CN110503962A - Speech recognition and setting method, device, computer equipment and storage medium - Google Patents
Speech recognition and setting method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110503962A CN110503962A CN201910739439.9A CN201910739439A CN110503962A CN 110503962 A CN110503962 A CN 110503962A CN 201910739439 A CN201910739439 A CN 201910739439A CN 110503962 A CN110503962 A CN 110503962A
- Authority
- CN
- China
- Prior art keywords
- word
- wake
- order
- order word
- database
- 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
- G10L17/00—Speaker identification or verification
- G10L17/02—Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/06—Decision making techniques; Pattern matching strategies
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/22—Interactive procedures; Man-machine interfaces
- G10L17/24—Interactive procedures; Man-machine interfaces the user being prompted to utter a password or a predefined phrase
Abstract
This application involves a kind of speech recognitions and setting method, device, computer equipment and storage medium.Method includes: that will wake up word sound to carry out feature extraction, obtains waking up word characteristic information；Word characteristic information will be waken up to match with word database is waken up；If successful match, obtains and whether continue matching wake-up word signal；Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；Word signal is waken up if getting stopping and continuing matching, obtains order word sound；Order word sound is subjected to feature extraction, obtains order word characteristic information；Order word characteristic information is matched with order word database；If successful match, obtains and whether continue to match order word signal；Otherwise, output order word it fails to match signal and continue to obtain order word sound；Continue to match order word signal if getting, repeats to obtain order word sound.
Description
Technical field
This application involves intelligent sound technical fields, more particularly to a kind of speech recognition and setting method, device, calculating
Machine equipment and storage medium.
Background technique
With the development of smart home, intelligent electric appliance, there is intelligent sound identification technology.Intelligent sound identification technology is
A technique for digital speech is converted into the text that computer is understood that.In intelligent communication system, intelligent sound interface
Electric appliance is being turned into from a simple service aid " supplier " and life " partner " of a service；Using electric appliance with
Communication network, people can easily inquire from Database Systems by voice command and extract related information.With meter
The miniaturization of calculation machine, keyboard have become the one very big obstacle of mobile platform.Speech recognition just gradually becomes in information technology
The key technology of man-machine interface, speech recognition technology enable people to get rid of keyboard in conjunction with speech synthesis technique, pass through voice
Order is operated.The application of voice technology, which has become one, has emulative emerging high-tech industry.
However, current audio recognition method is will to wake up word and order word is input in the equipment of speech recognition, user
Smart machine is opened by saying wake-up word, by order word come the order smart machine specific works.But traditional language
When voice recognition method needs user in the same space to a progress voice operating in multiple intelligent sound equipment, meeting
Error starting or maloperation other intelligent sound equipment simultaneously.
Summary of the invention
Based on this, it is necessary to need in the same space for user to a progress language in multiple intelligent sound equipment
When sound operates, can error starting or the technical issues of other intelligent sound equipment of maloperation simultaneously, a kind of speech recognition is provided and is set
Determine method, apparatus, computer equipment and storage medium.
A kind of speech recognition and setting method, which comprises
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
The wake-up word characteristic information includes waking up word language information, the wake-up word number in one of the embodiments,
It include waking up word kind database according to library, it is described to wake up the step of word characteristic information is matched with wake-up word database packet
It includes:
The wake-up word language information is matched with the wake-up word kind database, obtains matching result.
The wake-up word characteristic information further includes waking up word text information, the wake-up word in one of the embodiments,
Database further includes waking up word lteral data library, described to wake up the step of word characteristic information is matched with wake-up word database
Include:
The wake-up word text information is matched with wake-up word lteral data library, if the wake-up word text information with
The wake-up word lteral data storehouse matching success and the wake-up word language information and the wake-up word kind database carry out
Successful match then wakes up word characteristic information and wakes up the success of word database matching, otherwise wakes up word characteristic information and wakes up word number
Fail according to storehouse matching.
The order word characteristic information includes order word language information, the order word number in one of the embodiments,
It include order word kind database according to library, the step of order word characteristic information is matched with order word database packet
It includes:
The order word language information is matched with the order word kind database, obtains matching result.
The order word characteristic information further includes order word text information, the order word in one of the embodiments,
Database further includes order word lteral data library, described the step of being matched order word characteristic information with order word database
Include:
The order word text information is matched with the order word lteral data library, if the order word text information with
The order word lteral data storehouse matching success and the order word language information and the order word kind database carry out
Successful match, then order word characteristic information and order word database matching are successful, otherwise order word characteristic information and order word number
Fail according to storehouse matching.
The wake-up word lteral data library is called out including wake-up word acoustic model, multiple preset in one of the embodiments,
Text of waking up and default wake-up phrase；It is described that the wake-up word text information and wake-up word lteral data library progress is matched
Step includes: to be split as the wake-up word text information according to the wake-up word acoustic model to wake up word syllable, according to described
Default wake-up text, which converts each wake-up word syllable to, wakes up word text, is believed according to the wake-up word text received
The sequence of breath is ranked up the wake-up word text to obtain comparison wake-up phrase, and the comparison is waken up phrase and is preset with described
Phrase is waken up to be matched.
