out
Ym in information of waking up is respectively less than the number for exempting to wake up the included character of information corresponding to ym, then executes S608；
In S607, there are one be unsatisfactory for Rule of judgment exempt from wake up information, can confirm this exempt from wake up information institute
The character whole successful match for including；M=1,2,3 ..., x；
S608, Y=Y+1；
In S608, each of Y element carries out plus 1 operation；
S609, i=i+1；
S610, judge whether i is less than the number of the included character of semantic acoustic information；If so, executing S611；Otherwise it executes
S613；
S611, the i-th characteristic character for obtaining semantic acoustic information；
S612, traversal are interim to wake up information bank；Execute step S603；
S613, it fails to match；
S614, confirm the semanteme acoustic information belong to exempt from wake up information.
In the technical scheme, erroneous judgement can be reduced, the number of maloperation is reduced, improves the accuracy of speech recognition.For example,
It applies on smart phone, exempts to wake up information when " making visual telephone to ABC " is, then semantic acoustic information must strictly be " to beat
Visual telephone just can confirm that the semanteme acoustic information belongs to and exempt to wake up information to ABC ".
Optionally, judge whether to be successfully matched in S604 and exempt to wake up in information, if it is not, then executing S609.
In the technical scheme, intelligence degree can be improved.For example, apply on smart phone, when " beat visual telephone to
ABC " is to exempt to wake up information, then semantic acoustic information includes that " making visual telephone to ABC " can confirm the semanteme acoustic information category
In exempt from wake up information.Such as " I will make visual telephone to ABC " and " making visual telephone to ABC at once " belongs to exempt to wake up letter
Breath.
As shown in fig. 7, optionally, S501 may be selected to be embodied as following steps to S504:
S701, the i-th=0 characteristic character for obtaining semantic acoustic information；
S702, traversal are exempted to wake up information bank；
S703, matching exempt to wake up information comprising x (x=0,1,2,3 ...) item of the i-th characteristic character, and record every and exempt to call out
Succeeded matched number of characters Y (Y=y1, y2, y3 ...) in information of waking up；
In S703, exempts from wake-up information for the m articles and succeeded matched number of characters as ym；
S704, judge whether to be successfully matched to and exempt to wake up information: if so, executing S705；Otherwise S709 is executed；
S705, it empties and temporarily exempts to wake up information bank；
S706, x item is exempted to wake up information and be added temporarily to exempt to wake up information bank；
S707, successively judge that every is exempted to wake up whether the matched number of characters ym that succeeded in information is all larger than corresponding to ym
Exempt from the number of wake-up the included character of information；If so, executing S713；Otherwise S708 is executed；
S708, i=i+1；
S709, judge whether i is less than the number of the included character of semantic acoustic information；If so, executing S710；Otherwise it executes
S712；
S710, the i-th characteristic character for obtaining semantic acoustic information；
S711, traversal are interim to wake up information bank；Execute step S703；
S712, it fails to match；
S713, confirm the semanteme acoustic information belong to exempt from wake up information.
In the technical scheme, intelligence degree is further improved.Such as it applies in the intelligence that may be viewed by network video
When in equipment, using " continuing to play " as exempt from wake up information, then include " after ", " continuous ", " broadcasting ", " putting " this four words semantic sound
Message breath, which can be confirmed to be, exempts to wake up information.
According to a second aspect of the embodiments of the present invention, a kind of control device based on speech recognition is provided.
As shown in figure 8, in some alternative embodiments, the control device based on speech recognition includes:
First module 10, for obtaining acoustic information；
Second module 20, for when determine acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize
The function that acoustic information is characterized.
In some alternative embodiments, acoustic information includes waveform sound information and semantic acoustic information.
In some alternative embodiments, the control device based on speech recognition further includes meaning transferring module, for passing through
IDST parses waveform sound information to obtain semantic acoustic information.
As shown in figure 9, in some alternative embodiments, the second module 20 includes:
Matching unit 21 is exempted to wake up information bank, be matched to acoustic information for traversing；If successful match, it is determined that
Acoustic information, which belongs to, exempts to wake up information.
In some alternative embodiments, when matching to acoustic information, which is semantic sound letter
Breath.
In some alternative embodiments, matching unit 21 includes:
First subelement, for obtaining N characteristic character included in semantic acoustic information；
Second subelement temporarily exempts to wake up information bank, obtains and exempt to wake up comprising N characteristic character for traversing N-1
Information；
Third subelement, for when confirm successfully obtain comprising N characteristic character when exempting to wake up information, include the with this
The information of exempting to wake up of N characteristic character is temporarily exempted to wake up information bank as N；
4th subelement exempts from the first of wake-up information bank when confirmation and exempts from the character whole successful match that wake-up information is included
Afterwards, confirm that the acoustic information belongs to exempt to wake up information.
