CN109658925A - It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context - Google Patents
It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context Download PDFInfo
- Publication number
- CN109658925A CN109658925A CN201811435993.XA CN201811435993A CN109658925A CN 109658925 A CN109658925 A CN 109658925A CN 201811435993 A CN201811435993 A CN 201811435993A CN 109658925 A CN109658925 A CN 109658925A
- Authority
- CN
- China
- Prior art keywords
- voice
- state
- dialogue
- speech recognition
- context
- 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 58
- 230000002452 interceptive effect Effects 0.000 claims abstract description 38
- 230000008569 process Effects 0.000 claims description 23
- 238000010801 machine learning Methods 0.000 claims description 16
- 238000006243 chemical reaction Methods 0.000 claims description 14
- 230000004044 response Effects 0.000 claims description 14
- 238000011156 evaluation Methods 0.000 claims description 13
- 238000004891 communication Methods 0.000 claims description 3
- 235000013399 edible fruits Nutrition 0.000 claims description 2
- 230000006399 behavior Effects 0.000 claims 1
- 230000003993 interaction Effects 0.000 abstract description 9
- 230000001427 coherent effect Effects 0.000 abstract description 3
- 238000004378 air conditioning Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000036651 mood 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
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- 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
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/223—Execution procedure of a spoken command
-
- 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
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/225—Feedback of the input speech
-
- 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
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
- G10L2015/226—Procedures used during a speech recognition process, e.g. man-machine dialogue using non-speech characteristics
- G10L2015/228—Procedures used during a speech recognition process, e.g. man-machine dialogue using non-speech characteristics of application context
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)
- Navigation (AREA)
Abstract
It is a kind of that wake-up vehicle-mounted voice dialogue method is exempted from based on context, system, realize the electronic equipment and computer readable storage medium of this method, the state that vehicle-mounted voice is talked with is divided into init state by the method, dialogue state and end three kinds of states of state, voice input after init state is identified, analysis and judgement, the sound-type for responding reply is needed then to enter dialogue state if it is context-sensitive, otherwise ignore the input of this voice, further interactive voice is continued to determine whether, until time-out or user enter end state when having the intention for exiting dialogue.The speech dialog method and corresponding speech dialogue system can carry out coherent dialogue interaction on the basis of being based on context dependence with user, avoid reawaking in dialogue, largely the user experience is improved for this continuous dialogue.
Description
Technical field
The present invention relates to car-mounted terminal technical fields more particularly to a kind of vehicle-mounted voice of exempting to wake up based on context to talk with
Method and system.
Background technique
With the development of economy and social progress, automobile have been increasingly becoming the popular consumer goods.The consumer group of automobile
Body is constantly being expanded, and requirement of the consumer to automobile is also higher and higher.In recent years, in addition to economy, dynamic property, in driving conditions
In driving experience also become consumer's focus of attention.The technology development of mobile unit is also entered therewith deeper and deeper.
Vehicle-mounted voice conversational system is come into being, can be by the intention of speech recognition user, in addition to having and user's spare time
Function merely, moreover it is possible to complete control air-conditioning, window opening-closing, the function that vehicle is manipulated such as make a phone call and navigation, online,
The functions such as media entertainment, Weather information inquiry.For the considerations of avoiding misrecognition, speech dialogue system, which is mainly based upon, at present is called out
What awake method was realized.User wants and system engages in the dialogue it may first have to be waken up, then system can just identify user
Voice input.The benefit designed in this way is that conversational system will not identify unrelated voice input, always explicitly to wake up as mark
Knowledge starts to talk with, and still, insertion wakes up word or wake operation always in dialogue, sacrifices the continuity of dialogue, especially exists
Under vehicle-mounted scene, user is usually to pass through speech dialogue system to execute some associated tasks, between these associated tasks every time
It is all waken up, greatly reduces user experience.
Summary of the invention
Exempt to wake up vehicle-mounted voice dialogue method to solve the above problems, the invention proposes a kind of based on context and be
System, under the vehicle-mounted this stronger scene of task, the correlation based on context ensure that the continuity of dialogue, avoid every
The operation waken up before secondary dialogue.