The order word lteral data library includes order word acoustic model, multiple default lives in one of the embodiments,
Enable text and pre-set commands phrase；It is described that the order word text information and order word lteral data library progress is matched
Step includes: that the order word text information is split as order word syllable according to the order word acoustic model, according to described
Each order word syllable is converted order word text by pre-set commands text, is believed according to the order word text received
The sequence of breath is ranked up the order word text to obtain comparison order phrase, and the comparison order phrase is preset with described
Order phrase is matched.
A kind of speech recognition and setting device, described device include:
Speech reception module wakes up word sound and order word sound for receiving；
Whether whether command reception module continue matching wake-up word signal for reception and continue to match order word signal；
Characteristic extracting module wakes up word characteristic information for extracting from the wake-up word sound, from the order word sound
In extract order word characteristic information；
Matching module is used for for matching the wake-up word characteristic information with word database is waken up by the order word
Characteristic information is matched with order word database；
Output module, for export wake up word it fails to match signal, wake up word matching end signal, output order word setting terminates
Signal and order word it fails to match signal.
A kind of intelligent sound equipment, including memory, processor and storage can be run on a memory and on a processor
Computer program, the processor performs the steps of when executing the computer program
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
It is performed the steps of when row
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
Above-mentioned speech recognition and setting method, device, computer equipment and storage medium, by obtaining multiple wake-up words
Sound sets multiple and different wake-up words, sets multiple and different order words by obtaining multiple order words.So that user needs
Will in the same space to a progress voice operating in multiple intelligent sound equipment when, other intelligent sounds will not be interfered to set
Standby work.
Detailed description of the invention
Fig. 1 is the applied environment figure of speech recognition and setting method in one embodiment；
Fig. 2 is the flow diagram of speech recognition and setting method in one embodiment；
Fig. 3 is the flow diagram that word characteristic information matching step is waken up in one embodiment；
Fig. 4 is the flow diagram of order word characteristic information matching step in one embodiment；
Fig. 5 is the structural block diagram of speech recognition and setting device in one embodiment；
Fig. 6 is the internal structure chart of computer equipment in one embodiment.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
Speech recognition provided by the present application and setting method can be applied in application environment as shown in Figure 1.Wherein,
Terminal 102 is communicated with server 104 by network by network.Wherein, terminal 102 can be, but not limited to be various intelligence
Speech ciphering equipment, personal computer, laptop, smart phone, tablet computer and portable wearable device, server 104
It can be realized with the server cluster of the either multiple server compositions of independent server.
In one embodiment, as shown in Fig. 2, providing a kind of speech recognition and setting method, it is applied in this way
It is illustrated for terminal in Fig. 1, comprising the following steps:
Step 101: obtaining and wake up word sound.
Wherein, the indicative vocabulary of starting and closing that word is terminal 102 is waken up.Waking up word sound is that the external world is sent to end
The voice about wake-up word at end 102.
Specifically, terminal 102 obtains the extraneous wake-up word sound sended over.It should be noted that the external world sends over
Wake-up word for matching with the wake-up word in database.If the wake-up word waken up in word and database that the external world is sent
With success, that is to say, that the wake-up word that the external world is sent is identical as the wake-up word in database, can find the external world in the database
The wake-up word of transmission, then terminal 102 receives the setting of the wake-up word.If the wake-up waken up in word and database that the external world is sent
Word mismatch, that is to say, that the wake-up word that the external world is sent is different from the wake-up word in database, can not find the external world in the database
The wake-up word of transmission.Then terminal 102 receives the setting failure of the wake-up word.
Step 103: the wake-up word sound being subjected to feature extraction, obtains waking up word characteristic information.
Wherein, wake up word characteristic information refer to can with the wake-up word information in the database of terminal 102 carry out it is matched
Characteristic information.
Specifically, the wake-up word sound that sends over of the external world is carried out feature extraction by terminal 102, obtain can and database
In wake-up word information carry out matched wake-up word characteristic information.It should be noted that the extraneous wake-up word sound sended over
It cannot directly be matched with the wake-up word information in database.It needs to handle by terminal 102 and extracts necessary wake-up
Word characteristic information, by the way that the necessary wake-up word characteristic information extracted is matched with the wake-up word information in database,
Exact matching result can just be obtained.
Step 105: the wake-up word characteristic information is matched with word database is waken up.If successful match obtains
Whether continue matching and wakes up word signal.Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound.If obtaining
Word signal is waken up to matching is continued, then repeats to obtain wake-up word sound.Word signal, output are waken up if getting stopping and continuing matching
It wakes up word and matches end signal, and obtain order word sound.
