Content of the invention
The technical problem to be solved is for the deficiencies in the prior art, provides a kind of language based on sound localization
Voice recognition method and system and intelligent appliance equipment.
The technical scheme is that a kind of audio recognition method based on sound localization,
Comprise the following steps:
Step 1, when intelligent appliance equipment enters speech recognition state, gather voice signal, and control timer start
Timing;
Step 2, when described intervalometer timing be less than or equal to Preset Time when collect described voice signal, then root
Positioned according to described voice signal, determined the positional information of sound source corresponding to described voice signal;
Step 3, according to described positional information with the speech recognition area information that prestores, it is determined whether control described intelligent family
Electric equipment carries out the identification of described voice signal.
The invention has the beneficial effects as follows: when intelligent appliance equipment enters speech recognition state, control timer starts to count
When, collect voice signal when the timing of intervalometer is less than or equal to Preset Time, then it is fixed to carry out according to this voice signal
Position, determines the positional information of sound source corresponding to this voice signal, and according to positional information and the speech recognition area information prestoring,
Determine whether the identification controlling intelligent appliance equipment to carry out voice signal, such that it is able to exclusive PCR voice signal, effectively carry
The discrimination of high speech recognition system, the interactive experience of lifting user.
On the basis of technique scheme, the present invention can also do following improvement.
Further, step 31, correspond to when the corresponding position of described positional information is in described speech recognition area information
Speech recognition region in when, determine and control described intelligent appliance equipment to carry out the identification of described voice signal;
Or, step 32, when the corresponding position of described positional information is in the corresponding language of described speech recognition area information
When outside sound identification region, determine collection voice signal next time.
Beneficial effect using above-mentioned further scheme is: when sound source position is in speech recognition region, controls intelligence
Energy home appliance carries out the identification of voice signal, when outside sound source position being in speech recognition region, collection voice next time
Signal, can effectively exclude the interference of dummy order sound source.
Further, after step 31, described audio recognition method also includes:
Step 4, when described voice signal recognition failures, collection voice signal next time;Or,
Step 5, when described voice signal identifies successfully, it is fixed that the sound source corresponding to described voice signal is carried out again
Position, determines the current location information of described sound source;
Step 6, described positional information is updated according to described current location information, right to determine new speech recognition region institute
The speech recognition area information answered;
Step 7, the described intervalometer of replacement, make described intervalometer restart timing, and gather voice signal next time.
Further, described audio recognition method also includes:
Step 8, when the timing of described intervalometer is more than Preset Time, control described intelligent appliance equipment to enter dormancy shape
State.
Beneficial effect using above-mentioned further scheme is: when time t exceeds Preset Time, controls intelligent appliance equipment
Entering resting state, speech recognition system being avoided to be constantly in speech recognition state, thus reducing the situation of misrecognition.
Further, before step 1, described audio recognition method also includes:
Step 9, collection the first voice signal, described first voice signal is used for making described intelligent appliance equipment stop from described
Dormancy state enters described speech recognition state;
Step 10, positioned according to described first voice signal, and determined sound source corresponding to described first voice signal
Primary importance information;
Step 11, according to described primary importance information, determine described speech recognition area information.
Beneficial effect using above-mentioned further scheme is: makes intelligent appliance equipment enter language from resting state by basis
Sound identifies that the first voice signal of state is positioned, to determine speech recognition area information, to be arrived according to subsequent acquisition
The positional information of the sound source that voice signal is positioned and this speech recognition area information, it is determined whether control intelligent appliance equipment
Carry out the identification of voice signal, such that it is able to exclusive PCR voice signal, effectively improve the discrimination of speech recognition system, carry
Rise the interactive experience of user.
Another kind of technical scheme that the present invention solves above-mentioned technical problem is as follows: a kind of speech recognition based on sound localization
System, comprising:
Voice collector, for when intelligent appliance equipment enters speech recognition state, gathering voice signal;
Controller, for when intelligent appliance equipment enters speech recognition state, control timer starts timing;
Processor, for collecting described voice letter when the timing of described intervalometer is less than or equal to Preset Time
Number, then positioned according to the described voice signal of described voice collector collection, determined sound source corresponding to described voice signal
Positional information, and according to described positional information and the speech recognition area information that prestores, it is determined whether control described intelligent family
Electric equipment carries out the identification of described voice signal.
