CN110517670A - Promote the method and apparatus for waking up performance - Google Patents
Promote the method and apparatus for waking up performance Download PDFInfo
- Publication number
- CN110517670A CN110517670A CN201910801354.9A CN201910801354A CN110517670A CN 110517670 A CN110517670 A CN 110517670A CN 201910801354 A CN201910801354 A CN 201910801354A CN 110517670 A CN110517670 A CN 110517670A
- Authority
- CN
- China
- Prior art keywords
- wake
- speech frame
- input signal
- word
- frame
- 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 51
- 230000002618 waking effect Effects 0.000 title claims abstract description 48
- 230000003044 adaptive effect Effects 0.000 claims abstract description 50
- 238000013528 artificial neural network Methods 0.000 claims abstract description 9
- 230000015654 memory Effects 0.000 claims description 17
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 9
- 230000006854 communication Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 5
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 210000004218 nerve net Anatomy 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 5
- 230000006870 function Effects 0.000 description 8
- 230000001737 promoting effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 230000009467 reduction Effects 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000010295 mobile communication Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000037007 arousal Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 239000007787 solid Substances 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
- G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
-
- 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
- G10L15/16—Speech classification or search using artificial neural networks
-
- 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
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/78—Detection of presence or absence of voice signals
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)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Signal Processing (AREA)
- Telephone Function (AREA)
Abstract
The present invention discloses the method and apparatus for being promoted and waking up performance, wherein a method of it is promoted and wakes up performance, comprising: input signal is detected in the form of speech frame, determines that the speech frame is doubtful speech frame or non-speech frame;Efficient voice section is determined based on the doubtful speech frame and non-speech frame of judgement;The efficient voice section is input to adaptive wake up in model, wherein the adaptive model that wakes up is wake-up word positive example sample and the non-disaggregated model more than one for waking up word negative data and obtain after the deep neural network study of supervision ground based on collection;And the adaptive output for waking up model is obtained, wake-up device is determined whether based on the output.The scheme that the present processes and device provide is added the adaptive model that wakes up and is adaptively waken up to efficient voice section, and so as to effectively promote wake-up rate, especially effect is particularly evident in noisy environment.
Description
Technical field
The invention belongs to voice awakening technology fields, more particularly to promote the method and apparatus for waking up performance.
Background technique
In the related technology, there are no promote wake-up rate to reduce the sample of false wake-up again currently on the market.It wakes up and accidentally calls out
Waking up is shifting relationship, is all to select one and do to optimize currently on the market, and on the basis of keeping a side not drop substantially, optimization is another
One side.In terms of reducing false wake-up, comparing have representative is the secondary verifying of addition;In terms of promoting wake-up rate, mainly reduces and call out
Awake threshold.
Inventor has found during realizing the application, is primarily present following scheme in the prior art:
1, secondary authentication module is added, dual thresholding is set, engine is waken up and is tentatively judged received voice signal,
Secondary verifying is then opened when meeting preset condition, determines whether equipment executes wake operation by the result of secondary verifying;
2, it reduces and wakes up threshold, most simple and fast is exactly to reduce threshold wake-up value, achievees the purpose that be easy wake-up.
Wherein, on the one hand, it is secondary to verify the power consumption that will increase equipment, delay is increased, the interaction of user in practice is influenced
Experience;In addition, there is the risk of collapse in the limited equipment of some calculation resources, it cannot spread to that computing capability is weaker to be set
It is standby upper, also along with the reduction of wake-up rate while reducing rapidly false wake-up.On the other hand, reducing wake-up threshold can be effective
Wake-up rate is promoted, but the promotion of false wake-up can be brought simultaneously.
Summary of the invention
The embodiment of the present invention provides a kind of method and apparatus for being promoted and waking up performance, at least solving above-mentioned technical problem
One of.
In a first aspect, the embodiment of the present invention provides a kind of method for being promoted and waking up performance, comprising: in the form of speech frame pair
Input signal is detected, and determines that the speech frame is doubtful speech frame or non-speech frame;Doubtful speech frame based on judgement and
Non-speech frame determines efficient voice section;The efficient voice section is input to adaptive wake up in model, wherein described adaptive
Waking up model is that wake-up word positive example sample and non-wake-up word negative data based on collection have carried out supervision ground deep neural network
The disaggregated model more than one obtained after study;And the adaptive output for waking up model is obtained, it is determined based on the output
Whether wake-up device.
Second aspect, the embodiment of the present invention provide a kind of device for being promoted and waking up performance, comprising: note determination module, configuration
To detect in the form of speech frame to input signal, determine that the speech frame is doubtful speech frame or non-speech frame;Effectively
Voice segments determining module, is configured to the doubtful speech frame determined and non-speech frame signal determines efficient voice section;Adaptively
Wake-up module is configured to for the efficient voice section to be input to adaptive wake up in model, wherein the adaptive wake-up model
It is to be obtained after wake-up word positive example sample and non-wake-up word negative data based on collection have carried out the deep neural network study of supervision ground
Two disaggregated models taken;And result output module, it is configured to obtain the adaptive output for waking up model, is based on institute
It states output and determines whether wake-up device.
The third aspect provides a kind of electronic equipment comprising: at least one processor, and with described at least one
Manage the memory of device communication connection, wherein the memory is stored with the instruction that can be executed by least one described processor, institute
It states instruction to be executed by least one described processor, so that at least one described processor is able to carry out any embodiment of the present invention
Promotion wake up performance method the step of.