Wherein, waking up word database is a kind of database, the database refer to containing can with wake up the progress of word characteristic information
Matched wake-up word data, order word sound are the indicative vocabulary for executing other runnings of terminal 102, are extraneous order words
Sound is sent to the voice about order word of terminal 102.
Specifically, the wake-up word sound that sends over of the external world is carried out feature extraction by terminal 102, obtain can and database
In wake-up word data carry out matched wake-up word characteristic information.Again by obtained wake-up word characteristic information and wake-up word database
In wake-up word data matched.That is, comparison wakes up word characteristic information and wakes up the wake-up word number in word database
Whether according to identical, further, search to wake up whether there is and to extract obtained wake-up word characteristic information identical in word database
Wake-up word data.If waking up the identical wake-up word number of wake-up word characteristic information for existing and extracting in word database
According to then terminal 102 receives the setting of the wake-up word.If it is special to wake up the wake-up word for being not present and extracting in word database
Reference ceases identical wake-up word data, then terminal 102 does not receive to receive the setting of the wake-up word.It is called out when terminal 102 receives this
It wakes up after the setting of word, obtains whether the external world continues the order that matching wakes up word.If terminal 102, which receives the external world, continues matching wake-up word
Order then repeat step 101, step 103 and step 105.If terminal 102 receives the extraneous order for stopping matching wake-up word
Terminal 102 terminates to receive the setting for waking up word, and starts to receive the setting of order word.
Step 107: the order word sound being subjected to feature extraction, obtains order word characteristic information.
Wherein, order word characteristic information refer to can with the order word information in the database of terminal 102 carry out it is matched
Characteristic information.
Specifically, the order word sound that sends over of the external world is carried out feature extraction by terminal 102, obtain can and database
In order word information carry out matched order word characteristic information.It should be noted that the extraneous order word sound sended over
It cannot directly be matched with the order word information in database.It needs to handle by terminal 102 and extracts necessary order
Word characteristic information, by the way that the necessary order word characteristic information extracted is matched with the order word information in database,
Exact matching result can just be obtained.
Step 109: the order word characteristic information is matched with order word database.If successful match obtains
Whether continue to match order word signal.Otherwise, output order word it fails to match signal and continue to obtain order word sound.If obtaining
To matching order word signal is continued, then repeat to obtain order word sound.It is defeated if getting stopping to continue to match order word signal
Order word sets end signal out.
Wherein, order word database is a kind of database, which refers to containing can carry out with order word characteristic information
Matched order word data, order word sound are the indicative vocabulary for executing other runnings of terminal 102, are extraneous order words
Sound is sent to the voice about order word of terminal 102.
Specifically, the order word sound that sends over of the external world is carried out feature extraction by terminal 102, obtain can and database
In order word data carry out matched order word characteristic information.Again by obtained order word characteristic information and order word database
In order word data matched.That is, the order word number in comparison order word characteristic information and order word database
According to whether identical, further, whether there is in look-up command word database and to extract obtained order word characteristic information identical
Order word data.If the identical order word number of order word characteristic information for existing and extracting in order word database
According to then terminal 102 receives the setting of the order word.If the order word for being not present and extracting in order word database is special
Reference ceases identical order word data, then terminal 102 does not receive to receive the setting of the order word.When terminal 102 receives the life
After the setting for enabling word, obtain whether the external world continues to match the order of order word.Continue to match order word if terminal 102 receives the external world
Order then repeat step 107 and step 109.102 knot of terminal if terminal 102 receives the extraneous order for stopping matching order word
Beam receives the setting of order word, and starts to receive the setting of order word.
In above-mentioned speech recognition and setting method, terminal 102 is multiple and different to set by obtaining multiple wake-up word sounds
Wake-up word, terminal 102 sets multiple and different order words by obtaining multiple order words.So that user needs in same sky
In to a progresss voice operating in multiple terminals 102 when, the work of other terminals 102 will not be interfered.
The wake-up word characteristic information includes waking up word language information, the wake-up word number in one of the embodiments,
It include waking up word kind database according to library.Step 105: described that wake-up word characteristic information and wake-up word database progress is matched
Step includes step 205 and step 305, in which:
Step 205: the wake-up word language information being matched with the wake-up word kind database, obtains matching result.
Wherein, it wakes up word kind and refers to the languages for waking up word sound, for example, Chinese, English etc..Wake up word kind
Information refers to that the wake-up word sound passes through the languages characteristic information that processing obtains.This is the prior art, and so it will not be repeated.Institute
It states and wakes up the database that word kind database refers to the wake-up word information for being stored with multiple languages in terminal 102.
Specifically, the wake-up word sound that terminal 102 sends over the external world carries out languages feature extraction, obtains languages characteristic
Information, the languages characteristic information can be matched with the wake-up word information of languages in the wake-up word kind database.If
With success, it was demonstrated that the languages for the wake-up word sound that the external world sends over are 102 acceptable languages of terminal.In this way, waking up word
Languages database increases the languages range that terminal 102 can receive, and increases the scope of application of terminal 102.