In order to achieve the above objectives, the first aspect of the present invention provide it is a kind of based on context exempt from wake up vehicle-mounted voice pair
Words method, includes the following steps:
It is waken up according to predetermined instruction, opens speech recognition;
Into init state, first time interactive voice is received, is inputted according to voice and is carried out speech recognition and execute corresponding
Operation or answer;
Judge whether there is further interactive voice: if so, then judging that this voice inputs using machine learning model
Whether it is the context-sensitive sound-type for needing to respond reply, is then to enter dialogue state, it is defeated otherwise to ignore this voice
Enter, has continued to determine whether further interactive voice;If it is not, into end state;
Into dialogue state, is inputted according to voice and carry out speech recognition and execute corresponding operation or answer;The voice is known
Other result is intended to if there is exiting, then enters end state, otherwise continued to determine whether further interactive voice;
Into state is terminated, speech recognition is closed.
In certain embodiments, the predetermined instruction includes calling pre-set wake-up word or unlatching wake operation
Switch.
In certain embodiments, under init state, response reply is all carried out to the input of any kind of voice.
In certain embodiments, any kind of voice input includes Task, chat type or question and answer type
Voice input.
In certain embodiments, the context-sensitive sound-type for needing to respond reply includes context phase
The Task voice of pass inputs or the input of the chat type voice of evaluation system.
In certain embodiments, the context-sensitive Task voice input includes tasks in areas correlation or letter
Breath inherits relevant voice input;The tasks in areas correlation refers to the associated task in same field, including grasps to vehicle
The task of control, navigation task, media music task, Weather information query task and/or make a phone call task;The information
It inherits correlation and refers to the succession with information between two tasks, next task is executed based on the information in task before.
In certain embodiments, the chat type voice input of the evaluation system includes to speech dialogue system each side
The evaluation in face.
In certain embodiments, described the step of judging whether there is further interactive voice includes: in pre- timing
It is interior to be not received by any voice input, then enter end state.
The second aspect of the present invention provide it is a kind of based on context exempt from wake up vehicle-mounted voice conversational system, comprising:
Wake up instruction receiving module wakes up vehicle-mounted voice conversational system according to predetermined instruction, opens speech recognition module;
Speech recognition module receives voice input and carries out being identified and given out speech recognition result;
State conversion module, by the vehicle-mounted voice conversational system init state, dialogue state and terminate state it
Between converted;
Dialog process module executes corresponding operation or answer according to speech recognition result;
When init state, the speech recognition module receives first time interactive voice, is inputted according to voice and carries out voice
Identification provides speech recognition result, and the dialog process module executes corresponding operation or answer according to speech recognition result;
The state conversion module judges whether there is further interactive voice: if so, then utilizing machine learning model
Judge whether the input of this voice is the context-sensitive sound-type for needing to respond reply, is then to enter dialogue state, it is no
Then ignore the input of this voice, continues to determine whether further interactive voice;If it is not, into end state;
In dialogue state, the speech recognition module inputs progress speech recognition according to voice and provides speech recognition knot
Fruit, the dialog process module execute corresponding operation or answer according to speech recognition result;The speech recognition result if there is
Intention is exited, then enters end state, has otherwise continued to determine whether further interactive voice;
When terminating state, the speech recognition module is closed.
In certain embodiments, the state conversion module further includes overtime judging unit, is not had in the given time
Any voice input is received, then is judged as time-out, the state conversion module turns the state of vehicle-mounted voice conversational system
It is changed to end state.
The third aspect of the present invention provides a kind of electronic equipment, which includes:
Memory and one or more processors;
Wherein, the memory is connect with one or more of processor communications, and being stored in the memory can quilt
The instruction that one or more of processors execute, described instruction is executed by one or more of processors, so that described one
A or multiple processors be used for executes as it is preceding it is described in any item based on context exempt from wake-up vehicle-mounted voice dialogue method.
The fourth aspect of the present invention provides a kind of computer readable storage medium, is stored thereon with the executable finger of computer
It enables, when the computer executable instructions are executed by a computing apparatus, is operable to execute as preceding described in any item based on upper
Hereafter exempt from wake up vehicle-mounted voice dialogue method.
Wake-up vehicle-mounted voice dialogue method, system, realization are exempted from based on context in conclusion the present invention provides a kind of
The state that vehicle-mounted voice is talked with is divided into initialization shape by the electronic equipment and computer readable storage medium of this method, the method
State, dialogue state and end three kinds of states of state, input the voice after init state and are identified, analyzed and judged,
It needs the sound-type for responding reply then to enter dialogue state if it is context-sensitive, otherwise ignores the input of this voice,
Further interactive voice is continued to determine whether, until time-out or user enter end state when having the intention for exiting dialogue.