Fourth aspect, the embodiment of the present invention also provide a kind of computer program product, and the computer program product includes
The computer program being stored on non-volatile computer readable storage medium storing program for executing, the computer program include program instruction, when
When described program instruction is computer-executed, the computer is made to execute the side of the promotion wake-up performance of any embodiment of the present invention
The step of method.
The scheme that the present processes and device provide is by determining efficient voice by first detecting to input voice
Then the efficient voice section is input in the adaptive wake-up model of training in advance, is adaptively waken up, Ke Yiyou by section
Effect ground, which is promoted, wakes up accuracy, and the input voice that can constantly collect due to adaptively waking up model user has carried out prison
Superintend and direct it is trained, can be more complicated for usage scenario with so as to the usage scenario according to user is effectively optimized
Environment wake up accuracy it is especially high.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment
Attached drawing be briefly described, it should be apparent that, drawings in the following description are some embodiments of the invention, for ability
For the those of ordinary skill of domain, without creative efforts, it can also be obtained according to these attached drawings other attached
Figure.
Fig. 1 is the flow chart for the method that a kind of promotion that one embodiment of the invention provides wakes up performance;
Fig. 2 is the flow chart for another method for promoting wake-up performance that one embodiment of the invention provides;
Fig. 3 is the flow chart for the method that another promotion that one embodiment of the invention provides wakes up performance;
Fig. 4 is the flow chart for the method that another promotion that one embodiment of the invention provides wakes up performance;
Fig. 5 is the process of a specific example of the method that a kind of promotion that one embodiment of the invention provides wakes up performance
Figure;
Fig. 6 is the block diagram for the device that a kind of promotion that one embodiment of the invention provides wakes up performance;
Fig. 7 is the structural schematic diagram for the electronic equipment that one embodiment of the invention provides.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Referring to FIG. 1, it illustrates the flow chart that the promotion of the application wakes up one embodiment of method of performance, the present embodiment
Promotion wake up performance method can be adapted for the terminal for having Intelligent voice dialog arousal function, such as intelligent sound TV,
Intelligent sound box, Intelligent dialogue toy and other existing intelligent terminals for supporting voice to wake up etc..
As shown in Figure 1, in a step 101, detecting in the form of speech frame to input signal, determine that speech frame is doubtful
Like speech frame or non-speech frame;
Later, in a step 102, efficient voice section is determined based on the doubtful speech frame of judgement and non-speech frame signal;
Then, in step 103, efficient voice section is input to adaptive wake up in model;
Finally, at step 104, the adaptive output for waking up model is obtained, the wake-up device if output is wakes up word, if
Output is non-wake-up word then not wake-up device.
In the present embodiment, for step 101, the device for waking up performance is promoted first in the form of speech frame to each frame
Input signal carries out detection judgement, determines that the speech frame is doubtful speech frame or non-speech frame, i.e., each frame signal can all be judged to
Fixed, it is non-speech frame that the frame that judgement result has, which is the frame that doubtful speech frame has,.Later, it for step 102, is promoted and wakes up performance
Device determines efficient voice section based on the doubtful speech frame and non-speech frame signal of judgement.
Later, for step 103, efficient voice section is input to adaptive wake-up model by the device for promoting wake-up performance
In, wherein the adaptive model that wakes up is the wake-up word positive example sample and non-wake-up word counter-example sample under the various scenes based on collection
This carries out the disaggregated model more than one that obtains after the deep neural network study of supervision ground, by using the several scenes of collection
It is with having carried out supervision trained to the adaptive wake-up model to wake up word positive example sample and non-wake-up word negative data, this can be made
The adaptive model that wakes up can adaptively be made whether the judgement comprising waking up word to the input speech signal under various scenes,
By guiding with having supervision, the adaptive model that wakes up constantly is got a promotion to the wake-up accuracy of various scenes.
Finally, for step 104, the adaptive output for waking up model of device acquisition for waking up performance is promoted, if exporting as wake-up word
Wake up instruction is sent to equipment, does not send wake up instruction to equipment if exporting as non-wake-up word.
The method of the present embodiment determines efficient voice section, then by effective language by first detecting to input voice
Segment is input in the adaptive wake-up model of training in advance, is adaptively waken up, and it is accurate can effectively to promote wake-up
Degree, and the input voice that can constantly collect due to adaptively waking up model user is with having carried out supervision trained, can make
Obtaining it can effectively be optimized according to the usage scenario of user, and the environment more complicated for usage scenario wakes up accuracy
It is especially high.
With further reference to Fig. 2, another provided it illustrates one embodiment of the application promotes the method for wake-up performance
Flow chart.The flow chart is mainly process the step of further limiting to step 102 in Fig. 1.
As shown in Fig. 2, in step 201, if it is determined that occurring first continuously to preset the input signal of frame number being doubtful language
Sound frame determines that the first speech frame in the input signal of the first continuous default frame number is the starting point of efficient voice section;
In step 202, after the starting point for determining efficient voice section, if it is determined that there is the second input for continuously presetting frame number
Signal is non-speech frame or the end speech frame for detecting input signal, is determined in the second continuous input signal for presetting frame number
First speech frame be efficient voice section terminal;
In step 203, the terminal of the starting point based on determining efficient voice section and efficient voice section is from input signal
Select out efficient voice section.