As shown in Figure 10, in some alternative embodiments, the first module 10 includes:
First unit 11, for obtaining the first acoustic information of equipment end；
Second unit 12, for obtaining the second sound information in environment；
Filter element obtains acoustic information for filtering out the first acoustic information in second sound information.
In some alternative embodiments, first unit 11 in a manner of multichannel synchronousing collection for obtaining equipment end
The first acoustic information.Wherein, the first acoustic information is waveform sound information.
In some alternative embodiments, second unit 12 works as front ring for obtaining in a manner of multichannel synchronousing collection
Whole second sound information in border.Wherein, second sound information is waveform sound information.
As shown in figure 11, in some alternative embodiments, the first module 10 includes:
Third simple eye 13, for obtaining third acoustic information；
Noise reduction unit 15 obtains acoustic information for filtering out means of chaotic signals in third acoustic information.
In some alternative embodiments, third simple eye 13 can be used to obtain equipment in a manner of multichannel synchronousing collection
The second sound information at end.Wherein, third acoustic information is waveform sound information.
According to a third aspect of the embodiments of the present invention, a kind of intelligent apparatus is provided, comprising:
Processor；
Memory for storage processor executable instruction；
Wherein, processor is configured as:
Obtain acoustic information；
When determine acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize what acoustic information was characterized
Function.
Optionally, the previously described control method and device based on speech recognition can be real in network side server
It is existing, alternatively, can also realize in the terminal, alternatively, being realized in dedicated control equipment.
According to a fourth aspect of the embodiments of the present invention, a kind of computer readable storage medium is provided, calculating is stored thereon with
Machine program realizes previously described method when the computer program is executed by processor.Above-mentioned computer-readable storage medium
Matter includes read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access
Memory), tape and light storage device etc..
Those of ordinary skill in the art may be aware that list described in conjunction with the examples disclosed in the embodiments of the present disclosure
Member and algorithm steps can be realized with the combination of electronic hardware or computer software and electronic hardware.These functions are actually
It is implemented in hardware or software, the specific application and design constraint depending on technical solution.Those of skill in the art
Each specific application can be used different methods to achieve the described function, but this realization is it is not considered that exceed
The scope of the present invention.It is apparent to those skilled in the art that for convenience and simplicity of description, foregoing description
The specific work process of system and device, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
It should be understood that the invention is not limited to the process and structure that are described above and are shown in the accompanying drawings,
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is only limited by the attached claims
System.
Claims (10)
1. a kind of control method based on speech recognition characterized by comprising
Obtain acoustic information；
When determine the acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize acoustic information institute table
The function of sign.
2. control method according to claim 1, which is characterized in that the determination acoustic information, which belongs to, exempts to wake up letter
Breath, comprising:
Traversal is exempted to wake up information bank, matches to the acoustic information；
If successful match, it is determined that the acoustic information, which belongs to, exempts to wake up information.
3. control method according to claim 1 or 2, which is characterized in that the acquisition acoustic information, comprising:
Obtain the first acoustic information of equipment end；
Obtain the second sound information in environment；
First acoustic information is filtered out in the second sound information, obtains the acoustic information.
4. control method according to claim 1 or 2, which is characterized in that the acquisition acoustic information, comprising:
Obtain third acoustic information；
Means of chaotic signals is filtered out in the third acoustic information, obtains the acoustic information.
5. a kind of control device based on speech recognition characterized by comprising
First module, for obtaining acoustic information；
Second module, for when determine the acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize institute
State the function that acoustic information is characterized.
6. control device according to claim 5, which is characterized in that second module includes:
Matching unit is exempted to wake up information bank, be matched to the acoustic information for traversing；If successful match, it is determined that institute
State acoustic information belong to exempt from wake up information.
7. control device according to claim 5 or 6, which is characterized in that first module includes:
First unit, for obtaining the first acoustic information of equipment end；
Second unit, for obtaining the second sound information in environment；
Filter element obtains the acoustic information for filtering out first acoustic information in the second sound information.
8. control device according to claim 5 or 6, which is characterized in that first module includes:
Third unit, for obtaining third acoustic information；
Noise reduction unit obtains the acoustic information for filtering out means of chaotic signals in the third acoustic information.
