CN207458576U - A kind of real-time equipment for replacing wake-up word - Google Patents
A kind of real-time equipment for replacing wake-up word Download PDFInfo
- Publication number
- CN207458576U CN207458576U CN201721085378.1U CN201721085378U CN207458576U CN 207458576 U CN207458576 U CN 207458576U CN 201721085378 U CN201721085378 U CN 201721085378U CN 207458576 U CN207458576 U CN 207458576U
- Authority
- CN
- China
- Prior art keywords
- wake
- voice
- word
- circuit
- speech recognition
- 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.)
- Active
Links
Abstract
The utility model is related to field of speech recognition,Especially relate to a kind of real-time equipment for replacing wake-up word,The equipment includes speech signal collection process circuit,Voice wake-up control circuit,Speech recognition and network transmission circuit and high in the clouds training server,The speech signal collection process circuit gathers voice data,And the voice data is handled,The voice wake-up control circuit is attached with the speech signal collection process circuit and receives the voice data come from speech signal collection process circuit transmission,The speech recognition and network transmission circuit are bi-directionally connected with the voice wake-up control circuit,The high in the clouds training server is bi-directionally connected with the speech recognition and network transmission circuit,The utility model, which has used, can allow user directly to use the self-defined required wake-up word of voice,And ensure the wake-up word independent of some particular person by way of big data training,The wake-up of unspecified person can be realized after change,Improve the usage experience sense of user.
Description
Technical field
The utility model is related to field of speech recognition, especially relate to a kind of by offline and online speech recognition technology knot
The novel voice identification equipment of conjunction.
Background technology
Speech recognition at present is widely used in each electronic product as a kind of common human-computer interaction technology, with
Its naturally convenient interactive mode receives liking for consumer, has been increasingly becoming the mainstream interactive controlling side in intellectual product epoch
Formula, current speech recognition product facility can by the way of word is waken up, i.e., speech ciphering equipment normal use when be in standby shape
State will not respond the voice in environmental noise and user's normal life.Only user is telling equipment using voice
During the wake-up word pre-defined, speech ciphering equipment can be just waken up, starts that processing is identified to extraneous voice signal.So doing can
To reduce the misrecognition of speech recognition apparatus, widely used in current Related product equipment, this method is in the application
There are problems that, be to be cured to by producer before dispatching from the factory in product facility due to waking up word, user cannot change, therefore be difficult to full
The demand that sufficient user personality customizes, if single user's family has bought the speech recognition apparatus of multiple same models simultaneously, if
These equipment all have identical wake-up word, may be caused when user wakes up by the equipment for being not desired to wake up to false triggering
Puzzlement in use, this just needs to provide a kind of mode for allowing user that can voluntarily change wake-up word, to meet the needs of client,
Some current speech recognition apparatus can allow the wake-up word that user wants by recording to input user, then directly against the recording
Identify, but which be only capable of it is effective to the user of recording, when substitution is waken up, often because everyone phonetic feature
Difference, individually comparing the mode of the wake-up word phonetic feature of recording can cause to wake up failure, reduce the experience sense of user.
It wakes up word to solve the above problems, needing to invent a kind of can freely change but also makes the wake-up word of change from spy
Determine the real-time of the limitation of people and replace the device and method for waking up word.
Utility model content
A kind of real-time equipment for replacing wake-up word, including：
Speech signal collection process circuit, the speech signal collection process circuit gather voice data, and to the sound
Sound data carry out the processing such as noise reduction and enhancing；
Voice wake-up control circuit, the voice wake-up control circuit carry out single with the speech signal collection process circuit
To connection, i.e., described voice data is transmitted to the voice wake-up control circuit from the speech signal collection process circuit；
Speech recognition and network transmission circuit, the speech recognition and network transmission circuit wake up control electricity with the voice
Road is bi-directionally connected, i.e., described speech recognition and network transmission circuit can both have been transmitted to the voice wake-up control circuit data or
Data can be received from the voice wake-up control circuit；
High in the clouds training server, the high in the clouds training server carry out two-way with the speech recognition and network transmission circuit
Connection, i.e., described high in the clouds training server can think the speech recognition and network transmission circuit transmission data or receive
Data are received from the speech recognition and network transmission circuit.
As the preferred embodiment of the utility model, the voice wake-up control circuit is built-in with memory, the memory
The wake-up word pre-set is inside placed with, the voice wake-up control circuit decodes the voice data, and identifies institute's predicate
Whether sound data are the wake-up words, and the voice wake-up control circuit is combined and can actively repaiied with the high in the clouds training server
Change the wake-up word in the memory space.