In the present embodiment, for step 201, if promoting the device judgement for waking up performance there is the first continuous default frame
Several input signals is doubtful speech frame, it is determined that the first frame in the input signal of the first continuous default frame number is effective
The starting point of speech frame.Later, for step 202, after the starting point that efficient voice frame has been determined, if promoting the dress for waking up performance
It sets and determines to occur the input signal of the second continuous default frame number and be non-speech frame signal or have been detected by input signal
End speech frame, then can determine the first speech frame in input signal of the second continuous default frame number or change end frame and be
The terminal of efficient voice section.It, can be from this based on the beginning and end of the efficient voice frame of the determination finally, for step 203
Efficient voice section is selected out in input signal.Such as first continuously default frame number can be 10 frames, second continuously default frame number can also
Think that 10 frames or the first continuous default frame number are 20 frames, second presets frame number continuously as 15 frames, and the application does not limit herein
System.In addition, may also can have multiple efficient voice sections in input signal, it can be using one section of one section of ground detection of aforesaid way
Out, there is no limit herein by the application.
The method of the present embodiment is doubtful voice frame signal or non-language by the input signal for determining continuous default frame number
Sound frame signal can determine the beginning and end of one or more efficient voice sections with this, so as to by efficient voice section from
It is selected out in input signal, judges whether there is default wake-up word from the efficient voice section convenient for subsequent, be subsequent judgement pole
The earth reduces workload.
With further reference to Fig. 3, it illustrates another promotions that one embodiment of the application provides to wake up the method for performance
Flow chart.The process for the step of the step of flow chart is primarily directed to after step 101 in Fig. 1 further limits.
As shown in figure 3, in step 301, if it is determined that the input signal for not occurring the first continuous default frame number is doubtful
When speech frame, continue to detect input signal and determine until occurring first continuously to preset the input signal of frame number being doubtful
Like speech frame;
In step 302, after the starting point for determining efficient voice section, if it is determined that not occurring the defeated of the second continuous default frame number
When entering signal and being non-speech frame, continue to detect input signal and determine until there is the defeated of the second continuous default frame number
Entering signal is non-speech frame signal or the end speech frame up to input signal.
It in the present embodiment, is doubtful when there is no appearance first continuously to preset the input signal of frame number for step 301
When speech frame, need always to detect input signal and determine to know that it is successively doubtful for the input of the first continuous frame number occur
Like speech frame or detect the end speech frame of input signal.For step 302, it is determined that after starting point, if it is determined that not going out
When the input signal of existing second continuous default frame number is non-speech frame, continue to detect input signal and determine until going out
The input signal of existing second continuous default frame number is non-speech frame signal or the end speech frame up to input signal.To this
Embodiment can effectively handle various the case where being likely to occur.
In some alternative embodiments, before being detected in the form of speech frame to input signal, method is also wrapped
It includes: input signal is pre-processed to promote signal-to-noise ratio.It is further alternative, input signal is pre-processed to promote letter
It makes an uproar than including: to carry out analog-to-digital conversion to the analog signal of acquisition, and configure sample rate and quantizating index, is converted into digital signal;
And time-domain information, frequency domain information, delay and/or energy loss based on digital signal, to the target direction of digital signal
Voice is enhanced to promote signal-to-noise ratio.
Referring to FIG. 4, it illustrates the flow charts that the promotion that the application another embodiment provides wakes up the method for performance.It should
The process for the step of flow chart is further limited primarily directed to step 104 in Fig. 1.
As shown in figure 4, in step 401, waking up word for presetting based on the adaptive model calculating efficient voice section that wakes up
Confidence level, judge whether confidence level is greater than and default wake up word threshold value;
In step 402, if confidence level is more than or equal to default wake-up word threshold value, wake-up device;
In step 403, if confidence level be less than it is default wake up word threshold value, not wake-up device.
In the present embodiment, for step 401, promote the device for waking up performance has according to what adaptive wake-up model calculated
Voice segments are imitated for the default confidence level for waking up word, then judge whether the confidence level is greater than default wake-up word threshold value.For step
Rapid 402, it, can be with wake-up device if the confidence level is more than or equal to default wake-up word threshold value.For step 403, if the confidence level
Small space is default to wake up word threshold value, then is unable to wake-up device.So as to be sentenced by the confidence level for adaptively waking up model calculating
It is disconnected whether wake-up device.
In some alternative embodiments, efficient voice section is calculated for default wake-up word based on the adaptive model that wakes up
Confidence level includes: to calculate the posterior probability of each speech frame in efficient voice section to obtain each speech frame for default wake-up word
Confidence level;Confidence calculations efficient voice section based on each speech frame is for the default confidence level for waking up word.
Below to some problems encountered in the implementation of the present invention by description inventor and to finally determination
One specific embodiment of scheme is illustrated, so that those skilled in the art more fully understand the scheme of the application.
Inventor has found that the defect of prior art is mainly due to following original after carefully studying to the prior art
Because caused by:
1, secondary verifying, first passage wake-up module are secondary to pass through identification module;Basic thought is the voice that front end obtains
Signal passes through level-one wake-up module, checks current knowledge when the confidence level that present frame wakes up word is more than or equal to default wake-up word threshold value
The decoding result of other module executes wake operation if including to wake up word in the sentence decoded, does not otherwise do any feedback,
Secondary identification module at this time plays decisive role.
2, the identification model recognition accuracy of big parameter amount is high, and whether can effectively identify is real wake-up word,
False wake-up is reduced, but parameter amount is big, calculation amount is necessarily big, and the occupancy of CPU and memory is just come up, certainly delay and power consumption
Thereupon, the feeling of subjective experience is exactly that interaction is blunt, and reaction is slow, and equipment is easy hot;Small parameter amount model calculation amount is small,
But model identification is inaccurate, the identification mistake for waking up word causes wake-up rate to decline, and the audio that can be waken up originally is not due to identifying
Refuse to wake up out, such case maximum probability appears in some cacoepies or noisy environment.