The wake-up word characteristic information further includes waking up word text information, the wake-up word in one of the embodiments,
Database further includes wake-up word lteral data library, step 105: described to wake up word characteristic information and wake up the progress of word database
With the step of include:
Step 305: the wake-up word text information being matched with wake-up word lteral data library, if the wake-up word is literary
Word information and wake-up word lteral data storehouse matching success and the wake-up word language information and the wake-up word kind number
According to library carry out successful match, then wake up word characteristic information and wake up word database matching success, otherwise wake up word characteristic information with
Wake up the failure of word database matching.
Wherein, word text information is waken up, refers to that the wake-up word sound passes through the text feature information that processing obtains.It calls out
Awake word lteral data library refers to the database for the wake-up word information that multiple texts are stored in terminal 102.
Specifically, the wake-up word sound that sends over of the external world is carried out character features extraction by terminal 102, obtain can and institute
State the matched text feature information of wake-up word information progress for waking up text in word lteral data library.If successful match, it was demonstrated that outer
The text for the wake-up word sound that boundary sends over is 102 acceptable text of terminal.Increase in this way, waking up word kind database
The literal scope that terminal 102 can receive, increases the scope of application of terminal 102.
The order word characteristic information includes order word language information, the order word number in one of the embodiments,
It include order word kind database according to library, step 109: described that order word characteristic information and the progress of order word database is matched
Step includes step 209 and step 309, in which:
Step 209: the order word language information being matched with the order word kind database, obtains matching result.
Wherein, order word kind refers to the languages of the order word sound, for example, multiple languages such as Chinese, English.Life
Word language information is enabled to refer to that the order word sound passes through the languages characteristic information that processing obtains.This is the prior art, therefore not
It repeats again.The order word kind database refers to the database for the order word information that multiple languages are stored in terminal 102.
Specifically, the order word sound that terminal 102 sends over the external world carries out languages feature extraction, obtains languages characteristic
Information, the languages characteristic information can be matched with the order word information of languages in the order word kind database.If
With success, it was demonstrated that the languages for the order word sound that the external world sends over are 102 acceptable languages of terminal.In this way, order word
Languages database increases the languages range that terminal 102 can receive, and increases the scope of application of terminal 102.
The order word characteristic information further includes order word text information, the order word in one of the embodiments,
Database further includes order word lteral data library, described the step of being matched order word characteristic information with order word database
Include:
Step 309: the order word text information being matched with the order word lteral data library, if the order word is literary
Word information and order word lteral data storehouse matching success and the order word language information and the order word kind number
According to library carry out successful match, then order word characteristic information and order word database matching success, otherwise order word characteristic information with
The failure of order word database matching.
Wherein, order word text information refers to that the order word sound passes through the text feature information that processing obtains.Life
Word lteral data library is enabled to refer to the database for the order word information for being stored with multiple texts in terminal 102.
Specifically, the order word sound that sends over of the external world is carried out character features extraction by terminal 102, obtain can and institute
The order word information for stating text in order word lteral data library carries out matched text feature information.If successful match, it was demonstrated that outer
The text for the order word sound that boundary sends over is 102 acceptable text of terminal.In this way, order word kind database increases
The literal scope that terminal 102 can receive, increases the scope of application of terminal 102.
The wake-up word lteral data library is called out including wake-up word acoustic model, multiple preset in one of the embodiments,
Text of waking up and default wake-up phrase；Step 305: by the wake-up word text information and wake-up word lteral data library progress
With the step of include: step 405: according to the wake-up word acoustic model by the wake-up word text information be split as wake up word sound
Section converts each wake-up word syllable to according to the default wake-up text and wakes up word text, according to receiving
The sequence for waking up word text information is ranked up the wake-up word text to obtain comparison wake-up phrase, and the comparison is waken up word
Group is matched with the default wake-up phrase.
Wherein, it wakes up word acoustic model and refers to that can will wake up word text information is split as waking up the acoustic mode of word syllable
Type, in the present embodiment, wake-up word acoustic model are hidden Markov model.The default text that wakes up is the default wake-up phrase of composition
Text unit.The default phrase that wakes up is to wake up phrase with comparison to carry out matched wake-up word information.In this way, waking up word acoustic mode
Type improves terminal 102 and receives the precision for waking up word setting.
The order word lteral data library includes order word acoustic model, multiple default lives in one of the embodiments,
Enable text and pre-set commands phrase；Step 309: it is described by the order word text information and the order word lteral data library into
The step of row matching includes: step 409: the order word text information being split as order according to the order word acoustic model
Word syllable converts order word text for each order word syllable according to the pre-set commands text, according to what is received
The sequence of the order word text information is ranked up the order word text to obtain comparison order phrase, and the comparison is ordered
Phrase is enabled to be matched with the pre-set commands phrase.