As the preferred embodiment of the utility model, the voice wake-up control circuit controls the speech recognition and network to pass
The open and close on transmission of electricity road, it is described when the speech signal collection process circuit judges the voice data to wake up word
Speech recognition and network transmission circuit are opened, on the contrary then close.
The utility model realizes that process is as follows：
The extraneous voice signal of the speech signal collection process circuit acquisition simultaneously drops the voice signal
It makes an uproar and enhancing processing forms new voice signal and the new voice signal is sent to voice wake-up control circuit, institute's predicate
Sound wake-up control circuit detects whether the new voice signal is effective wake-up word, and word then institute is waken up if not effective
Predicate sound signal acquisition process circuit continues the extraneous voice signal of acquisition, then will be described effective if effectively waking up word
It wakes up word to be transferred to the speech recognition and network transmission circuit and open the speech recognition and network transmission circuit, institute's predicate
Sound identifies and network transmission circuit sends indicator, and user issues real-time voice according to the indicator and instructs, the voice
Signal acquisition processing circuit gathers the real-time voice and instructs and be sent to the speech recognition and network transmission circuit, institute's predicate
Sound identifies and network transmission circuit detects whether the real-time voice instruction is effective phonetic control command, if described real-time
Phonetic order is that effective phonetic control command then performs corresponding control operation, if real-time voice instruction is invalid
, then operation is terminated, if real-time voice instruction is the instruction that preset change wakes up word, user inputs according to prompting
The new wake-up word to be changed, the speech signal collection process circuit gather and handle the new wake-up word, the speech recognition
The new wake-up word come from speech signal collection process circuit transmission is received with network transmission circuit, and is newly called out described
Awake word sends high in the clouds training server to, and the high in the clouds training server handles the new wake-up word, the high in the clouds instruction
Practice server and the new wake-up word is converted into text, and given birth to according to the speech model that the text is transferred by big data training
The speech recognition and network transmission circuit are transferred into the wake-up word of user demand, then by the wake-up word of the user demand,
The wake-up word of the user demand can realize the wake-up command of unspecified person, the speech recognition and network transmission circuit by institute
The wake-up word for stating user demand is transferred to the voice wake-up control circuit, and the voice wake-up control circuit needs the user
The wake-up word asked, which is written in built-in memory space, overrides original wake-up word, realizes the replacement for waking up word.
Compared with prior art, the beneficial effects of the utility model：
The utility model can allow user directly to use the self-defined required wake-up word of voice, and pass through the side that big data is trained
Formula has ensured that the wake-up word independent of some particular person, can realize the wake-up of unspecified person, improve user's after change
Usage experience sense.
Description of the drawings
Fig. 1 is the utility model block diagram；
Fig. 2 is the utility model embodiment block diagram.
Specific embodiment
The utility model is described in further detail with reference to embodiment and specific embodiment, but should not be by this
The scope for being interpreted as the utility model aforementioned body is only limitted to following embodiment, all to be realized based on the utility model content
Technology belongs to the scope of the utility model.
As shown in Fig. 2, a kind of real-time equipment for replacing wake-up word, is called out including speech signal collection process circuit, voice
Awake control circuit, speech recognition and network transmission circuit and high in the clouds training server, speech signal collection process circuit acquisition sound
Sound data, and voice data is handled, voice wake-up control circuit is attached simultaneously with speech signal collection process circuit
The voice data come from the transmission of speech signal collection process circuit is received, speech recognition and network transmission circuit wake up with voice to be controlled
Circuit processed is bi-directionally connected, and high in the clouds training server is bi-directionally connected with speech recognition and network transmission circuit, in the present embodiment
Middle speech signal collection process circuit is made of microphone, Dolby circuit and enhancing circuit, the extraneous sound number of microphone acquisition
According to and by Dolby circuit and enhancing processing of circuit voice data and generate new voice data, speech signal collection process circuit
Give above-mentioned new data transmission in network telephony to voice wake-up control circuit, voice wake-up control circuit is according to above-mentioned new voice data
Control voice identification control and the unlatching of network transmission circuit, speech recognition and network transmission circuit transmission data are trained to high in the clouds
Server, high in the clouds training server will send back voice through speech recognition and network transmission circuit after data processing and wake up control electricity
Road covers original wake-up word.
Claims (3)
1. a kind of real-time equipment for replacing wake-up word, including：
Speech signal collection process circuit, the speech signal collection process circuit gather voice data, and to the sound number
According to being handled；
Voice wake-up control circuit, the voice wake-up control circuit are attached simultaneously with the speech signal collection process circuit
Receive the voice data come from speech signal collection process circuit transmission；
Speech recognition and network transmission circuit, the speech recognition and network transmission circuit are double with the voice wake-up control circuit
To connection；
High in the clouds training server, the high in the clouds training server carry out two-way company with the speech recognition and network transmission circuit
It connects.