3, threshold wake-up value is reduced in order to promote wake-up rate, threshold wake-up value is to wake up the sequencing that occurs according to word of root to it
What posteriority weighted, threshold value reduces the requirement reduction meant that partial words in word are waken up, causes to delete and replace accidentally to call out
Awake increase.
Those skilled in the art may use following scheme to solve drawbacks described above:
It usually will recognize that the condition for suitably loosening secondary verifying, reduce the size of secondary verifying model, reduce a part meter
Calculation amount, or reduce the reduction ratio of threshold wake-up value.
Inventor has found that these methods can quickly do defect before some excellent during realizing the application
Change, but does not all tackle the problem at its root.For Item drive type industry, everybody consider it is more be that product is quickly fallen
Ground goes to consider the mode that other are more deep so as to no time enough and energy.
The scheme of the application proposes a kind of device for being promoted and waking up performance:
By big data analysis, the usage scenario of product, frequency of use are obtained, the distribution of user's use state is based on depth
Learning art excavates user and is intended to, and real-time monitoring currently and history acoustic environmental information, is converted into oneself by original fixed module
Adaptive learning module, to adapt to different scene demands, false wake-up reduces by 60% and (is based on the basis of calculation amount is not increased
600 hours household scrnario testings);Simultaneously by learning to the above-mentioned information counted on, it is adaptive after module can be with
It effectively monitors rapidly and obtains effective doubtful wake-up voice segments, so as to effectively promote wake-up rate, especially noisy
Effect is particularly evident in environment.
Referring to FIG. 5, it illustrates the flow charts of a specific embodiment of the scheme of the application, it should be noted that with
Although referring to some specific examples in lower embodiment, the scheme being not intended to limit this application.
As shown in figure 5, detailed, steps are as follows:
1, pre-processing is carried out to the signal of equipment end acquisition, carries out analog-to-digital conversion, configures suitable sample rate and quantization
Index is converted into manageable digital signal;
2, pre-processing is done to the collected digital signal of step 1, main purpose is to inhibit noise, enhances voice.It is based on
Signal time domain, frequency domain information are delayed in the communication process of binding signal, and energy loss enhances the voice of target direction,
Promote signal-to-noise ratio;
3, the enhanced signal of step 2 is detected, is based on big data and deep learning, it is less to obtain a number of plies,
Two few depth of assortment neural network models of number of nodes are voice or non-voice by calculating posterior probability differentiation, when after
When testing threshold value of the probability greater than a certain setting, it is determined as doubtful speech frame, is otherwise determined as non-voice, it is doubtful for accumulating several frames all
When like speech frame, it is determined as that voice starts, once voice starts to enter step 3, otherwise continues this step;
4, effective voice segments are sent into wake-up model and are calculated by the detection information obtained based on step 3.Wake-up module
Middle wake-up model is the sample information based on collection, and positive example sample and non-wake-up word negative data comprising waking up word are had
The disaggregated model more than one that the deep neural network study of supervision obtains.In conjunction with current acoustic enviroment, usage scenario, analysis is used
The use at family is intended to, autonomous learning, adaptive adjustment wake-up module.The efficient voice input that step 2 detects is adjusted to call out
In module of waking up, the posteriority of every frame data is calculated, the confidence level for specified wake-up word of present frame is obtained, if the confidence level is big
Then start wake-up mechanism when being equal to the specified preset threshold for waking up word, otherwise enters step 3.
Wake-up module after adaptive can distinguish noise scenarios well, and then can effectively inhibit the mistake of noise field
Wake up, on the basis of calculation amount is not substantially increased, false wake-up reduce by 60%, while if current speech signal due to various originals
Cause, environment is too noisy or human hair sound of speaking is inaccurate etc. cause it is artificial for the first time do not reach when send wake up instruction it is default
Threshold wake-up value starting wakes up, then after adaptive learning, next equal ambient, under equivalent sound even more rugged environment
When sending wake up instruction again, meeting maximum probability gets the wake up instruction of transmission and correctly starts wake-up states, thus very great Cheng
Wake-up rate is improved on degree, to low signal-to-noise ratio, the acquisition promotion of the wake up instruction with accent is especially apparent adaptation mechanism.
Inventor also used following alternative, and summarize the alternative during realizing the application
Merits and demerits.
Although current scheme can promote wake-up performance to a certain extent, calculating is not increased substantially under equal conditions
Amount, can also be a little painstaking but use in the especially deficient equipment of computing resource.
Inventor also attempted following scheme during realizing the application:
One of alternative is the structure of reduced-order models, but the reduction of model structure brings and wakes up under performance
Drop reduces calculation amount so mainstream way still does fixed point to wake-up model at present, can be nervous in script computing resource
It promotes in equipment.
Another alternative is: in terms of reducing false wake-up, being limited to each word waken up in word, if waking up word
In each word reach the confidence level of requirement and the word and be more than or equal to preset threshold value and then open wake-up, otherwise refuse.It is this
The advantage of mode is exactly very effective to false wake-up is reduced, but the requirement for speaker is also relatively stringent, wakes up some in word
Word cacoepy or pronunciation have deviation, pronounce indistinctly, and word speed is too fast to flood one of word, all wake-up rate can be caused not high,
Be not suitable for popularizing, can only be used in certain specific demands.