The speech dialog method and corresponding speech dialogue system can carry out on the basis of being based on context dependence with user
Coherent dialogue interaction, avoids reawaking in dialogue, and this continuous dialogue largely improves user's body
It tests.
Detailed description of the invention
Fig. 1 is the flow diagram of the invention that exempt to wake up vehicle-mounted voice dialogue method based on context;
Fig. 2 is the block diagram of the invention that exempt to wake up vehicle-mounted voice conversational system based on context.
Specific embodiment
In order to make the objectives, technical solutions and advantages of the present invention clearer, With reference to embodiment and join
According to attached drawing, the present invention is described in more detail.It should be understood that these descriptions are merely illustrative, and it is not intended to limit this hair
Bright range.In addition, in the following description, descriptions of well-known structures and technologies are omitted, to avoid this is unnecessarily obscured
The concept of invention.
As shown in Figure 1, the first aspect of the present invention provide it is a kind of based on context exempt from wake up vehicle-mounted voice dialogue side
Method 100, includes the following steps:
Step 110, it is waken up according to predetermined instruction, opens speech recognition.
The predetermined instruction includes calling pre-set wake-up word or opening wake operation to switch.For example, to the vehicle-mounted language
The wake-up of sound conversational system can be by way of greeting to the speech dialogue system and carry out, such as " hello " " Hi "
" Hello " etc..It is then waken up when speech dialogue system receives wake-up word voice input signal, and makes corresponding answer, for example,
" please say ", " please tell " etc..Or onboard setting voice shortcut key is switched as wake operation, user presses shortcut key
I.e. the speech dialogue system is waken up, specifically, voice shortcut key can be set on the steering wheel or on operation panel.It is vehicle-mounted
After speech dialogue system is waken up, speech recognition module is turned on, and is inputted and is identified for receiving voice, provides identification
As a result.
Step 120, into init state, first time interactive voice is received, is inputted according to voice and carries out speech recognition simultaneously
Execute corresponding operation or answer.
The init state refer to user wake up conversational system after first time interaction when state, in this state, use
Any speech dialogue system that family is said all carries out response reply, including Task, chat type and question and answer type etc..Dialog process module
Identify the intention of user using machine learning model, selection enters dialogue state or terminates state, such as user say be
" exiting ", conversational system can be set as state end state, and close speech recognition;If what user said is " to close vehicle
State is then set as dialogue state by window ", system, is continued starting speech recognition after carrying out response reply to user, is waited user
It speaks.At this moment user can be directly entered second of interaction there is no need to reawake.
I.e. under init state, conversational system is intended to carry out state conversion according to user, exits intention and then enters knot
Otherwise pencil state enters dialogue state.
Step 130, further interactive voice is judged whether there is.
Step 140, if there is further interactive voice, then whether the input of this voice is judged using machine learning model
It needs to respond the sound-type replied to be context-sensitive, is then to enter dialogue state, otherwise ignore the input of this voice, after
It is continuous to judge whether there is further interactive voice;If without further interactive voice, conversational system if, enters end state.
The dialogue state refers to that conversational system state in which after interaction for the first time, end state refer to that conversational system terminates
With the dialogue of user, and the state of speech recognition is closed.
Judged specifically, this judges whether there is the settable timeout mechanism of further interactive voice, if pre-
Interior conversational system of fixing time is not received by any voice input, then enters end state and close speech recognition.The pre- timing
Between can be set to 30 seconds or other durations.
If conversational system receives further interactive voice within the above-mentioned predetermined time, dialog process module passes through
Recall context (dialog history, i.e. dialogue in this dialogue before the input of this voice) and judges this using machine learning model
Whether secondary voice input has correlation, then selects the state that enter.Specifically, context-sensitive needs respond back
Multiple sound-type includes the chat type voice input etc. of context-sensitive Task voice input or evaluation system.Context
Relevant Task voice input includes that tasks in areas is related or the relevant voice of Information inheriting inputs.The tasks in areas correlation refers to
The associated task in same field, including the task, navigation task, media music task, weather manipulated to vehicle
Information query task and/or task is made a phone call, for example, user says " opening air-conditioning " or " closing vehicle window ", conversational system determines
It is that there is the Task voice of tasks in areas correlation to input, conversational system then enters dialogue state, continues starting speech recognition.