Wherein, order word acoustic model refers to the acoustic mode that order word text information can be split as to order word syllable
Type, in the present embodiment, order word acoustic model are hidden Markov model.Pre-set commands text is composition pre-set commands phrase
Text unit.Pre-set commands phrase is to carry out matched order word information with comparison order phrase.In this way, order word acoustic mode
Type improves the precision that terminal 102 receives the setting of order word.
It should be understood that although each step in the flow chart of Fig. 2 to Fig. 4 is successively shown according to the instruction of arrow,
But these steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly state otherwise herein, these
There is no stringent sequences to limit for the execution of step, these steps can execute in other order.Moreover, Fig. 2 is into Fig. 4
At least part step may include that perhaps these sub-steps of multiple stages or stage are not necessarily same to multiple sub-steps
One moment executed completion, but can execute at different times, and the execution in these sub-steps or stage sequence is also not necessarily
Be successively carry out, but can at least part of the sub-step or stage of other steps or other steps in turn or
Alternately execute.
In one embodiment, as shown in figure 5, providing a kind of speech recognition and setting device, comprising: phonetic incepting mould
Block 501, command reception module 503, characteristic extracting module 505, matching module 507 and output module 509, in which:
Speech reception module 501 wakes up word sound and order word sound for receiving.
Whether command reception module 503 continues matching wake-up word signal and whether continues matching order word to believe for receiving
Number.
Characteristic extracting module 505 wakes up word characteristic information for extracting from the wake-up word sound, specifically, special
Sign extraction module 505 wakes up word language information and wake-up word text information for extracting from the wake-up word sound.From institute
It states and extracts order word characteristic information in order word sound, specifically, characteristic extracting module 505 is used for from the order word sound
In extract order word language information and order word text information.
Matching module 507, for the wake-up word characteristic information to be matched with word database is waken up, specifically,
It is matched for word language information will to be waken up with word kind database is waken up with module 507.Matching module 507 will be for that will wake up
Word text information is matched with word lteral data library is waken up, and further, matching module 507 is used for the order word feature
Information is matched with order word database.Specifically, according to the wake-up word acoustic model by the wake-up word text information
It is split as waking up word syllable, is converted each wake-up word syllable to according to the default wake-up text and wake up word text, root
The wake-up word text is ranked up to obtain comparison wake-up phrase according to the sequence of the wake-up word text information received, it will
The comparison wakes up phrase and is matched with the default wake-up phrase.Matching module 507 be used for by order word language information with
Order word kind database is matched.Matching module 507 be used for by order word text information and order word lteral data library into
Row matching, further, matching module 507 is for splitting the order word text information according to the order word acoustic model
For order word syllable, order word text is converted for each order word syllable according to the pre-set commands text, according to connecing
The sequence of the order word text information received is ranked up the order word text to obtain comparison order phrase, will be described
Comparison order phrase is matched with the pre-set commands phrase.
Output module 509 wakes up word it fails to match signal, wakes up word matching end signal, output order word for exporting
Set end signal and order word it fails to match signal.
Specific restriction about speech recognition and setting device may refer to above for speech recognition and setting method
Restriction, details are not described herein.Modules in above-mentioned speech recognition and setting device can be fully or partially through software, hard
Part and combinations thereof is realized.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment,
It can also be stored in a software form in the memory in computer equipment, execute the above modules in order to which processor calls
Corresponding operation.
In one embodiment, a kind of computer equipment is provided, which can be terminal, internal structure
Figure can be as shown in Figure 6.The computer equipment includes processor, the memory, network interface, display connected by system bus
Screen and input unit.Wherein, the processor of the computer equipment is for providing calculating and control ability.The computer equipment is deposited
Reservoir includes non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system and computer journey
Sequence.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The network interface of machine equipment is used to communicate with external terminal by network connection.When the computer program is executed by processor with
Realize a kind of speech recognition and setting method.The display screen of the computer equipment can be liquid crystal display or electric ink is aobvious
Display screen, the input unit of the computer equipment can be the touch layer covered on display screen, be also possible to computer equipment shell
Key, trace ball or the Trackpad of upper setting can also be external keyboard, Trackpad or mouse etc..
It will be understood by those skilled in the art that structure shown in Fig. 6, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment is provided, including memory, processor and storage are on a memory
And the computer program that can be run on a processor, processor perform the steps of when executing computer program
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
In one embodiment, the wake-up word feature letter is also performed the steps of when processor executes computer program
Breath includes waking up word language information, and the wake-up word database includes waking up word kind database, described to wake up word feature letter
The step of breath is matched with wake-up word database includes: by the wake-up word language information and the wake-up word kind database
It is matched, obtains matching result.