2. a kind of real-time equipment for replacing wake-up word according to claim 1, it is characterised in that：The voice wakes up control
Circuit is built-in with memory, and the wake-up word pre-set is placed in the memory, and the wake-up word decodes the sound
Data, and identify whether the voice data is to wake up word, the voice wake-up control circuit and the high in the clouds training server
With reference to the wake-up word that can actively change in the memory space.
3. a kind of real-time equipment for replacing wake-up word according to claim 2, it is characterised in that：The voice wakes up control
Circuit controls the speech recognition and the open and close of network transmission circuit, when the speech signal collection process circuit judges
When the voice data is wakes up word, the speech recognition and network transmission circuit are opened, on the contrary then close.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201721085378.1U CN207458576U (en) | 2017-08-29 | 2017-08-29 | A kind of real-time equipment for replacing wake-up word |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201721085378.1U CN207458576U (en) | 2017-08-29 | 2017-08-29 | A kind of real-time equipment for replacing wake-up word |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN207458576U true CN207458576U (en) | 2018-06-05 |
Family
ID=62248460
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201721085378.1U Active CN207458576U (en) | 2017-08-29 | 2017-08-29 | A kind of real-time equipment for replacing wake-up word |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN207458576U (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358954A (en) * | 2017-08-29 | 2017-11-17 | 成都启英泰伦科技有限公司 | It is a kind of to change the device and method for waking up word in real time |
| CN109859757A (en) * | 2019-03-19 | 2019-06-07 | 百度在线网络技术（北京）有限公司 | A kind of speech ciphering equipment control method, device and terminal |
| CN111354341A (en) * | 2018-12-04 | 2020-06-30 | 阿里巴巴集团控股有限公司 | Voice awakening method and device, processor, sound box and television |
-
2017
- 2017-08-29 CN CN201721085378.1U patent/CN207458576U/en active Active
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107358954A (en) * | 2017-08-29 | 2017-11-17 | 成都启英泰伦科技有限公司 | It is a kind of to change the device and method for waking up word in real time |
| CN111354341A (en) * | 2018-12-04 | 2020-06-30 | 阿里巴巴集团控股有限公司 | Voice awakening method and device, processor, sound box and television |
| CN109859757A (en) * | 2019-03-19 | 2019-06-07 | 百度在线网络技术（北京）有限公司 | A kind of speech ciphering equipment control method, device and terminal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107358954A (en) | It is a kind of to change the device and method for waking up word in real time | |
| WO2018059405A1 (en) | Voice control system, wakeup method and wakeup apparatus therefor, electrical appliance and co-processor | |
| CN207458576U (en) | A kind of real-time equipment for replacing wake-up word | |
| CN105493180B (en) | Electronic device and method for speech recognition | |
| CN107339786B (en) | A kind of system and method for air-conditioning, regulation air-conditioning loudspeaker casting volume | |
| CN107564518A (en) | Smart machine control method, device and computer equipment | |
| CN106465006B (en) | The operating method of microphone and the electronic equipment for supporting this method | |
| US11295760B2 (en) | Method, apparatus, system and storage medium for implementing a far-field speech function | |
| WO2017206725A1 (en) | Smart refrigerator, server and language control system and method | |
| CN106814639A (en) | Speech control system and method | |
| CN107277904A (en) | A kind of terminal and voice awakening method | |
| CN108877802A (en) | A kind of circuit, TV and the sound control method of integrated speech arousal function | |
| KR20170103925A (en) | Speech identification system and identification method of a kind of robot system | |
| CN102932212A (en) | Intelligent household control system based on multichannel interaction manner | |
| KR102563817B1 (en) | Method for processing user voice input and electronic device supporting the same | |
| CN107147792A (en) | A kind of method for automatically configuring audio, device, mobile terminal and storage device | |
| US11631406B2 (en) | Method for responding to user utterance and electronic device for supporting same | |
| WO2018006374A1 (en) | Function recommending method, system, and robot based on automatic wake-up | |
| CN110033776A (en) | A kind of virtual image interactive system and method applied to screen equipment | |
| CN109474658A (en) | Electronic equipment, server and the recording medium of task run are supported with external equipment | |
| KR102391298B1 (en) | electronic device providing speech recognition service and method thereof | |
| CN204678698U (en) | A kind of voice-operated gas heater | |
| CN106057201A (en) | Household electrical appliance intelligent voice interaction control method and apparatus | |
| WO2018023515A1 (en) | Gesture and emotion recognition home control system | |
| CN108965605A (en) | The control method and device of smart phone |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |