CN110503938A - The recognition methods of machine conversational language and device, identification engine switching method and device - Google Patents
The recognition methods of machine conversational language and device, identification engine switching method and device Download PDFInfo
- Publication number
- CN110503938A CN110503938A CN201910811586.2A CN201910811586A CN110503938A CN 110503938 A CN110503938 A CN 110503938A CN 201910811586 A CN201910811586 A CN 201910811586A CN 110503938 A CN110503938 A CN 110503938A
- Authority
- CN
- China
- Prior art keywords
- language
- category
- classifier
- identification
- voice messaging
- 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
- 238000000034 method Methods 0.000 title claims abstract description 70
- 230000000694 effects Effects 0.000 abstract description 3
- 230000003993 interaction Effects 0.000 abstract description 3
- 238000012360 testing method Methods 0.000 abstract description 3
- 238000004590 computer program Methods 0.000 description 22
- 230000006870 function Effects 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 241001672694 Citrus reticulata Species 0.000 description 7
- 238000012545 processing Methods 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 239000000284 extract Substances 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 239000000203 mixture Substances 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000000429 assembly Methods 0.000 description 1
- 230000000712 assembly Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000035935 pregnancy Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/005—Language recognition
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Machine Translation (AREA)
Abstract
The present invention provides a kind of machine conversational language recognition methods and devices, identification engine switching method and device.The machine conversational language recognition methods includes obtaining guidance voice messaging;According to the guidance voice messaging, the classifier feature of the guidance voice messaging is extracted;Obtain category of language classifier;According to the classifier feature, category of language identification is carried out to the guidance voice messaging by the category of language classifier.Machine conversational language recognition methods through the invention, can detect automatically the category of language of client, to allow speech robot people that can automatically switch language identification engine, to reach better interaction effect according to the testing result of this method.
Description
Technical field
The present invention relates to technical field of language recognition, in particular to a kind of machine conversational language recognition methods, machine dialogue
Speech recognition equipment, machine conversational language identification engine switching method and machine conversational language identify engine switching device.
Background technique
It is the dialect that can not detect the different zones of various foreign languages or various countries automatically in the speech robot people of the prior art
's.The corresponding speech recognition engine of this speech robot people can only be set when speech robot people uses.
And in practical applications, it can not go to be set according to the actual situation.By taking dialect as an example, in south such as Guangdong, Zhejiang
Square developed regions say that the people of dialect and the people to speak Mandarin mix residence or even quite a few people and will not speak Mandarin.When by dialect
People when being exchanged with speech robot people, if speech robot people still uses mandarin pronunciation to identify engine, basic nothing
Method is normally exchanged with user, this brings very big difficulty to the practical application of speech robot people.
Thus, it is desirable to have a kind of technical solution overcomes or at least mitigates at least one drawbacks described above of the prior art.
Summary of the invention
It is an object of that present invention to provide a kind of machine conversational language recognition methods and devices, identification engine switching method and dress
It sets to overcome or at least mitigate at least one drawbacks described above of the prior art.
To achieve the above object, the present invention provides a kind of machine conversational language recognition methods, and the machine conversational language is known
Other method includes: to obtain guidance voice messaging;According to the guidance voice messaging, the classifier of the guidance voice messaging is extracted
Use feature;Obtain category of language classifier;According to the classifier feature, drawn by the category of language classifier to described
It leads voice messaging and carries out category of language identification.
Optionally, include: before the acquisition guides voice messaging
Export leading question;
The acquisition guidance voice messaging includes providing user according to the leading question to have corresponding close with leading question
The guidance voice messaging of system.
Optionally, described according to the guidance voice messaging, extract the classifier feature packet of the guidance voice messaging
It includes:
The guidance voice is converted into guidance text information;
The character features in the guidance text information are extracted as classifier feature.
Optionally, described according to the classifier feature, by the category of language classifier to the guidance voice
Information carries out category of language identification
The classifier is input to the category of language classifier with feature;
The identification label for obtaining the category of language classifier output, determines the language to be identified according to the identification label
Say category of language corresponding to information, wherein at least one described identification label represents the guidance voice messaging and exists
There is no corresponding category of language in the category of language classifier, which is known as Unknown Label.
Optionally, when the identification label of the category of language classifier of acquisition is Unknown Label, the guidance is stored
Voice messaging.
Optionally, the machine conversational language recognition methods includes:
After the storage language category information to be identified, identify that the language category information institute to be identified stored is right
The category of language answered, and using the corresponding category of language as category of language to be learned.
Optionally, the machine conversational language recognition methods includes:
Identify category of language corresponding to the language category information to be identified stored as language to be learned described
After saying type, the category of language classifier is updated according to the category of language to be learned.
Optionally, the update category of language classifier includes:
Obtain the language feature library of category of language to be learned;
The language classification device is updated according to the language feature library of the category of language to be learned, so that it includes described for being formed
The language classification device of category of language to be learned.