Above-described embodiment at least can be realized following technical effect:
Wake-up module after adaptive can distinguish noise scenarios well, and then can effectively inhibit the mistake of noise field
Wake up, on the basis of calculation amount is not substantially increased, false wake-up reduce by 60%, while if current speech signal due to various originals
Cause, environment is too noisy or human hair sound of speaking is inaccurate etc. cause it is artificial for the first time do not reach when send wake up instruction it is default
Threshold wake-up value starting wakes up, then after adaptive learning, next equal ambient, under equivalent sound even more rugged environment
When sending wake up instruction again, meeting maximum probability gets the wake up instruction of transmission and correctly starts wake-up states, thus very great Cheng
Wake-up rate is improved on degree, to low signal-to-noise ratio, the acquisition promotion of the wake up instruction with accent is especially apparent adaptation mechanism.
Referring to FIG. 6, it illustrates the block diagrams that the promotion that one embodiment of the invention provides wakes up the device of performance.
As shown in fig. 6, promoting the device 600 for waking up performance, module 610, efficient voice section determining module are included determining whether
620, adaptive wake-up module 630 and result output module 640.
Wherein, determination module 610 are configured to detect input signal in the form of speech frame, determine the voice
Frame is doubtful speech frame or non-speech frame;Efficient voice section determining module 620 is configured to the doubtful speech frame and non-determined
Voice frame signal determines efficient voice section;Adaptive wake-up module 630, is configured to for the efficient voice section being input to adaptively
It wakes up in model, wherein the adaptive wake-up model is wake-up word positive example sample and non-wake-up word counter-example sample based on collection
This carries out two disaggregated models obtained after the deep neural network study of supervision ground;And result output module 640, configuration
To obtain the adaptive output for waking up model, wake-up device is determined whether based on the output.
It should be appreciated that all modules recorded in Fig. 6 with reference to each in method described in Fig. 1, Fig. 2, Fig. 3 and Fig. 4
Step is corresponding.The operation above with respect to method description and feature and corresponding technical effect are equally applicable in Fig. 6 as a result,
All modules, details are not described herein.
It is worth noting that, the scheme that the module in embodiments herein is not intended to limit this application, such as determine
Module can be described as detecting input signal in the form of speech frame, determine that the speech frame is doubtful speech frame or non-
The module of speech frame.Furthermore it is also possible to realize related function module by hardware processor, such as determination module can also be used
Processor realizes that details are not described herein.
In further embodiments, the embodiment of the invention also provides a kind of nonvolatile computer storage medias, calculate
Machine storage medium is stored with computer executable instructions, which can be performed in above-mentioned any means embodiment
Promotion wake up performance method;
As an implementation, nonvolatile computer storage media of the invention is stored with the executable finger of computer
It enables, computer executable instructions setting are as follows:
Input signal is detected in the form of speech frame, determines that the speech frame is doubtful speech frame or non-voice
Frame;
Efficient voice section is determined based on the doubtful speech frame and non-speech frame of judgement;
The efficient voice section is input to adaptive wake up in model, wherein the adaptive wake-up model is to be based on
The wake-up word positive example sample of collection and the non-word negative data that wakes up carry out obtain after supervision ground deep neural network learns one
A more disaggregated models;
The adaptive output for waking up model is obtained, wake-up device is determined whether based on the output.
Non-volatile computer readable storage medium storing program for executing may include storing program area and storage data area, wherein storage journey
It sequence area can application program required for storage program area, at least one function;Storage data area can be stored to be waken up according to promotion
The device of performance uses created data etc..In addition, non-volatile computer readable storage medium storing program for executing may include high speed with
Machine access memory, can also include nonvolatile memory, a for example, at least disk memory, flush memory device or its
His non-volatile solid state memory part.In some embodiments, it includes opposite that non-volatile computer readable storage medium storing program for executing is optional
In the remotely located memory of processor, these remote memories can be by being connected to the network to the device for promoting wake-up performance.
The example of above-mentioned network includes but is not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
The embodiment of the present invention also provides a kind of computer program product, and computer program product is non-volatile including being stored in
Computer program on computer readable storage medium, computer program include program instruction, when program instruction is held by computer
When row, so that computer is executed any of the above-described and promote the method for waking up performance.
Fig. 7 is the structural schematic diagram of electronic equipment provided in an embodiment of the present invention, as shown in fig. 7, the equipment includes: one
Or multiple processors 710 and memory 720, in Fig. 7 by taking a processor 710 as an example.Promote setting for the method for waking up performance
Standby can also include: input unit 730 and output device 740.Processor 710, memory 720, input unit 730 and output dress
Setting 740 can be connected by bus or other modes, in Fig. 7 for being connected by bus.Memory 720 is above-mentioned non-
Volatile computer readable storage medium storing program for executing.The non-volatile software journey that processor 710 is stored in memory 720 by operation
Sequence, instruction and module, thereby executing the various function application and data processing of server, i.e. realization above method embodiment
Promote the method for waking up performance.Input unit 730 can receive the number or character information of input, and generates and promote wake-up property
The related key signals input of the user setting and function control of the device of energy.Output device 740 may include the display such as display screen
Equipment.
Method provided by the embodiment of the present invention can be performed in the said goods, has the corresponding functional module of execution method and has
Beneficial effect.The not technical detail of detailed description in the present embodiment, reference can be made to method provided by the embodiment of the present invention.