9. a kind of intelligent apparatus characterized by comprising
Processor；
Memory for storage processor executable instruction；
Wherein, the processor is configured to:
Obtain acoustic information；
When determine the acoustic information belong to exempt to wake up information when, control corresponding executing agency and realize acoustic information institute table
The function of sign.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that when the computer journey
The control method based on speech recognition as described in any one of Claims 1-4 is realized when sequence is executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810404609.3A CN110415691A (en) | 2018-04-28 | 2018-04-28 | Control method, device and computer readable storage medium based on speech recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810404609.3A CN110415691A (en) | 2018-04-28 | 2018-04-28 | Control method, device and computer readable storage medium based on speech recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110415691A true CN110415691A (en) | 2019-11-05 |
Family
ID=68357421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810404609.3A Pending CN110415691A (en) | 2018-04-28 | 2018-04-28 | Control method, device and computer readable storage medium based on speech recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415691A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112929724A (en) * | 2020-12-31 | 2021-06-08 | 海信视像科技股份有限公司 | Display device, set top box and far-field pickup awakening control method |
| WO2022247244A1 (en) * | 2021-05-24 | 2022-12-01 | 青岛海尔空调器有限总公司 | Voice control method for air conditioner, and air conditioner |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004234273A (en) * | 2003-01-30 | 2004-08-19 | Hitachi Ltd | Interactive terminal device and interactive application providing method |
| CN105225662A (en) * | 2015-08-24 | 2016-01-06 | 深圳市冠旭电子有限公司 | Smart bluetooth earphone plays method and the smart bluetooth earphone of external voice automatically |
| CN105551498A (en) * | 2015-10-28 | 2016-05-04 | 东莞酷派软件技术有限公司 | Voice recognition method and device |
| WO2017071182A1 (en) * | 2015-10-26 | 2017-05-04 | 乐视控股（北京）有限公司 | Voice wakeup method, apparatus and system |
| CN107274889A (en) * | 2017-06-19 | 2017-10-20 | 北京紫博光彦信息技术有限公司 | A kind of method and device according to speech production business paper |
| CN107564518A (en) * | 2017-08-21 | 2018-01-09 | 百度在线网络技术（北京）有限公司 | Smart machine control method, device and computer equipment |
-
2018
- 2018-04-28 CN CN201810404609.3A patent/CN110415691A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004234273A (en) * | 2003-01-30 | 2004-08-19 | Hitachi Ltd | Interactive terminal device and interactive application providing method |
| CN105225662A (en) * | 2015-08-24 | 2016-01-06 | 深圳市冠旭电子有限公司 | Smart bluetooth earphone plays method and the smart bluetooth earphone of external voice automatically |
| WO2017071182A1 (en) * | 2015-10-26 | 2017-05-04 | 乐视控股（北京）有限公司 | Voice wakeup method, apparatus and system |
| CN105551498A (en) * | 2015-10-28 | 2016-05-04 | 东莞酷派软件技术有限公司 | Voice recognition method and device |
| CN107274889A (en) * | 2017-06-19 | 2017-10-20 | 北京紫博光彦信息技术有限公司 | A kind of method and device according to speech production business paper |
| CN107564518A (en) * | 2017-08-21 | 2018-01-09 | 百度在线网络技术（北京）有限公司 | Smart machine control method, device and computer equipment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112929724A (en) * | 2020-12-31 | 2021-06-08 | 海信视像科技股份有限公司 | Display device, set top box and far-field pickup awakening control method |
| WO2022247244A1 (en) * | 2021-05-24 | 2022-12-01 | 青岛海尔空调器有限总公司 | Voice control method for air conditioner, and air conditioner |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3610396B1 (en) | Voice identification feature optimization and dynamic registration methods, client, and server | |
| CN1307589C (en) | Method and apparatus of managing information about a person | |
| KR102309031B1 (en) | Apparatus and Method for managing Intelligence Agent Service | |
| CN112269895A (en) | Vibration control method and device and computer readable storage medium | |
| CN107610698A (en) | A kind of method for realizing Voice command, robot and computer-readable recording medium | |
| CN112735439A (en) | Environmentally regulated speaker identification | |
| CN107644638A (en) | Audio recognition method, device, terminal and computer-readable recording medium | |
| CN105378708A (en) | Environmentally aware dialog policies and response generation | |
| CN105511638B (en) | Input method application method and device | |
| CN106649236B (en) | Modify the method and device of prompt | |
| CN109658953A (en) | A kind of vagitus recognition methods, device and equipment | |
| CN106936991A (en) | The method and terminal of a kind of automatic regulating volume | |
| CN111145756A (en) | Voice recognition method and device for voice recognition | |
| CN111883117B (en) | Voice wake-up method and device | |
| CN110298463A (en) | Meeting room preordering method, device, equipment and storage medium based on speech recognition | |
| CN110415691A (en) | Control method, device and computer readable storage medium based on speech recognition | |
| CN108038243A (en) | Music recommends method, apparatus, storage medium and electronic equipment | |
| CN112102833B (en) | Speech recognition method, device, equipment and storage medium | |
| CN106844335A (en) | Natural language processing method and device | |
| KR20190032026A (en) | Method for providing natural language expression and electronic device supporting the same | |
| CN109994106A (en) | A kind of method of speech processing and equipment | |
| CN109147801B (en) | Voice interaction method, system, terminal and storage medium | |
| CN110910898B (en) | Voice information processing method and device | |
| CN113053362A (en) | Method, device, equipment and computer readable medium for speech recognition | |
| US20220276067A1 (en) | Method and apparatus for guiding voice-packet recording function, device and computer storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191105 |
|
| RJ01 | Rejection of invention patent application after publication |