Optionally, described according to the guidance voice messaging, extract the classifier feature packet of the guidance voice messaging
It includes:
According to the guidance voice, the audio frequency characteristics of the guidance voice are obtained as the classifier feature.
The present invention also provides a kind of machine conversational language identification device, the machine conversational language identification device includes:
Voice is guided to obtain module, the guidance voice obtains module for obtaining guidance voice messaging;
Classifier characteristic extracting module, the classifier characteristic extracting module is for extracting the guidance voice messaging
Classifier feature;
Classifier obtains module, and the classifier obtains module for obtaining category of language classifier;
Corresponding category of language identification module, the corresponding category of language identification module are used for special according to the classifier
Sign carries out category of language identification to the guidance voice messaging by the category of language classifier.
Optionally, the machine conversational language identification device includes:
Leading question output module, the leading question output module is for exporting leading question;
It includes providing user according to the leading question that the guidance voice, which obtains the guidance voice messaging that module obtains,
There is the guidance voice messaging of corresponding relationship with leading question.
Optionally, the classifier includes: with characteristic extracting module
Conversion module, the conversion module are used to for the guidance voice to be converted into guidance text information;
Character features extraction module, the character features extraction module are used to extract the text in the guidance text information
Feature is as classifier feature.
Optionally, the corresponding category of language identification module includes:
Input module, the input module are used to the classifier being input to the category of language classifier with feature;
Identify that label acquisition module, the identification label acquisition module are used to obtain the category of language classifier output
It identifies label, category of language corresponding to the language category information to be identified is determined according to the identification label, wherein at least
Have the identification label to represent the guidance voice messaging does not have corresponding language kind in the category of language classifier
Class, the label are known as Unknown Label.
Optionally, the machine conversational language identification device includes:
Memory module, the memory module are used to when the identification label of the category of language classifier obtained be unknown mark
When label, the guidance voice messaging is stored.
Optionally, the machine conversational language identification device includes:
Category of language identification module to be learned, the category of language identification module to be learned are stored described for identification
Category of language corresponding to voice messaging is guided, and using the corresponding category of language as category of language to be learned.
Optionally, the machine conversational language identification device includes:
Classifier update module, the classifier update module are used to update institute's predicate according to the category of language to be learned
Say type classification device.
Optionally, the classifier update module includes:
Language feature library obtains module, and the language feature library obtains the language that module is used to obtain category of language to be learned
Feature database;
Training update module, the trained update module are used for according to the language feature library of the category of language to be learned more
The new language classification device, to form the language classification device including the category of language to be learned.
The present invention also provides a kind of electronic equipment, including memory, processor and storage are in the memory simultaneously
The computer program that can be run on the processor, which is characterized in that when the processor executes the computer program
Realize machine conversational language recognition methods as described above.
The present invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage has calculating
Machine program, which is characterized in that the computer program can be realized machine conversational language as described above when being executed by processor
Recognition methods.
The present invention also provides a kind of machine conversational languages to identify engine switching method, and the machine conversational language identification is drawn
Holding up switching method includes:
According to category of language used in machine conversational language recognition methods as described above identification user;
Switch machine for identification user's voice language identification engine be the category of language corresponding to language know
Other engine.
The present invention also provides a kind of machine conversational languages to identify engine switching device, and the machine conversational language identification is drawn
Holding up switching device includes: machine conversational language identification device, machine conversational language identification device, the machine conversational language identification
Device is machine conversational language identification device as described above, for being known using machine conversational language recognition methods as described above
Category of language is not corresponded to;
Switching module, the language identification engine that the switching module is used to switch machine user's voice for identification is institute
State language identification engine corresponding to category of language.
The present invention also provides a kind of electronic equipment, including memory, processor and storage are in the memory simultaneously
The computer program that can be run on the processor, which is characterized in that when the processor executes the computer program
Realize machine conversational language identification engine switching method as described above.
The present invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage has calculating
Machine program, which is characterized in that the computer program can be realized machine conversational language as described above when being executed by processor
Identify engine switching method.
Machine conversational language recognition methods through the invention, can detect automatically the category of language of client, to allow language
Sound robot can automatically switch language identification engine, to reach better interaction effect according to the testing result of this method.
Detailed description of the invention
Fig. 1 is the flow diagram of the machine conversational language recognition methods of one embodiment of the invention.
Fig. 2 is the structural schematic diagram of the machine conversational language identification device in one embodiment of the invention.
Fig. 3 is the composition schematic diagram using the electronic equipment of machine conversational language recognition methods of the invention.
Specific embodiment
To keep the purposes, technical schemes and advantages of the invention implemented clearer, below in conjunction in the embodiment of the present invention
Attached drawing, technical solution in the embodiment of the present invention is further described in more detail.In the accompanying drawings, identical from beginning to end or class
As label indicate same or similar element or element with the same or similar functions.Described embodiment is the present invention
A part of the embodiment, instead of all the embodiments.The embodiments described below with reference to the accompanying drawings are exemplary, it is intended to use
It is of the invention in explaining, and be not considered as limiting the invention.Based on the embodiments of the present invention, ordinary skill people
Member's every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.Under
Face is described in detail the embodiment of the present invention in conjunction with attached drawing.