As an implementation, above-mentioned electronic apparatus application is in the device for promoting wake-up performance, comprising: at least one
Processor;And the memory being connect at least one processor communication;Wherein, be stored with can be by least one for memory
Manage device execute instruction, instruction executed by least one processor so that at least one processor can:
Input signal is detected in the form of speech frame, determines that the speech frame is doubtful speech frame or non-voice
Frame;
Efficient voice section is determined based on the doubtful speech frame and non-speech frame of judgement;
The efficient voice section is input to adaptive wake up in model, wherein the adaptive wake-up model is to be based on
The wake-up word positive example sample of collection and the non-word negative data that wakes up carry out obtain after supervision ground deep neural network learns one
A more disaggregated models;
The adaptive output for waking up model is obtained, wake-up device is determined whether based on the output.
The electronic equipment of the embodiment of the present application exists in a variety of forms, including but not limited to:
(1) mobile communication equipment: the characteristics of this kind of equipment is that have mobile communication function, and to provide speech, data
Communication is main target.This Terminal Type includes: smart phone (such as iPhone), multimedia handset, functional mobile phone and low
Hold mobile phone etc..
(2) super mobile personal computer equipment: this kind of equipment belongs to the scope of personal computer, there is calculating and processing function
Can, generally also have mobile Internet access characteristic.This Terminal Type includes: PDA, MID and UMPC equipment etc., such as iPad.
(3) portable entertainment device: this kind of equipment can show and play multimedia content.Such equipment include: audio,
Video player (such as iPod), handheld device, e-book and intelligent toy and portable car-mounted navigation equipment.
(4) server: providing the equipment of the service of calculating, and the composition of server includes that processor, hard disk, memory, system are total
Line etc., server is similar with general computer architecture, but due to needing to provide highly reliable service, in processing energy
Power, stability, reliability, safety, scalability, manageability etc. are more demanding.
(5) other electronic devices with data interaction function.
The apparatus embodiments described above are merely exemplary, wherein unit can be as illustrated by the separation member
Or may not be and be physically separated, component shown as a unit may or may not be physical unit, i.e.,
It can be located in one place, or may be distributed over multiple network units.It can select according to the actual needs therein
Some or all of the modules achieves the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creative labor
In the case where dynamic, it can understand and implement.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features;
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of promote the method for waking up performance, comprising:
Input signal is detected in the form of speech frame, determines that the speech frame is doubtful speech frame or non-speech frame;
Efficient voice section is determined based on the doubtful speech frame and non-speech frame of judgement;
The efficient voice section is input to adaptive wake up in model, wherein the adaptive wake-up model is based on collection
Wake-up word positive example sample and it is non-wake up word negative data carry out supervision ground deep neural network study after obtain more than one
Disaggregated model;
The adaptive output for waking up model is obtained, wake-up device is determined whether based on the output.
2. according to the method described in claim 1, wherein the doubtful speech frame and non-speech frame based on judgement determines effectively
Voice segments include:
If it is determined that occurring first continuously to preset the input signal of frame number being doubtful speech frame, determine that described first continuously presets frame
First speech frame in several input signals is the starting point of the efficient voice section;
After the starting point for determining the efficient voice section, if it is determined that occurring second continuously to preset the input signal of frame number being non-language
Sound frame or the end speech frame for detecting the input signal determine the head in the input signal of the described second continuous default frame number
A speech frame is the terminal of the efficient voice section;
The terminal of starting point and the efficient voice section based on the determining efficient voice section is selected out from the input signal
The efficient voice section.
3. according to the method described in claim 2, wherein, determining that the speech frame is doubtful speech frame or non-speech frame described
Later, the method also includes:
If it is determined that continuing when the input signal for not occurring the first continuous default frame number is doubtful speech frame to the input signal
It is detected and is determined until occurring first continuously to preset the input signal of frame number being doubtful speech frame;
After the starting point for determining the efficient voice section, if it is determined that the input signal for not occurring the second continuous default frame number is non-
When speech frame, continue to detect the input signal and determine until there is the input letter of the second continuous default frame number
Number be non-speech frame signal or until the input signal end speech frame.
4. according to the method described in claim 1, wherein, carrying out detecting it to input signal in the form of speech frame described
Before, the method also includes:
Input signal is pre-processed to promote signal-to-noise ratio.
5. described to be pre-processed to input signal to promote signal-to-noise ratio and include: according to the method described in claim 4, wherein
Analog-to-digital conversion is carried out to the analog signal of acquisition, and configures sample rate and quantizating index, is converted into digital signal;
Time-domain information, frequency domain information, delay and/or energy loss based on the digital signal, to the mesh of the digital signal
The voice in mark direction is enhanced to promote signal-to-noise ratio.
6. method according to any one of claims 1-5, wherein described to obtain the defeated of the adaptive wake-up model
Out, the wake-up device if the output is wakes up word, determines whether that wake-up device includes: based on the output
The adaptive efficient voice section for waking up model output is obtained for the default confidence level for waking up word, described in judgement
Whether confidence level is greater than default wake-up word threshold value;
If the confidence level is more than or equal to the default wake-up word threshold value, the equipment is waken up;
If the confidence level is less than the default wake-up word threshold value, the equipment is not waken up.
7. described based on adaptive the wake-ups model calculating efficient voice according to the method described in claim 6, wherein
Section includes: for the default confidence level for waking up word
The posterior probability of each speech frame in the efficient voice section is calculated to obtain each speech frame for default wake-up word
Confidence level;
Efficient voice section described in confidence calculations based on each speech frame is for the default confidence level for waking up word.