The Information inheriting correlation refers to the succession with information between two tasks, and next task is executed based on the information in task before,
Such as " how is Beijing weather " and " navigation past ", the second word needs to inherit " Beijing " this information, has Information inheriting
Correlation.The chat of evaluation system type refers to evaluation of the user to speech dialogue system, including to speech dialogue system each side
The evaluation in face, such as " you are very clever " etc, system also can carry out response reply to such dialogue.
If user say be one not the least concerned if, the not input of context-sensitive Task voice or evaluation system
The chat type voice of system inputs, then conversational system ignores the input of this voice, has continued to determine whether further interactive voice,
Keep dialogue state.
Step 150, into dialogue state, progress speech recognition is inputted according to voice and executes corresponding operation or answer,
Judge whether the speech recognition result exits intention, be intended to then enter step 160 if there is exiting, otherwise return step 130,
Further interactive voice is continued to determine whether.
Such as above-mentioned " opening air-conditioning " or " closing vehicle window ", the dialog process module in conversational system carries out corresponding operation
Afterwards, speech recognition is continued starting, the further interactive voice of user is waited.
Under dialogue state, the chat type voice of context-sensitive Task, evaluation system type is inputted, dialogue system
System carries out response reply, and the input of this voice is ignored in the case of other.
Step 160, into state is terminated, speech recognition is closed.
By above-mentioned steps 110-160, the interactive process of a user and speech dialogue system are completed.Due to the vehicle-mounted language
Sound conversational system judges whether the input of this voice is up and down during dialogue continues, through machine learning model trained in advance
The relevant Task voice input of text or the chat type voice input of evaluation system, are if it is carried out by conversational system corresponding
Operation replies, and needs to wake up the operation of conversational system before avoiding each voice input, maintains the continuity of dialogue, mention
The high experience sense of user.
The second aspect of the present invention provide it is a kind of based on context exempt from wake up vehicle-mounted voice conversational system 200, such as Fig. 2
It is shown, comprising: wake up instruction receiving module 210 wakes up vehicle-mounted voice conversational system according to predetermined instruction, opens speech recognition mould
Block 220;Speech recognition module 220 receives voice input and carries out being identified and given out speech recognition result;State conversion module 230,
The vehicle-mounted voice conversational system is converted between init state, dialogue state and end state;Dialog process mould
Block 240 executes corresponding operation or answer according to speech recognition result;When init state, the speech recognition module 220 is connect
First time interactive voice is received, progress speech recognition is inputted according to voice and provides speech recognition result, the dialog process module 240
Corresponding operation or answer are executed according to speech recognition result;The state conversion module 230 judges whether there is further language
Sound interaction: if so, then judging whether the input of this voice is that context-sensitive needs respond back using machine learning model
Multiple sound-type is then to enter dialogue state, otherwise ignores the input of this voice, has continued to determine whether further voice
Interaction;If it is not, into end state;In dialogue state, the speech recognition module 220 is inputted according to voice and is carried out
Speech recognition provides speech recognition result, the dialog process module 240 according to speech recognition result execute corresponding operation or
It replies;When terminating state, the speech recognition module 220 is closed.
In certain embodiments, state conversion module 230 further includes overtime judging unit 231, in the given time
It is not received by the input of any voice, then is judged as time-out, the state conversion module 230 is by the shape of vehicle-mounted voice conversational system
State is converted to end state.
The third aspect of the present invention provides a kind of electronic equipment, which includes: memory and one or more
A processor;Wherein, the memory is connect with one or more of processor communications, and being stored in the memory can quilt
The instruction that one or more of processors execute, described instruction is executed by one or more of processors, so that described one
A or multiple processors, which are used to execute, exempts from wake-up vehicle-mounted voice dialogue method based on context as previously described.
The fourth aspect of the present invention provides a kind of computer readable storage medium, is stored thereon with the executable finger of computer
It enables, when the computer executable instructions are executed by a computing apparatus, is operable to execute as previously described based on context
Exempt to wake up vehicle-mounted voice dialogue method.
The present invention is further detailed with specific embodiment below.
Embodiment one
User: " Hi ".User initiates interactive voice.
Conversational system: " you are good, please tell ".The speech dialogue system is waken up, and is carried out replying response and is opened voice knowledge
Not, into init state.
User: " exiting ".
Conversational system: " good, goodbye ".Dialog process module, which recognizes user using machine learning model, has end to talk with
Intention, then select to enter end state, and close speech recognition.