It should be noted that in the description of the present invention, term " first ", " second " are used for description purposes only, and cannot
It is interpreted as indication or suggestion relative importance.
Machine conversational language recognition methods as shown in Figure 1 includes:
Step 101: obtaining guidance voice messaging;
Step 102: according to the guidance voice messaging, extracting the classifier feature of the guidance voice messaging;
Step 103: obtaining category of language classifier;
Step 104: according to the classifier feature, by the category of language classifier to the guidance voice messaging
Carry out category of language identification.
Machine conversational language recognition methods through the invention, can detect automatically the category of language of client, to allow language
Sound robot can automatically switch language identification engine, to reach better interaction effect according to the testing result of this method.
In the present embodiment, further comprising before obtaining guidance voice messaging includes:
Previous step 1: output leading question;
Acquired guidance voice messaging includes providing user according to leading question to have drawing for corresponding relationship with leading question
Lead voice messaging.
For example, when user and machine carry out interactive conversation, machine exports leading question first, for example, hello, or
Person be you be where this kind of leading question.
At this point, user can engage in the dialogue according to the leading question, for example, user can answer, hello or I is Hubei
People etc..
According to the answer of user, guidance voice messaging, i.e., the guidance voice messaging in above-described embodiment are as follows: hello are obtained
Or I is the audio-frequency information of people from Hubei.
In the present embodiment, step 102: according to guidance voice messaging, extracting the classifier feature of guidance voice messaging
Include:
Step 1021: guidance voice is converted into guidance text information;
Step 1022: extracting the character features in guidance text information as classifier feature
For example, character features can be the key word library of mandarin identification, mode according to keywords is retrieved, and calculates phase
Like degree, for example, in Guangdong language: feeding that hello, following content may be identified as in mandarin pronunciation identification engine: pregnancy
After, Wang Lihong, I one blush, feed hello, just in case the inside, for texts such as reasons, at this point, these above-mentioned texts can be switched to spell
Sound, in existing characteristics library, when in actual use, a new client is said with Guangdong language again: feeding that hello, at this moment new user can be said
Words, switch to phonetic, then see whether the phonetic in feature database is included in new phonetic.If comprising, then it is assumed that matching.
By taking the guidance voice messaging that above-mentioned user provides is " I is Hubei people " as an example, when getting " I is Hubei people "
Audio-frequency information after, by the guidance voice (audio-frequency information of " I is people from Hubei ") be converted into guidance text information (" I is Hubei
The text information of people "), the character features in guidance text information are extracted as classifier feature (i.e. category of language classifier
Input feature value).
In the present embodiment, step 104: according to classifier feature, by category of language classifier to leading question message
Breath carries out category of language identification
Step 1041: classifier is input to category of language classifier with feature;
Step 1042: obtaining the identification label of category of language classifier output, language to be identified is determined according to identification label
Category of language corresponding to information, wherein at least one identification label represents guidance voice messaging in category of language point
There is no corresponding category of language in class device, which is known as Unknown Label.
It is understood that in an alternative embodiment, being also possible to no Unknown Label.For example, it is assumed that can unite
All dialect types in the world Ji Chu, at this point, Unknown Label can be not provided with.
In the present embodiment, machine conversational language recognition methods further comprises:
Step 105: when the identification label of institute's category of language classifier of acquisition is Unknown Label, storing leading question message
Breath.
It is understood that in the present embodiment, the guidance voice messaging of storage be include client provide guidance voice,
It also include the guidance text information after voice will be guided to convert.It is understood that can also include classifier feature.
In the present embodiment, machine conversational language recognition methods further comprises:
Step 106: after storing language category information to be identified, identifying that the language category information to be identified institute stored is right
The category of language answered, and using the corresponding category of language as category of language to be learned.
For example, it can be used manual identified, i.e. human ear listens the guidance voice messaging of storage, guidance text from the point of view of people
Information, according to guidance voice messaging and guidance text information come language kind corresponding to the artificial judgment guidance voice messaging
Class.
It is understood that can also be judged by way of artificial intelligence, for example, will guidance voice messaging input
In the classifier that can carry out phonological detection to other, automatic identification is carried out.
In the present embodiment, machine conversational language recognition methods further comprises:
Step 107: the category of language corresponding to the language category information to be identified that identification is stored is as language to be learned
After saying type, category of language classifier is updated according to category of language to be learned.
For example, when carrying out the identification of machine conversational language next time, in order to allow machine recognition to go out specifically without energy
The category of language to be learned enough identified needs to be updated category of language classifier, so that updated category of language be allowed to divide
Class device can recognize that the category of language to be learned not identified specifically when identifying next time.
In the present embodiment, step 107: updating category of language classifier further comprises:
Step 1071: obtaining the language feature library of category of language to be learned;
Step 1072: according to the language feature library of category of language to be learned update language classification device, thus formed include to
Learn the language classification device of category of language.