In one embodiment, the wake-up word feature letter is also performed the steps of when processor executes computer program
Breath further includes waking up word text information, and the wake-up word database further includes waking up word lteral data library, described to wake up word spy
The step of reference breath is matched with wake-up word database includes: by the wake-up word text information and the wake-up word text number
It is matched according to library, if the wake-up word text information and wake-up word lteral data storehouse matching success and the wake-up word
Language information and the wake-up word kind database carry out successful match, then wake up word characteristic information and wake up word database matching
Otherwise success wakes up word characteristic information and wakes up the failure of word database matching.
In one embodiment, the order word feature letter is also performed the steps of when processor executes computer program
Breath includes order word language information, and the order word database includes order word kind database, described to believe order word feature
The step of breath is matched with order word database includes: by the order word language information and the order word kind database
It is matched, obtains matching result.
In one embodiment, the order word feature letter is also performed the steps of when processor executes computer program
Breath further includes order word text information, and the order word database further includes order word lteral data library, described that order word is special
The step of reference breath is matched with order word database includes: by the order word text information and the order word text number
It is matched according to library, if the order word text information and order word lteral data storehouse matching success and the order word
Language information and the order word kind database carry out successful match, then order word characteristic information and order word database matching
Success, otherwise order word characteristic information and order word database matching fail.
In one embodiment, the wake-up word text number is also performed the steps of when processor executes computer program
It include waking up word acoustic model, multiple default wake-up texts and default wake-up phrase according to library；It is described to believe the wake-up word text
Breath the step of being matched with wake-up word lteral data library includes: according to the wake-up word acoustic model by the wake-up word
Text information is split as waking up word syllable, converts wake-up word for each wake-up word syllable according to the default wake-up text
Text is ranked up the wake-up word text according to the sequence of the wake-up word text information received to obtain comparison wake-up
The comparison is waken up phrase and matched with the default wake-up phrase by phrase.
In one embodiment, the order word text number is also performed the steps of when processor executes computer program
It include order word acoustic model, multiple pre-set commands texts and pre-set commands phrase according to library；It is described to believe the order word text
Breath the step of being matched with the order word lteral data library includes: according to the order word acoustic model by the order word
Text information is split as order word syllable, converts order word for each order word syllable according to the pre-set commands text
Text is ranked up the order word text according to the sequence of the order word text information received to obtain comparison order
Phrase matches the comparison order phrase with the pre-set commands phrase.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program performs the steps of when being executed by processor
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
In one embodiment, the wake-up word feature letter is also performed the steps of when processor executes computer program
Breath includes waking up word language information, and the wake-up word database includes waking up word kind database, described to wake up word feature letter
The step of breath is matched with wake-up word database includes: by the wake-up word language information and the wake-up word kind database
It is matched, obtains matching result.
In one embodiment, the wake-up word feature letter is also performed the steps of when processor executes computer program
Breath further includes waking up word text information, and the wake-up word database further includes waking up word lteral data library, described to wake up word spy
The step of reference breath is matched with wake-up word database includes: by the wake-up word text information and the wake-up word text number
It is matched according to library, if the wake-up word text information and wake-up word lteral data storehouse matching success and the wake-up word
Language information and the wake-up word kind database carry out successful match, then wake up word characteristic information and wake up word database matching
Otherwise success wakes up word characteristic information and wakes up the failure of word database matching.
In one embodiment, the order word feature letter is also performed the steps of when processor executes computer program
Breath includes order word language information, and the order word database includes order word kind database, described to believe order word feature
The step of breath is matched with order word database includes: by the order word language information and the order word kind database
It is matched, obtains matching result.
In one embodiment, the order word feature letter is also performed the steps of when processor executes computer program
Breath further includes order word text information, and the order word database further includes order word lteral data library, described that order word is special
The step of reference breath is matched with order word database includes: by the order word text information and the order word text number
It is matched according to library, if the order word text information and order word lteral data storehouse matching success and the order word
Language information and the order word kind database carry out successful match, then order word characteristic information and order word database matching
Success, otherwise order word characteristic information and order word database matching fail.
In one embodiment, the wake-up word text number is also performed the steps of when processor executes computer program
It include waking up word acoustic model, multiple default wake-up texts and default wake-up phrase according to library；It is described to believe the wake-up word text
Breath the step of being matched with wake-up word lteral data library includes: according to the wake-up word acoustic model by the wake-up word
Text information is split as waking up word syllable, converts wake-up word for each wake-up word syllable according to the default wake-up text
Text is ranked up the wake-up word text according to the sequence of the wake-up word text information received to obtain comparison wake-up
The comparison is waken up phrase and matched with the default wake-up phrase by phrase.
In one embodiment, the order word text number is also performed the steps of when processor executes computer program
It include order word acoustic model, multiple pre-set commands texts and pre-set commands phrase according to library；It is described to believe the order word text
Breath the step of being matched with the order word lteral data library includes: according to the order word acoustic model by the order word
Text information is split as order word syllable, converts order word for each order word syllable according to the pre-set commands text
Text is ranked up the order word text according to the sequence of the order word text information received to obtain comparison order
Phrase matches the comparison order phrase with the pre-set commands phrase.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM(EPROM), electrically erasable ROM(EEPROM) or flash memory.Volatile memory may include
Random-access memory (ram) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM(SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM(ESDRAM), synchronization link (Synchlink) DRAM(SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of above embodiments can be combined arbitrarily, for simplicity of description, not to above-described embodiment
In each technical characteristic it is all possible combination be all described, as long as however, the combination of these technical characteristics be not present lance
Shield all should be considered as described in this specification.