Embodiment two
User: " Hi ".User initiates interactive voice.
Conversational system: " you are good, please tell ".The speech dialogue system is waken up, and is carried out replying response and is opened voice knowledge
Not, into init state.
User: " closing vehicle window ".
Conversational system: " good, to close vehicle window for you ".Dialog process module executes corresponding according to speech recognition result
Operation replies, into dialogue state.
User: " opening air-conditioning ".
Conversational system: " good, this is turned on ".Dialog process module is judged as context-sensitive using machine learning model
Task input, execute corresponding operation or answer, and keep dialogue state.
User: " exiting ".
Conversational system: " good, goodbye ".Dialog process module, which recognizes user using machine learning model, has end to talk with
Intention, then select to enter end state, and close speech recognition.
Embodiment three
User: " Hi ".User initiates interactive voice.
Conversational system: " you are good, please tell ".The speech dialogue system is waken up, and is carried out replying response and is opened voice knowledge
Not, into init state.
User: " closing vehicle window ".
Conversational system: " good, to close vehicle window for you ".Dialog process module executes corresponding according to speech recognition result
Operation replies, into dialogue state.
User: " you are very clever ".
Conversational system: " thanks compliment ".Dialog process module is judged as the context-sensitive spare time using machine learning model
Merely type inputs, and reply response, system thinks that this kind of evaluation type chats the intention for having end to talk with, and into end state, and closes
Speech recognition.
Example IV
User: " Hi ".User initiates interactive voice.
Conversational system: " you are good, please tell ".The speech dialogue system is waken up, and is carried out replying response and is opened voice knowledge
Not, into init state.
User: " closing vehicle window ".
Conversational system: " good, to close vehicle window for you ".Dialog process module executes corresponding according to speech recognition result
Operation replies, into dialogue state.
User does not carry out voice input in for a period of time, trigger timeout mechanism, and conversational system enters end state, closes language
Sound identification.
Embodiment five
User: " Hi ".User initiates interactive voice.
Conversational system: " you are good, please tell ".The speech dialogue system is waken up, and is carried out replying response and is opened voice knowledge
Not, into init state.
User: " closing vehicle window ".
Conversational system: " good, to close vehicle window for you ".Dialog process module executes corresponding according to speech recognition result
Operation replies, into dialogue state.
User: " guess my mood of today how ".
Conversational system is judged as that the incoherent voice of context inputs using machine learning model, and it is defeated to ignore this voice
Enter, responded without replying, but keep dialogue state, waits user speech input.
User: " opening air-conditioning ".
Conversational system: " good, this is turned on ".Dialog process module is judged as context-sensitive using machine learning model
Task input, reply response, and keep dialogue state.
User: " exiting ".
Conversational system: " good, goodbye ".Dialog process module, which recognizes user using machine learning model, has end to talk with
Intention, then select to enter end state, and close speech recognition.
It should be noted that specific appellation can be arranged for the conversational system, in this way, above-mentioned initiation interaction, uses " hello "
When modes such as " Hi " " Hello ", their specific appellation can be added.
Wake-up vehicle-mounted voice dialogue method, system, realization are exempted from based on context in conclusion the present invention provides a kind of
The state that vehicle-mounted voice is talked with is divided into initialization shape by the electronic equipment and computer readable storage medium of this method, the method
State, dialogue state and end three kinds of states of state, input the voice after init state and are identified, analyzed and judged,
It needs the sound-type for responding reply then to enter dialogue state if it is context-sensitive, otherwise ignores the input of this voice,
Further interactive voice is continued to determine whether, until time-out or user enter end state when having the intention for exiting dialogue.
The speech dialog method and corresponding speech dialogue system can carry out on the basis of being based on context dependence with user
Coherent dialogue interaction, avoids reawaking in dialogue, and this continuous dialogue largely improves user's body
It tests.
It should be understood that above-mentioned specific embodiment of the invention is used only for exemplary illustration or explains of the invention
Principle, but not to limit the present invention.Therefore, that is done without departing from the spirit and scope of the present invention is any
Modification, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.In addition, appended claims purport of the present invention
Covering the whole variations fallen into attached claim scope and boundary or this range and the equivalent form on boundary and is repairing
Change example.