In an alternative embodiment, step 102: according to guidance voice messaging, extracting the classifier of guidance voice messaging
Include: with feature
According to guidance voice, the audio frequency characteristics of guidance voice are obtained as classifier feature.
In this embodiment, it is identified by audio frequency characteristics, for example, passing through the features such as the frequency of audio.
For example, it can be carried out using following algorithm:
1, feature audio deposits its corresponding mfcc (mel cepstrum) in the database in advance.
2, the mfcc of audio user is calculated
3, by dtw algorithm (dynamic time adjustment algorithm), the distance of two audios is calculated
4, think to match less than the audio of threshold values.
The present invention is further elaborated by way of example below.It is understood that the citing is not constituted pair
Any restrictions of the invention.
Previous step 1: output leading question, for example, the leading question of output are as follows: may I ask where you are
Step 101: obtaining guidance voice messaging;Specifically, guidance voice messaging, example are obtained according to the answer of user
Such as, user makes the answer that content is " I is Shanghai people " using dialect.At this point, the guidance voice messaging obtained is for example are as follows:
I is that Shanghai is greasy.
Step 102: according to the guidance voice messaging, extracting the classifier feature of the guidance voice messaging;Specifically
Ground, in the present embodiment, step 1021: guidance voice is converted into guidance text information;It is converted into word segment: A Lashi
Shanghai is greasy;
Step 1022: it extracts and guides the character features in text information as classifier feature, in one embodiment,
Extracting me is the greasy phonetic in Shanghai, i.e. " a la shi shang hai ni ", and phonetic is converted to classifier feature,
That is one group of feature vector.
Step 103: obtaining category of language classifier, it is to be understood that the category of language classifier is after training
Category of language classifier;
Step 104: according to the classifier feature, by the category of language classifier to the guidance voice messaging
Progress category of language identification, specific address, in the present embodiment, step 1041: classifier is input to category of language with feature
Classifier;
Step 1042: obtaining the identification label of category of language classifier output, language to be identified is determined according to identification label
Category of language corresponding to information.
When having prestored the corresponding identification label of the languages (such as Shanghai native language) in classifier, that is, it may recognize that this
Languages are which kind of dialect, for example, Shanghai native language.
When not prestoring the corresponding identification label of the languages (such as Shanghai native language) in classifier, that is, export a knowledge
Distinguishing label, which represents guidance voice messaging does not have corresponding category of language in category of language classifier, the identification
Label is known as Unknown Label.
When the identification label is Unknown Label, step 105 is carried out: when the identification mark of institute's category of language classifier of acquisition
When label are Unknown Label, storage guidance voice messaging.
Step 106: after storing language category information to be identified, identifying that the language category information to be identified institute stored is right
The category of language answered, and using the corresponding category of language as category of language to be learned.It in the present embodiment, can be by artificial
Mode identified, for example, after people hears the guidance voice messaging, be judged as that the guidance voice messaging is Shanghai native language, this
When, it can be using Shanghai native language as category of language to be learned.
Step 107: the category of language corresponding to the language category information to be identified that identification is stored is as language to be learned
After saying type, category of language classifier is updated according to category of language to be learned.For example, the training set of Shanghai native language is obtained, is led to
It crosses training set and extracts feature set, category of language classifier before is trained by feature set, to update category of language
Classifier.It is understood that can also add using manually the speech recognition result of this Shanghai native language and audio frequency characteristics
Enter into feature database.
It is above-mentioned to be illustrated by taking Shanghai dialect as an example.But the present invention is certainly not limited to this, can be applied to each languages,
Dialect.
The present invention also provides a kind of machine conversational language identification device, the machine conversational language identification device includes drawing
Lead sound obtains module 1, classifier characteristic extracting module 2, and classifier obtains module 3 and corresponding category of language identification module
4, wherein guidance voice obtains module 1 for obtaining guidance voice messaging;Classifier characteristic extracting module 2 is drawn for extracting
Lead the classifier feature of voice messaging;Classifier obtains module 3 for obtaining category of language classifier;Corresponding category of language is known
Other module 4 is used to carry out the guidance voice messaging by the category of language classifier according to the classifier feature
Category of language identification.
In the present embodiment, the machine conversational language identification device includes leading question output module, and leading question exports mould
Block is for exporting leading question;It includes providing user according to leading question that guidance voice, which obtains the guidance voice messaging that module obtains,
There is the guidance voice messaging of corresponding relationship with leading question.
Referring to fig. 2, in the present embodiment, classifier characteristic extracting module includes that conversion module and character features extract
Module, wherein conversion module will be for that will guide voice to be converted into guidance text information;Character features extraction module draws for extracting
The character features in text information are led as classifier feature.
Referring to fig. 2, in the present embodiment, the corresponding category of language identification module includes input module and identification label
Module is obtained, input module is used to classifier being input to category of language classifier with feature;Identification label acquisition module is used for
The identification label for obtaining the output of category of language classifier determines language corresponding to language category information to be identified according to identification label
Say type, wherein at least one identification label represents the guidance voice messaging and do not correspond in category of language classifier
Category of language, which is known as Unknown Label.
Referring to fig. 2, in the present embodiment, the machine conversational language identification device includes memory module, and memory module is used
In when the identification label of the category of language classifier of acquisition is Unknown Label, storage guides voice messaging.
Referring to fig. 2, in the present embodiment, the machine conversational language identification device includes category of language identification mould to be learned
Block, category of language corresponding to the guidance voice messaging that category of language identification module to be learned is stored for identification, and should
Corresponding category of language is as category of language to be learned.
Referring to fig. 2, in the present embodiment, machine conversational language identification device includes classifier update module, and classifier is more
New module is used to update the category of language classifier according to the category of language to be learned.
Referring to fig. 2, in the present embodiment, classifier update module includes that language feature library obtains module and training update
Module, language feature library obtain the language feature library that module is used to obtain category of language to be learned;Training update module is used for root
The language classification device is updated according to the language feature library of category of language to be learned, to form the language including category of language to be learned
Say classifier.
The present invention also provides a kind of electronic equipment, including memory, processor and storage are in the memory simultaneously
The computer program that can be run on the processor, which is characterized in that when the processor executes the computer program
Realize machine conversational language recognition methods as described above.
Fig. 3 is that by the electronic equipment of the machine conversational language recognition methods provided according to an embodiment of the present invention
Exemplary block diagram.
As shown in figure 3, electronic equipment 500 includes input equipment 501, input interface 502, central processing unit 503, memory
504, output interface 505 and output equipment 506.Wherein, input interface 502, central processing unit 503, memory 504 and defeated
Outgoing interface 505 is connected with each other by bus 507, and input equipment 501 and output equipment 506 are respectively by input interface 502 and defeated
Outgoing interface 505 is connect with bus 507, and then is connect with the other assemblies for calculating equipment 500.Specifically, input equipment 504 receives
Central processing unit 503 is transmitted to by information is inputted from external input information, and by input interface 502;Central processing unit
503 are handled to generate output information input information based on the computer executable instructions stored in memory 504, will be defeated
Information is temporarily or permanently stored in memory 504 out, and output information is then transmitted to output by output interface 505
Equipment 506;Output information is output to the outside of calculating equipment 500 for users to use by output equipment 506.
That is, electronic equipment shown in Fig. 3 also may be implemented as including: to be stored with computer executable instructions
Memory;And one or more processors, the one or more processors can be real when executing computer executable instructions
Now in conjunction with the machine conversational language recognition methods of Fig. 1 description.
In one embodiment, electronic equipment shown in Fig. 3 may be implemented as including: memory 504, be configured as depositing
Store up executable program code;One or more processors 503 are configured as the executable program stored in run memory 504
Code, to execute the machine conversational language recognition methods in above-described embodiment.
The present invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage has calculating
Machine program, the computer program can be realized machine conversational language recognition methods as described above when being executed by processor.
The present invention also provides a kind of machine conversational languages to identify engine switching method, and the machine conversational language identification is drawn
Holding up switching method includes:
According to category of language used in machine conversational language recognition methods as described above identification user;
Switch machine for identification user's voice language identification engine be the category of language corresponding to language know
Other engine.
For example, machine itself uses Mandarin Chinese language identification engine to be identified, passes through language as described above
Speech identification method is identified, when the category of language for having identified user is Shanghai native language, switching machine makes for identification
The language identification engine of user's voice is language identification engine corresponding to the category of language.Mandarin Chinese language identification is drawn
It holds up and switches to Shanghai native language language identification engine.
The present invention also provides a kind of machine conversational languages to identify engine switching device, for realizing machine as described above
Conversational language identifies engine switching method, and the machine conversational language identification engine switching device includes the identification of machine conversational language
Device and switching module, wherein machine conversational language identification device corresponds to category of language for identification, and machine conversational language is known
Other device is machine conversational language identification device as described above;Switching module is for switching machine user's voice for identification
Language identification engine be category of language corresponding to language identification engine.
The present invention also provides a kind of electronic equipment, including memory, processor and storage are in the memory simultaneously
The computer program that can be run on the processor, which is characterized in that when the processor executes the computer program
Realize machine conversational language identification engine switching method as described above.
The present invention also provides a kind of computer readable storage medium, the computer-readable recording medium storage has calculating
Machine program, which is characterized in that the computer program can be realized machine conversational language as described above when being executed by processor
Identify engine switching method.
It is not in fact for limiting the present invention, any this field although the present invention is disclosed as above with preferred embodiment
Technical staff without departing from the spirit and scope of the present invention, can make possible variation and modification, therefore, of the invention
Protection scope should be subject to the range that the claims in the present invention are defined.
In a typical configuration, calculating equipment includes one or more processors (CPU), input/output interface, net
Network interface and memory.