The several embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the concept of this application, various modifications and improvements can be made, these belong to the protection of the application
Range.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of speech recognition and setting method, which comprises
It obtains and wakes up word sound；
The wake-up word sound is subjected to feature extraction, obtains waking up word characteristic information；
The wake-up word characteristic information is matched with word database is waken up；
If successful match, obtains and whether continue matching wake-up word signal；
Otherwise, output, which wakes up word it fails to match signal and continues to obtain, wakes up word sound；
Word signal is waken up if getting and continuing matching, repeats to obtain wake-up word sound；
Word signal is waken up if getting stopping and continuing matching, output wakes up word and matches end signal, and obtains order word sound；
The order word sound is subjected to feature extraction, obtains order word characteristic information；
The order word characteristic information is matched with order word database；
If successful match, obtains and whether continue to match order word signal；
Otherwise, output order word it fails to match signal and continue to obtain order word sound；
Continue to match order word signal if getting, repeats to obtain order word sound；
Word signal is waken up if getting stopping and continuing matching, exports order word setting end signal.
2. the method according to claim 1, wherein the wake-up word characteristic information includes waking up word kind letter
Breath, the wake-ups word database include wake up word kind database, it is described will wake up word characteristic information with wake-up word database into
Row matching the step of include:
The wake-up word language information is matched with the wake-up word kind database, obtains matching result.
3. according to the method described in claim 2, it is characterized in that, the wake-up word characteristic information further includes waking up word text letter
Breath, the wake-up word database further include waking up word lteral data library, described to wake up word characteristic information and wake up word database
The step of being matched include:
The wake-up word text information is matched with wake-up word lteral data library, if the wake-up word text information with
The wake-up word lteral data storehouse matching success and the wake-up word language information and the wake-up word kind database carry out
Successful match then wakes up word characteristic information and wakes up the success of word database matching, otherwise wakes up word characteristic information and wakes up word number
Fail according to storehouse matching.
4. the method according to claim 1, wherein the order word characteristic information includes order word kind letter
Breath, the order word database includes order word kind database, it is described by order word characteristic information and order word database into
Row matching the step of include:
The order word language information is matched with the order word kind database, obtains matching result.
5. according to the method described in claim 4, it is characterized in that, the order word characteristic information further includes order word text letter
Breath, the order word database further includes order word lteral data library, described by order word characteristic information and order word database
The step of being matched include:
The order word text information is matched with the order word lteral data library, if the order word text information with
The order word lteral data storehouse matching success and the order word language information and the order word kind database carry out
Successful match, then order word characteristic information and order word database matching are successful, otherwise order word characteristic information and order word number
Fail according to storehouse matching.
6. according to the method described in claim 3, it is characterized in that, the wake-up word lteral data library includes waking up word acoustic mode
Type, multiple default wake-up texts and default wake-up phrase；It is described by the wake-up word text information and the wake-up word text number
The step of being matched according to library includes: to be split as the wake-up word text information according to the wake-up word acoustic model to wake up word
Syllable converts each wake-up word syllable to according to the default wake-up text and wakes up word text, according to the institute received
The sequence for stating wake-up word text information is ranked up the wake-up word text to obtain comparison wake-up phrase, and the comparison is waken up
Phrase is matched with the default wake-up phrase.
7. according to the method described in claim 5, it is characterized in that, the order word lteral data library includes order word acoustic mode
Type, multiple pre-set commands texts and pre-set commands phrase；It is described by the order word text information and the order word text number
The step of being matched according to library includes: that the order word text information is split as order word according to the order word acoustic model
Syllable converts order word text for each order word syllable according to the pre-set commands text, according to the institute received
The sequence for stating order word text information is ranked up the order word text to obtain comparison order phrase, and the comparison is ordered
Phrase is matched with the pre-set commands phrase.
8. a kind of speech recognition and setting device, which is characterized in that described device includes:
Speech reception module wakes up word sound and order word sound for receiving；
Whether whether command reception module continue matching wake-up word signal for reception and continue to match order word signal；
Characteristic extracting module wakes up word characteristic information for extracting from the wake-up word sound, from the order word sound
In extract order word characteristic information；
Matching module is used for for matching the wake-up word characteristic information with word database is waken up by the order word
Characteristic information is matched with order word database；
Output module, for export wake up word it fails to match signal, wake up word matching end signal, output order word setting terminates
Signal and order word it fails to match signal.