Claims (12)
1. a kind of exempt from wake-up vehicle-mounted voice dialogue method based on context, which comprises the steps of:
It is waken up according to predetermined instruction, opens speech recognition;
Into init state, first time interactive voice is received, is inputted according to voice and is carried out speech recognition and execute corresponding behaviour
Make or replies;
Judge whether there is further interactive voice: if so, then whether judging the input of this voice using machine learning model
It needs to respond the sound-type replied to be context-sensitive, is then to enter dialogue state, otherwise ignore the input of this voice, after
It is continuous to judge whether there is further interactive voice;If it is not, into end state;
Into dialogue state, is inputted according to voice and carry out speech recognition and execute corresponding operation or answer;The speech recognition knot
Fruit is intended to if there is exiting, then enters end state, otherwise returns and has continued to determine whether further interactive voice;
Into state is terminated, speech recognition is closed.
2. according to claim 1 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that described pre-
Fixed instruction includes calling pre-set wake-up word or unlatching wake operation switch.
3. according to claim 1 or 2 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that
Under init state, response reply is all carried out to the input of any kind of voice.
4. according to claim 3 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that described
The voice input of what type includes that the voice of Task, chat type or question and answer type inputs.
5. according to claim 1 or 2 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that institute
Stating the context-sensitive sound-type for needing response to reply includes context-sensitive Task voice input or evaluation system
Chat type voice input.
6. according to claim 5 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that on described
Hereafter relevant Task voice input includes that tasks in areas is related or the relevant voice of Information inheriting inputs;The tasks in areas
Correlation refers to the associated task in same field, and task, navigation task, media music including being manipulated to vehicle are appointed
Business and/or makes a phone call task at Weather information query task;The Information inheriting correlation refer between two tasks with information after
It holds, next task is executed based on the information in task before.
7. according to claim 5 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that institute's commentary
The chat type voice input of valence system includes the evaluation to speech dialogue system various aspects.
8. according to claim 1 or 2 exempt from wake-up vehicle-mounted voice dialogue method based on context, which is characterized in that institute
Stating the step of judging whether there is further interactive voice includes: to be not received by any voice input in the given time, then
Into end state.
9. a kind of exempt from wake-up vehicle-mounted voice conversational system based on context characterized by comprising
Wake up instruction receiving module wakes up vehicle-mounted voice conversational system according to predetermined instruction, opens speech recognition module;
Speech recognition module receives voice input and carries out being identified and given out speech recognition result;
State conversion module, by the vehicle-mounted voice conversational system between init state, dialogue state and end state into
Row conversion;
Dialog process module executes corresponding operation or answer according to speech recognition result;
When init state, the speech recognition module receives first time interactive voice, is inputted according to voice and carries out speech recognition
Speech recognition result is provided, the dialog process module executes corresponding operation or answer according to speech recognition result;
The state conversion module judges whether there is further interactive voice: if so, then being judged using machine learning model
Whether the input of this voice is the context-sensitive sound-type for needing to respond reply, is then to enter dialogue state, otherwise neglects
Slightly this voice input, has continued to determine whether further interactive voice;If it is not, into end state;
In dialogue state, the speech recognition module inputs progress speech recognition according to voice and provides speech recognition result, institute
It states dialog process module and corresponding operation or answer is executed according to speech recognition result;The speech recognition result is if there is exiting meaning
Figure then enters end state, has otherwise continued to determine whether further interactive voice;
When terminating state, the speech recognition module is closed.
10. according to claim 9 exempt from wake-up vehicle-mounted voice conversational system based on context, which is characterized in that described
State conversion module further includes overtime judging unit, is not received by any voice input in the given time, is then judged as super
When, the state of vehicle-mounted voice conversational system is converted to end state by the state conversion module.
11. a kind of electronic equipment, which is characterized in that the electronic equipment includes:
Memory and one or more processors;
Wherein, the memory is connect with one or more of processor communications, and being stored in the memory can be described
The instruction that one or more processors execute, described instruction executed by one or more of processors so that it is one or
Multiple processors require the described in any item methods of 1-8 for perform claim.