Memory may include the non-volatile memory in computer-readable medium, random access memory (RAM) and/or
The forms such as Nonvolatile memory, such as read-only memory (ROM) or flash memory (flash RAM).Memory is computer-readable medium
Example.
For computer-readable medium including permanent and non-permanent, removable and non-removable, media can be by any side
Method or technology realize that information stores.Information can be computer readable instructions, data structure, the module of program or other numbers
According to.The example of the storage medium of computer includes, but are not limited to phase change memory (PRAM), static random access memory
(SRAM), dynamic random access memory (DRAM), other kinds of random access memory (RAM), read-only memory
(ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory techniques, CD-ROM are read-only
Memory (CD-ROM), data multifunctional optical disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or its
His magnetic storage device or any other non-transmission medium, can be used for storing and can be accessed by a computing device information.
It will be understood by those skilled in the art that the embodiment of the present invention can provide as method, system or computer program product.
Therefore, complete hardware embodiment, complete software embodiment or embodiment combining software and hardware aspects can be used in the present invention
Form.It is deposited moreover, the present invention can be used to can be used in the computer that one or more wherein includes computer usable program code
The shape for the computer program product implemented on storage media (including but not limited to magnetic disk storage, CD-ROM, optical memory etc.)
Formula.
Furthermore, it is to be understood that one word of " comprising " does not exclude other units or steps.Multiple units for being stated in device claim,
Module or device can also be realized by a unit or overall apparatus by software or hardware.First, second equal words are used to mark
Title is known, without identifying any specific sequence.
Flow chart and block diagram in attached drawing are illustrated according to the system of various embodiments of the invention, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of module, program segment or code are used for including one or more
The executable instruction of logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
Function can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly identified actually may be used
To be basically executed in parallel, they can also execute in the opposite order sometimes, and this depends on the function involved.It is also noted that
, the combination of each box in block diagram and or flow chart and the box in block diagram and/or general flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Alleged processor can be central processing unit (Central Processing Unit, CPU), can also be it
His general processor, digital signal processor (Digital Signal Processor, DSP), specific integrated circuit
(Application Specific Integrated Circuit, ASIC), ready-made programmable gate array (Field-
Programmable Gate Array, FPGA) either other programmable logic device, discrete gate or transistor logic,
Discrete hardware components etc..General processor can be microprocessor or the processor is also possible to any conventional processor
Deng processor is above-mentioned apparatus/terminal device control centre of the invention, utilizes this entire hair of various interfaces and connection
Bright above-mentioned apparatus/terminal device various pieces.
Memory can be used for storing computer program and/or module, and processor is stored in memory by operation or execution
Interior computer program and/or module, and the data being stored in memory are called, realization device/terminal device is various
Function.Memory can mainly include storing program area and storage data area, wherein storing program area can storage program area, extremely
Application program (such as sound-playing function, image player function etc.) needed for a few function etc.;Storage data area can store
Created data (such as audio data, phone directory etc.) etc. are used according to mobile phone.In addition, memory may include high speed with
Machine accesses memory, can also include nonvolatile memory, such as hard disk, memory, plug-in type hard disk, intelligent memory card
(Smart Media Card, SMC), secure digital (Secure Digital, SD) card, flash card (Flash Card), at least
One disk memory, flush memory device or other volatile solid-state parts.
If the integrated module/unit of the device of the invention/terminal device is realized in the form of SFU software functional unit and makees
It is independent product when selling or using, can store in a computer readable storage medium.Based on this understanding,
The present invention realizes all or part of the process in above-described embodiment method, can also be instructed by computer program relevant hard
Part is completed, computer program can be stored in a computer readable storage medium, which holds by processor
When row, it can be achieved that the step of above-mentioned each embodiment of the method.Wherein, computer program includes computer program code, computer
Program code can be source code form, object identification code form, executable file or certain intermediate forms etc..Computer-readable Jie
Matter may include: can carry computer program code any entity or device, recording medium, USB flash disk, mobile hard disk, magnetic disk,
CD, computer storage, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random
Access Memory), electric carrier signal, telecommunication signal and software distribution medium etc..It should be noted that computer-readable
The content that medium includes can carry out increase and decrease appropriate according to the requirement made laws in jurisdiction with patent practice, such as at certain
A little jurisdictions do not include electric carrier signal and telecommunication signal according to legislation and patent practice, computer-readable medium.
Finally it is noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations.To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, those skilled in the art should understand that: it is still
It is possible to modify the technical solutions described in the foregoing embodiments, or part of technical characteristic is equally replaced
It changes;And these are modified or replaceed, the essence for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution
Mind and range.
Claims (10)
1. a kind of machine conversational language recognition methods, which is characterized in that the machine conversational language recognition methods includes:
Obtain guidance voice messaging;
According to the guidance voice messaging, the classifier feature of the guidance voice messaging is extracted;
Obtain category of language classifier;
According to the classifier feature, category of language is carried out to the guidance voice messaging by the category of language classifier
Identification.