8. a kind of promote the device for waking up performance, comprising:
Determination module is configured to detect input signal in the form of speech frame, determines that the speech frame is doubtful voice
Frame or non-speech frame;
Efficient voice section determining module, is configured to the doubtful speech frame determined and non-speech frame signal determines efficient voice
Section;
Adaptive wake-up module is configured to for the efficient voice section to be input to adaptive wake up in model, wherein described adaptive
It is that wake-up word positive example sample and non-wake-up word negative data based on collection have carried out supervision ground depth nerve net that model, which should be waken up,
Two disaggregated models obtained after network study;
As a result output module is configured to obtain the adaptive output for waking up model, determines whether to wake up based on the output
Equipment.
9. a kind of electronic equipment comprising: at least one processor, and deposited with what at least one described processor communication was connect
Reservoir, wherein the memory be stored with can by least one described processor execute instruction, described instruction by it is described at least
One processor executes, so that at least one described processor is able to carry out the step of any one of claim 1 to 7 the method
Suddenly.
10. a kind of storage medium, is stored thereon with computer program, which is characterized in that real when described program is executed by processor
The step of any one of existing claim 1 to 7 the method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910801354.9A CN110517670A (en) | 2019-08-28 | 2019-08-28 | Promote the method and apparatus for waking up performance |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910801354.9A CN110517670A (en) | 2019-08-28 | 2019-08-28 | Promote the method and apparatus for waking up performance |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110517670A true CN110517670A (en) | 2019-11-29 |
Family
ID=68628694
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910801354.9A Pending CN110517670A (en) | 2019-08-28 | 2019-08-28 | Promote the method and apparatus for waking up performance |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110517670A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807909A (en) * | 2019-12-09 | 2020-02-18 | 深圳云端生活科技有限公司 | Radar and voice processing combined control method |
CN111223497A (en) * | 2020-01-06 | 2020-06-02 | 苏州思必驰信息科技有限公司 | Nearby wake-up method and device for terminal, computing equipment and storage medium |
CN111599371A (en) * | 2020-05-19 | 2020-08-28 | 苏州奇梦者网络科技有限公司 | Voice adding method, system, device and storage medium |
CN111596882A (en) * | 2020-04-02 | 2020-08-28 | 云知声智能科技股份有限公司 | Distributed array alignment method |
CN111653274A (en) * | 2020-04-17 | 2020-09-11 | 北京声智科技有限公司 | Method, device and storage medium for awakening word recognition |
CN111653276A (en) * | 2020-06-22 | 2020-09-11 | 四川长虹电器股份有限公司 | Voice awakening system and method |
CN111833869A (en) * | 2020-07-01 | 2020-10-27 | 中关村科学城城市大脑股份有限公司 | Voice interaction method and system applied to urban brain |
CN111899722A (en) * | 2020-08-11 | 2020-11-06 | Oppo广东移动通信有限公司 | Voice processing method and device and storage medium |
CN112114886A (en) * | 2020-09-17 | 2020-12-22 | 北京百度网讯科技有限公司 | Method and device for acquiring false wake-up audio |
CN112151015A (en) * | 2020-09-03 | 2020-12-29 | 腾讯科技(深圳)有限公司 | Keyword detection method and device, electronic equipment and storage medium |
CN112669822A (en) * | 2020-12-16 | 2021-04-16 | 爱驰汽车有限公司 | Audio processing method and device, electronic equipment and storage medium |
EP3923272A1 (en) * | 2020-06-10 | 2021-12-15 | Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method and apparatus for adapting a wake-up model |
CN117012206A (en) * | 2023-10-07 | 2023-11-07 | 山东省智能机器人应用技术研究院 | Man-machine voice interaction system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105632486A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Voice wake-up method and device of intelligent hardware |
US20170323202A1 (en) * | 2016-05-06 | 2017-11-09 | Fujitsu Limited | Recognition apparatus based on deep neural network, training apparatus and methods thereof |
CN107871506A (en) * | 2017-11-15 | 2018-04-03 | 北京云知声信息技术有限公司 | The awakening method and device of speech identifying function |
CN108010515A (en) * | 2017-11-21 | 2018-05-08 | 清华大学 | A kind of speech terminals detection and awakening method and device |
CN108877778A (en) * | 2018-06-13 | 2018-11-23 | 百度在线网络技术(北京)有限公司 | Sound end detecting method and equipment |
-
2019
- 2019-08-28 CN CN201910801354.9A patent/CN110517670A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105632486A (en) * | 2015-12-23 | 2016-06-01 | 北京奇虎科技有限公司 | Voice wake-up method and device of intelligent hardware |
US20170323202A1 (en) * | 2016-05-06 | 2017-11-09 | Fujitsu Limited | Recognition apparatus based on deep neural network, training apparatus and methods thereof |
CN107871506A (en) * | 2017-11-15 | 2018-04-03 | 北京云知声信息技术有限公司 | The awakening method and device of speech identifying function |
CN108010515A (en) * | 2017-11-21 | 2018-05-08 | 清华大学 | A kind of speech terminals detection and awakening method and device |
CN108877778A (en) * | 2018-06-13 | 2018-11-23 | 百度在线网络技术(北京)有限公司 | Sound end detecting method and equipment |
Cited By (23)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110807909A (en) * | 2019-12-09 | 2020-02-18 | 深圳云端生活科技有限公司 | Radar and voice processing combined control method |