9. a kind of intelligent sound equipment, can run on a memory and on a processor including memory, processor and storage
Computer program, which is characterized in that the processor realizes any one of claims 1 to 7 when executing the computer program
The step of the method.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 7 is realized when being executed by processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910739439.9A CN110503962A (en) | 2019-08-12 | 2019-08-12 | Speech recognition and setting method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910739439.9A CN110503962A (en) | 2019-08-12 | 2019-08-12 | Speech recognition and setting method, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110503962A true CN110503962A (en) | 2019-11-26 |
Family
ID=68587230
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910739439.9A Pending CN110503962A (en) | 2019-08-12 | 2019-08-12 | Speech recognition and setting method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110503962A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111312250A (en) * | 2020-02-21 | 2020-06-19 | 珠海荣邦电子科技有限公司 | Voice-based multi-device adaptation control method, device and system |
| CN112164388A (en) * | 2020-11-05 | 2021-01-01 | 佛山市顺德区美的电子科技有限公司 | Voice equipment and awakening method and device thereof and storage medium |
| CN113593554A (en) * | 2021-07-21 | 2021-11-02 | 深圳市芯中芯科技有限公司 | Voice recognition offline command word awakening application method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104660792A (en) * | 2013-11-21 | 2015-05-27 | 腾讯科技（深圳）有限公司 | Method and device for awakening applications |
| CN108335695A (en) * | 2017-06-27 | 2018-07-27 | 腾讯科技（深圳）有限公司 | Sound control method, device, computer equipment and storage medium |
| CN109949808A (en) * | 2019-03-15 | 2019-06-28 | 上海华镇电子科技有限公司 | The speech recognition appliance control system and method for compatible mandarin and dialect |
-
2019
- 2019-08-12 CN CN201910739439.9A patent/CN110503962A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104660792A (en) * | 2013-11-21 | 2015-05-27 | 腾讯科技（深圳）有限公司 | Method and device for awakening applications |
| CN108335695A (en) * | 2017-06-27 | 2018-07-27 | 腾讯科技（深圳）有限公司 | Sound control method, device, computer equipment and storage medium |
| CN109949808A (en) * | 2019-03-15 | 2019-06-28 | 上海华镇电子科技有限公司 | The speech recognition appliance control system and method for compatible mandarin and dialect |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111312250A (en) * | 2020-02-21 | 2020-06-19 | 珠海荣邦电子科技有限公司 | Voice-based multi-device adaptation control method, device and system |
| CN112164388A (en) * | 2020-11-05 | 2021-01-01 | 佛山市顺德区美的电子科技有限公司 | Voice equipment and awakening method and device thereof and storage medium |
| CN113593554A (en) * | 2021-07-21 | 2021-11-02 | 深圳市芯中芯科技有限公司 | Voice recognition offline command word awakening application method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11727914B2 (en) | Intent recognition and emotional text-to-speech learning | |
| CN111627418B (en) | Training method, synthesizing method, system, device and medium for speech synthesis model | |
| WO2021051544A1 (en) | Voice recognition method and device | |
| CN102543071B (en) | Voice recognition system and method used for mobile equipment | |
| WO2020073530A1 (en) | Customer service robot session text classification method and apparatus, and electronic device and computer-readable storage medium | |
| CN106981290B (en) | Voice control device and voice control method | |
| CN113327609B (en) | Method and apparatus for speech recognition | |
| CN107134279A (en) | A kind of voice awakening method, device, terminal and storage medium | |
| CN110503962A (en) | Speech recognition and setting method, device, computer equipment and storage medium | |
| CN111081217B (en) | Voice wake-up method and device, electronic equipment and storage medium | |
| CN110910903B (en) | Speech emotion recognition method, device, equipment and computer readable storage medium | |
| WO2020024620A1 (en) | Voice information processing method and device, apparatus, and storage medium | |
| CN111833845A (en) | Multi-language speech recognition model training method, device, equipment and storage medium | |
| CN110047484A (en) | A kind of speech recognition exchange method, system, equipment and storage medium | |
| CN110992955A (en) | Voice operation method, device, equipment and storage medium of intelligent equipment | |
| CN109074804A (en) | Voice recognition processing method, electronic equipment and storage medium based on accent | |
| CN103514882A (en) | Voice identification method and system | |
| CN112669842A (en) | Man-machine conversation control method, device, computer equipment and storage medium | |
| CN108899028A (en) | Voice awakening method, searching method, device and terminal | |
| CN105353957A (en) | Information display method and terminal | |
| CN111192586A (en) | Voice recognition method and device, electronic equipment and storage medium | |
| CN114242093A (en) | Voice tone conversion method and device, computer equipment and storage medium | |
| CN113611316A (en) | Man-machine interaction method, device, equipment and storage medium | |
| WO2020073839A1 (en) | Voice wake-up method, apparatus and system, and electronic device | |
| CN115064160B (en) | Voice wake-up method and device |
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: 20191126 |
|
| RJ01 | Rejection of invention patent application after publication |