2. machine conversational language recognition methods as described in claim 1, which is characterized in that guide voice messaging in the acquisition
Include: before
Export leading question;
The acquisition guidance voice messaging includes providing user according to the leading question to have corresponding relationship with leading question
Guide voice messaging.
3. machine conversational language recognition methods as described in claim 1, which is characterized in that described according to the leading question message
Breath, the classifier for extracting the guidance voice messaging include: with feature
The guidance voice is converted into guidance text information;
The character features in the guidance text information are extracted as classifier feature.
4. machine conversational language recognition methods as described in claim 1, which is characterized in that described special according to the classifier
Sign, carrying out category of language identification to the guidance voice messaging by the category of language classifier includes:
The classifier is input to the category of language classifier with feature;
The identification label for obtaining the category of language classifier output, determines the language kind to be identified according to the identification label
Category of language corresponding to category information, wherein at least one described identification label represents the guidance voice messaging described
There is no corresponding category of language in category of language classifier, which is known as Unknown Label.
5. machine conversational language recognition methods as claimed in claim 4, which is characterized in that when the category of language point of acquisition
When the identification label of class device is Unknown Label, the guidance voice messaging is stored.
6. machine conversational language recognition methods as claimed in claim 5, which is characterized in that machine conversational language identification side
Method includes:
After the storage language category information to be identified, identify corresponding to the language category information to be identified stored
Category of language, and using the corresponding category of language as category of language to be learned.
7. machine conversational language recognition methods as claimed in claim 6, which is characterized in that machine conversational language identification side
Method includes:
Identify category of language corresponding to the language category information to be identified stored as language kind to be learned described
After class, the category of language classifier is updated according to the category of language to be learned.
8. a kind of machine conversational language identification device, which is characterized in that the machine conversational language identification device includes:
Voice is guided to obtain module, the guidance voice obtains module for obtaining guidance voice messaging;
Classifier characteristic extracting module, the classifier characteristic extracting module are used to extract point of the guidance voice messaging
Class device feature;
Classifier obtains module, and the classifier obtains module for obtaining category of language classifier;
Corresponding category of language identification module, the corresponding category of language identification module are used to be led to according to the classifier feature
It crosses the category of language classifier and category of language identification is carried out to the guidance voice messaging.
9. a kind of machine conversational language identifies engine switching method, which is characterized in that the machine conversational language identification engine is cut
The method of changing includes:
Language used in machine conversational language recognition methods identification user as claimed in any of claims 1 to 7
Say type;
Switching machine for identification user's voice language identification engine be the category of language corresponding to language identification draw
It holds up.
10. a kind of machine conversational language identifies engine switching device, which is characterized in that the machine conversational language identification engine is cut
Changing device includes:
Machine conversational language identification device, the machine conversational language identification device are machine as claimed in claim 8 dialogue
Speech recognition equipment, for using machine conversational language recognition methods as claimed in any of claims 1 to 7 in one of claims identification pair
Answer category of language;
Switching module, the language identification engine that the switching module is used to switch machine user's voice for identification is institute's predicate
Say language identification engine corresponding to type.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910811586.2A CN110503938A (en) | 2019-08-30 | 2019-08-30 | The recognition methods of machine conversational language and device, identification engine switching method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910811586.2A CN110503938A (en) | 2019-08-30 | 2019-08-30 | The recognition methods of machine conversational language and device, identification engine switching method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110503938A true CN110503938A (en) | 2019-11-26 |
Family
ID=68590615
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910811586.2A Pending CN110503938A (en) | 2019-08-30 | 2019-08-30 | The recognition methods of machine conversational language and device, identification engine switching method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110503938A (en) |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101825953A (en) * | 2010-04-06 | 2010-09-08 | 朱建政 | Chinese character input product with combined voice input and Chinese phonetic alphabet input functions |
CN102521221A (en) * | 2011-11-30 | 2012-06-27 | 江苏奇异点网络有限公司 | Multilingual conference information output method with text output function |