CN111223497A (en) * | 2020-01-06 | 2020-06-02 | 苏州思必驰信息科技有限公司 | Nearby wake-up method and device for terminal, computing equipment and storage medium |
CN111223497B (en) * | 2020-01-06 | 2022-04-19 | 思必驰科技股份有限公司 | Nearby wake-up method and device for terminal, computing equipment and storage medium |
CN111596882A (en) * | 2020-04-02 | 2020-08-28 | 云知声智能科技股份有限公司 | Distributed array alignment method |
CN111596882B (en) * | 2020-04-02 | 2023-05-26 | 云知声智能科技股份有限公司 | Distributed array alignment method |
CN111653274A (en) * | 2020-04-17 | 2020-09-11 | 北京声智科技有限公司 | Method, device and storage medium for awakening word recognition |
CN111653274B (en) * | 2020-04-17 | 2023-08-04 | 北京声智科技有限公司 | Wake-up word recognition method, device and storage medium |
CN111599371A (en) * | 2020-05-19 | 2020-08-28 | 苏州奇梦者网络科技有限公司 | Voice adding method, system, device and storage medium |
CN111599371B (en) * | 2020-05-19 | 2023-10-20 | 苏州奇梦者网络科技有限公司 | Voice adding method, system, device and storage medium |
EP3923272A1 (en) * | 2020-06-10 | 2021-12-15 | Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method and apparatus for adapting a wake-up model |
US11587550B2 (en) | 2020-06-10 | 2023-02-21 | Apollo Intelligent Connectivity (Beijing) Technology Co., Ltd. | Method and apparatus for outputting information |
CN111653276A (en) * | 2020-06-22 | 2020-09-11 | 四川长虹电器股份有限公司 | Voice awakening system and method |
CN111833869B (en) * | 2020-07-01 | 2022-02-11 | 中关村科学城城市大脑股份有限公司 | Voice interaction method and system applied to urban brain |
CN111833869A (en) * | 2020-07-01 | 2020-10-27 | 中关村科学城城市大脑股份有限公司 | Voice interaction method and system applied to urban brain |
CN111899722A (en) * | 2020-08-11 | 2020-11-06 | Oppo广东移动通信有限公司 | Voice processing method and device and storage medium |
CN111899722B (en) * | 2020-08-11 | 2024-02-06 | Oppo广东移动通信有限公司 | Voice processing method and device and storage medium |
CN112151015A (en) * | 2020-09-03 | 2020-12-29 | 腾讯科技(深圳)有限公司 | Keyword detection method and device, electronic equipment and storage medium |
CN112151015B (en) * | 2020-09-03 | 2024-03-12 | 腾讯科技(深圳)有限公司 | Keyword detection method, keyword detection device, electronic equipment and storage medium |
CN112114886A (en) * | 2020-09-17 | 2020-12-22 | 北京百度网讯科技有限公司 | Method and device for acquiring false wake-up audio |
CN112114886B (en) * | 2020-09-17 | 2024-03-29 | 北京百度网讯科技有限公司 | Acquisition method and device for false wake-up audio |
CN112669822A (en) * | 2020-12-16 | 2021-04-16 | 爱驰汽车有限公司 | Audio processing method and device, electronic equipment and storage medium |
CN117012206A (en) * | 2023-10-07 | 2023-11-07 | 山东省智能机器人应用技术研究院 | Man-machine voice interaction system |
CN117012206B (en) * | 2023-10-07 | 2024-01-16 | 山东省智能机器人应用技术研究院 | Man-machine voice interaction system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110517670A (en) | Promote the method and apparatus for waking up performance | |
CN110473539A (en) | Promote the method and apparatus that voice wakes up performance | |
US11127416B2 (en) | Method and apparatus for voice activity detection | |
CN110364143B (en) | Voice awakening method and device and intelligent electronic equipment | |
CN105632486B (en) | Voice awakening method and device of intelligent hardware | |
CN108597505B (en) | Voice recognition method and device and terminal equipment | |
CN108899044A (en) | Audio signal processing method and device | |
DE102018126133A1 (en) | Generate dialog based on verification values | |
CN107767863A (en) | voice awakening method, system and intelligent terminal | |
CN110570873B (en) | Voiceprint wake-up method and device, computer equipment and storage medium | |
CN107360157A (en) | A kind of user registering method, device and intelligent air conditioner | |
CN110600008A (en) | Voice wake-up optimization method and system | |
CN112562742B (en) | Voice processing method and device | |
CN110335593A (en) | Sound end detecting method, device, equipment and storage medium | |
CN108917283A (en) | A kind of intelligent refrigerator control method, system, intelligent refrigerator and cloud server | |
US20080215318A1 (en) | Event recognition | |
CN111722696B (en) | Voice data processing method and device for low-power-consumption equipment | |
CN109697981B (en) | Voice interaction method, device, equipment and storage medium | |
CN111145763A (en) | GRU-based voice recognition method and system in audio | |
CN111179915A (en) | Age identification method and device based on voice | |
CN110503944A (en) | The training of voice wake-up model and application method and device | |
CN111323783A (en) | Scene recognition method and device, storage medium and electronic equipment | |
CN109994129A (en) | Speech processing system, method and apparatus | |
CN109377993A (en) | Intelligent voice system and its voice awakening method and intelligent sound equipment | |
CN110197663B (en) | Control method and device and electronic equipment |
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 | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 215123 building 14, Tengfei Innovation Park, 388 Xinping street, Suzhou Industrial Park, Suzhou City, Jiangsu Province Applicant after: Sipic Technology Co.,Ltd. Address before: 215123 building 14, Tengfei Innovation Park, 388 Xinping street, Suzhou Industrial Park, Suzhou City, Jiangsu Province Applicant before: AI SPEECH Co.,Ltd. |
|
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191129 |