12. a kind of computer readable storage medium, is stored thereon with computer executable instructions, refer to when the computer is executable
When order is executed by a computing apparatus, it is operable to perform claim and requires the described in any item methods of 1-8.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811435993.XA CN109658925A (en) | 2018-11-28 | 2018-11-28 | It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811435993.XA CN109658925A (en) | 2018-11-28 | 2018-11-28 | It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109658925A true CN109658925A (en) | 2019-04-19 |
Family
ID=66111767
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811435993.XA Pending CN109658925A (en) | 2018-11-28 | 2018-11-28 | It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109658925A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110182155A (en) * | 2019-05-14 | 2019-08-30 | 中国第一汽车股份有限公司 | Sound control method, vehicle control syetem and the vehicle of vehicle control syetem |
CN110517678A (en) * | 2019-08-28 | 2019-11-29 | 南昌保莱科技有限公司 | A kind of AI voice answer-back response system of view-based access control model induction |
CN110706702A (en) * | 2019-10-15 | 2020-01-17 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
CN110751948A (en) * | 2019-10-18 | 2020-02-04 | 珠海格力电器股份有限公司 | Voice recognition method, device, storage medium and voice equipment |
CN111776943A (en) * | 2020-06-18 | 2020-10-16 | 北京起重运输机械设计研究院有限公司 | Voice control method and system for crane |
CN112017629A (en) * | 2020-07-15 | 2020-12-01 | 马上消费金融股份有限公司 | Conversation control method and equipment of voice robot and storage medium |
CN112092820A (en) * | 2020-09-03 | 2020-12-18 | 广州小鹏汽车科技有限公司 | Initialization setting method for vehicle, and storage medium |
CN112787899A (en) * | 2021-01-08 | 2021-05-11 | 青岛海尔特种电冰箱有限公司 | Equipment voice interaction method, computer readable storage medium and refrigerator |
CN113470634A (en) * | 2020-04-28 | 2021-10-01 | 海信集团有限公司 | Control method of voice interaction equipment, server and voice interaction equipment |
CN115914453A (en) * | 2021-09-30 | 2023-04-04 | 华为技术有限公司 | Method, terminal and server for operating application program |
CN116662654A (en) * | 2023-05-26 | 2023-08-29 | 惠州市西子湖畔网络有限公司 | Big data-based affinity matching system and method |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106776578A (en) * | 2017-01-03 | 2017-05-31 | 竹间智能科技(上海)有限公司 | Talk with the method and device of performance for lifting conversational system |
CN107004410A (en) * | 2014-10-01 | 2017-08-01 | 西布雷恩公司 | Voice and connecting platform |
CN107145329A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | Apparatus control method, device and smart machine |
CN107424607A (en) * | 2017-07-04 | 2017-12-01 | 珠海格力电器股份有限公司 | Voice command mode switching method, device and the equipment with the device |
CN108073275A (en) * | 2016-11-18 | 2018-05-25 | 联想(新加坡)私人有限公司 | Information processing method, information processing equipment and program product |
CN108182943A (en) * | 2017-12-29 | 2018-06-19 | 北京奇艺世纪科技有限公司 | A kind of smart machine control method, device and smart machine |
CN108647002A (en) * | 2018-03-30 | 2018-10-12 | 联想(北京)有限公司 | Information processing method and electronic equipment |
-
2018
- 2018-11-28 CN CN201811435993.XA patent/CN109658925A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107004410A (en) * | 2014-10-01 | 2017-08-01 | 西布雷恩公司 | Voice and connecting platform |
CN108073275A (en) * | 2016-11-18 | 2018-05-25 | 联想(新加坡)私人有限公司 | Information processing method, information processing equipment and program product |
CN106776578A (en) * | 2017-01-03 | 2017-05-31 | 竹间智能科技(上海)有限公司 | Talk with the method and device of performance for lifting conversational system |
CN107145329A (en) * | 2017-04-10 | 2017-09-08 | 北京猎户星空科技有限公司 | Apparatus control method, device and smart machine |
CN107424607A (en) * | 2017-07-04 | 2017-12-01 | 珠海格力电器股份有限公司 | Voice command mode switching method, device and the equipment with the device |
CN108182943A (en) * | 2017-12-29 | 2018-06-19 | 北京奇艺世纪科技有限公司 | A kind of smart machine control method, device and smart machine |
CN108647002A (en) * | 2018-03-30 | 2018-10-12 | 联想(北京)有限公司 | Information processing method and electronic equipment |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110182155B (en) * | 2019-05-14 | 2021-04-30 | 中国第一汽车股份有限公司 | Voice control method of vehicle-mounted control system, vehicle-mounted control system and vehicle |
CN110182155A (en) * | 2019-05-14 | 2019-08-30 | 中国第一汽车股份有限公司 | Sound control method, vehicle control syetem and the vehicle of vehicle control syetem |