CN102915731A (en) * | 2012-10-10 | 2013-02-06 | 百度在线网络技术(北京)有限公司 | Method and device for recognizing personalized speeches |
CN103456303A (en) * | 2013-08-08 | 2013-12-18 | 四川长虹电器股份有限公司 | Method for controlling voice and intelligent air-conditionier system |
CN104252287A (en) * | 2014-09-04 | 2014-12-31 | 广东小天才科技有限公司 | Interactive device and method for improving expressive ability on basis of same |
CN105185375A (en) * | 2015-08-10 | 2015-12-23 | 联想(北京)有限公司 | Information processing method and electronic equipment |
CN105654953A (en) * | 2016-03-22 | 2016-06-08 | 美的集团股份有限公司 | Voice control method and system |
CN106710586A (en) * | 2016-12-27 | 2017-05-24 | 北京智能管家科技有限公司 | Speech recognition engine automatic switching method and device |
CN106997762A (en) * | 2017-03-08 | 2017-08-01 | 广东美的制冷设备有限公司 | The sound control method and device of household electrical appliance |
CN107146612A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | Voice guide method, device, smart machine and server |
CN107305769A (en) * | 2016-04-20 | 2017-10-31 | 斑马网络技术有限公司 | Voice interaction processing method, device, equipment and operating system |
CN108364646A (en) * | 2018-02-08 | 2018-08-03 | 上海智臻智能网络科技股份有限公司 | Embedded speech operating method, device and system |
CN109448712A (en) * | 2018-11-12 | 2019-03-08 | 百度在线网络技术(北京)有限公司 | Voice interactive method, device, equipment and storage medium |
-
2019
- 2019-08-30 CN CN201910811586.2A patent/CN110503938A/en active Pending
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101825953A (en) * | 2010-04-06 | 2010-09-08 | 朱建政 | Chinese character input product with combined voice input and Chinese phonetic alphabet input functions |
CN102521221A (en) * | 2011-11-30 | 2012-06-27 | 江苏奇异点网络有限公司 | Multilingual conference information output method with text output function |
CN102915731A (en) * | 2012-10-10 | 2013-02-06 | 百度在线网络技术(北京)有限公司 | Method and device for recognizing personalized speeches |
CN103456303A (en) * | 2013-08-08 | 2013-12-18 | 四川长虹电器股份有限公司 | Method for controlling voice and intelligent air-conditionier system |
CN104252287A (en) * | 2014-09-04 | 2014-12-31 | 广东小天才科技有限公司 | Interactive device and method for improving expressive ability on basis of same |
CN105185375A (en) * | 2015-08-10 | 2015-12-23 | 联想(北京)有限公司 | Information processing method and electronic equipment |
CN105654953A (en) * | 2016-03-22 | 2016-06-08 | 美的集团股份有限公司 | Voice control method and system |
CN107305769A (en) * | 2016-04-20 | 2017-10-31 | 斑马网络技术有限公司 | Voice interaction processing method, device, equipment and operating system |
CN106710586A (en) * | 2016-12-27 | 2017-05-24 | 北京智能管家科技有限公司 | Speech recognition engine automatic switching method and device |
CN106997762A (en) * | 2017-03-08 | 2017-08-01 | 广东美的制冷设备有限公司 | The sound control method and device of household electrical appliance |
CN107146612A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | Voice guide method, device, smart machine and server |
CN108364646A (en) * | 2018-02-08 | 2018-08-03 | 上海智臻智能网络科技股份有限公司 | Embedded speech operating method, device and system |
CN109448712A (en) * | 2018-11-12 | 2019-03-08 | 百度在线网络技术(北京)有限公司 | Voice interactive method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108847241B (en) | Method for recognizing conference voice as text, electronic device and storage medium | |
US11475897B2 (en) | Method and apparatus for response using voice matching user category | |
CN108447471A (en) | Audio recognition method and speech recognition equipment | |
WO2017054122A1 (en) | Speech recognition system and method, client device and cloud server | |
CN108428446A (en) | Audio recognition method and device | |
CN110473566A (en) | Audio separation method, device, electronic equipment and computer readable storage medium | |
CN105895103A (en) | Speech recognition method and device | |
CN110517664A (en) | Multi-party speech recognition methods, device, equipment and readable storage medium storing program for executing | |
US20100145710A1 (en) | Data-Driven Voice User Interface | |
CN111445898B (en) | Language identification method and device, electronic equipment and storage medium | |
CN104538034A (en) | Voice recognition method and system | |
CN109976702A (en) | A kind of audio recognition method, device and terminal | |
CN109213856A (en) | A kind of method for recognizing semantics and system | |
CN108735200A (en) | A kind of speaker's automatic marking method | |
CN111951779A (en) | Front-end processing method for speech synthesis and related equipment | |
CN110428853A (en) | Voice activity detection method, Voice activity detection device and electronic equipment | |
CN112463942A (en) | Text processing method and device, electronic equipment and computer readable storage medium | |
CN110597958B (en) | Text classification model training and using method and device | |
CN111178081A (en) | Semantic recognition method, server, electronic device and computer storage medium | |
WO2022260794A1 (en) | Reducing biases of generative language models | |
CN114860938A (en) | Statement intention identification method and electronic equipment | |
US11132999B2 (en) | Information processing device, information processing method, and non-transitory computer readable storage medium | |
US11615787B2 (en) | Dialogue system and method of controlling the same | |
CN112466287B (en) | Voice segmentation method, device and computer readable storage medium | |
CN111046674B (en) | Semantic understanding method and device, electronic equipment and 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 | ||
CB03 | Change of inventor or designer information |
Inventor after: Dai Jian Inventor after: Zhou Weidong Inventor after: Liu Hua Inventor after: Liu Kai Inventor after: Yu Ling Inventor before: Dai Jian |
|
CB03 | Change of inventor or designer information | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191126 |
|
RJ01 | Rejection of invention patent application after publication |