CN110517678A (en) * | 2019-08-28 | 2019-11-29 | 南昌保莱科技有限公司 | A kind of AI voice answer-back response system of view-based access control model induction |
CN110517678B (en) * | 2019-08-28 | 2022-04-08 | 南昌保莱科技有限公司 | AI voice response system based on visual sense |
CN110706702A (en) * | 2019-10-15 | 2020-01-17 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
CN110706702B (en) * | 2019-10-15 | 2021-11-09 | 芜湖盟博科技有限公司 | Infinite stage multi-turn dialogue method for speech recognition |
CN110751948A (en) * | 2019-10-18 | 2020-02-04 | 珠海格力电器股份有限公司 | Voice recognition method, device, storage medium and voice equipment |
CN113470634B (en) * | 2020-04-28 | 2024-05-17 | 海信集团有限公司 | Voice interaction equipment control method, server and voice interaction equipment |
CN113470634A (en) * | 2020-04-28 | 2021-10-01 | 海信集团有限公司 | Control method of voice interaction equipment, server and voice interaction equipment |
CN111776943A (en) * | 2020-06-18 | 2020-10-16 | 北京起重运输机械设计研究院有限公司 | Voice control method and system for crane |
CN112017629A (en) * | 2020-07-15 | 2020-12-01 | 马上消费金融股份有限公司 | Conversation control method and equipment of voice robot and storage medium |
CN112092820A (en) * | 2020-09-03 | 2020-12-18 | 广州小鹏汽车科技有限公司 | Initialization setting method for vehicle, and storage medium |
CN112787899A (en) * | 2021-01-08 | 2021-05-11 | 青岛海尔特种电冰箱有限公司 | Equipment voice interaction method, computer readable storage medium and refrigerator |
CN112787899B (en) * | 2021-01-08 | 2022-10-28 | 青岛海尔特种电冰箱有限公司 | Equipment voice interaction method, computer readable storage medium and refrigerator |
CN115914453A (en) * | 2021-09-30 | 2023-04-04 | 华为技术有限公司 | Method, terminal and server for operating application program |
CN116662654A (en) * | 2023-05-26 | 2023-08-29 | 惠州市西子湖畔网络有限公司 | Big data-based affinity matching system and method |
CN116662654B (en) * | 2023-05-26 | 2024-04-09 | 惠州市西子湖畔网络有限公司 | Big data-based affinity matching system and method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109658925A (en) | It is a kind of that wake-up vehicle-mounted voice dialogue method and system are exempted from based on context | |
US20220122607A1 (en) | Controlling an engagement state of an agent during a human-machine dialog | |
US9211854B2 (en) | System and method for incorporating gesture and voice recognition into a single system | |
Jiang et al. | Automatic online evaluation of intelligent assistants | |
CN103187055B (en) | Data processing system based on vehicular applications | |
CN109712624A (en) | A kind of more voice assistant coordination approach, device and system | |
CN111639168A (en) | Multi-turn conversation processing method and device, electronic equipment and storage medium | |
CN103187056B (en) | Speech processing system based on vehicular applications | |
CN109637548A (en) | Voice interactive method and device based on Application on Voiceprint Recognition | |
CN110727346A (en) | Man-machine interaction method and device, vehicle and storage medium | |
EP3646318B1 (en) | Electronic device and system for deciding duration of receiving voice input based on context information | |
CN106384591A (en) | Method and device for interacting with voice assistant application | |
Trognon | Speech acts and the logic of mutual understanding | |
CN105912111A (en) | Method for ending voice conversation in man-machine interaction and voice recognition device | |
CN106681499A (en) | Method and device for automobile control based on mobile terminal and electronic device | |
CN109286561A (en) | Vehicle instant communication method and onboard system | |
CN109686368A (en) | Voice wakes up response process method and device, electronic equipment and storage medium | |
CN103187061A (en) | Speech conversational system in vehicle | |
CN110111789A (en) | Voice interactive method, calculates equipment and computer-readable medium at device | |
CN110400564A (en) | A kind of chat robots dialogue management method based on stack | |
CN108882202A (en) | A kind of vehicle-mounted exchange method and device based on smart phone | |
CN112614491A (en) | Vehicle-mounted voice interaction method and device, vehicle and readable medium | |
CN112286485A (en) | Method and device for controlling application through voice, electronic equipment and storage medium | |
CN111091819A (en) | Voice recognition device and method, voice interaction system and method | |
CN109036392A (en) | Robot interactive system |
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 |