The invention has the beneficial effects as follows: when intelligent appliance equipment enters speech recognition state, controller control timer
Start timing, voice collector collects voice signal, processor when the timing of intervalometer is less than or equal to Preset Time
Positioned according to this voice signal, determined the positional information of sound source corresponding to voice signal, and according to positional information and prestored
Speech recognition area information, it is determined whether control intelligent appliance equipment to carry out the identification of voice signal, dry such that it is able to exclude
Disturb voice signal, effectively improve the discrimination of speech recognition system, the interactive experience of lifting user.
On the basis of technique scheme, the present invention can also do following improvement.
Further, described processor is specifically for being in described speech recognition area when the corresponding position of described positional information
When in domain information corresponding speech recognition region, determine the identification controlling described intelligent appliance equipment to carry out described voice signal;
Or, when outside the corresponding position of described positional information being in the corresponding speech recognition region of described speech recognition area information,
Determine collection voice signal next time.
Further, described voice collector is additionally operable to when described voice signal recognition failures, collection language next time
Message number;
Or, described processor is additionally operable to: when described voice signal identifies successfully, to corresponding to described voice signal
Sound source is positioned again, determines the current location information of described sound source, and updates institute's rheme according to described current location information
Confidence ceases, to determine the speech recognition area information corresponding to new speech recognition region;Described controller is additionally operable to reset institute
State intervalometer, make described intervalometer restart timing;Described voice collector is additionally operable to gather voice signal next time.
Further, described controller is additionally operable to, when the timing of described intervalometer is more than Preset Time, control described intelligence
Resting state can be entered by home appliance.
Further, described voice collector is additionally operable to: collection the first voice signal, and described first voice signal is used for making
Described intelligent appliance equipment enters described speech recognition state from described resting state;
Described processor is additionally operable to: is positioned according to described first voice signal, and determines described first voice signal
The primary importance information of corresponding sound source, and according to described primary importance information, determine described speech recognition area information.
Further, described voice collector adopts microphone array.
Another kind of technical scheme that the present invention solves above-mentioned technical problem is as follows: a kind of intelligent appliance equipment, comprising: as above
State the speech recognition system based on sound localization described in any one embodiment.
Further, described intelligent appliance equipment includes: intelligent refrigerator, intelligent air condition and Intelligent air purifier.
The advantage of the aspect that the present invention adds will be set forth in part in the description, and partly will become from the following description
Obtain substantially, or recognized by present invention practice.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes it is clear that described embodiment is a part of embodiment of the present invention, rather than whole embodiments.Based on this
Embodiment in bright, the every other reality that those of ordinary skill in the art are obtained on the premise of not making creative work
Apply example, all should belong to the scope of protection of the invention.
In home environment, it is frequently present of many people session operational scenarios.In such circumstances, if speech recognition system works
State, needs to process various voices.Due to there is phonetic entry always, speech recognition system can hardly enter resting state, and one
Directly it is in speech recognition state, carry out the identification of voice signal, and the voice now identifying is incoherent interference voice mostly,
The situation of misrecognition easily occurs.
And in the embodiment of the present invention it is assumed that position of sound source will not occur quickly to change, by waking up home appliance
Voice signal carries out sound localization, then is realized and the interacting of home appliance by voice signal afterwards, and updates the position of sound source
Confidence ceases, and only the voice signal of this position is identified, can exclude the interference in dummy order source.
A kind of executive agent of audio recognition method 100 based on sound localization as shown in Figure 1 can be speech recognition
System or intelligent appliance equipment, this audio recognition method 100 includes:
110, when intelligent appliance equipment enters speech recognition state, gather voice signal, and control timer starts to count
When.
120, collect voice signal when the timing of intervalometer is less than or equal to Preset Time, then believed according to this voice
Number positioned, determined the positional information of sound source corresponding to this voice signal.
130, according to positional information and the speech recognition area information prestoring, it is determined whether control intelligent appliance equipment to carry out
The identification of voice signal.
Specifically, in this embodiment, as shown in figure 1, audio recognition method 100 can also include: step 125, when fixed
When device timing be more than Preset Time when, then can control intelligent appliance equipment enter resting state.
In above-described embodiment, when intelligent appliance equipment enters speech recognition state, control timer starts timing, when fixed
When device timing be less than or equal to Preset Time when collect voice signal, then positioned according to this voice signal, determine
The positional information of sound source corresponding to this voice signal, and according to positional information and the speech recognition area information prestoring, determination is
The no identification controlling intelligent appliance equipment to carry out voice signal, such that it is able to exclusive PCR voice signal, effectively improves voice
The discrimination of identifying system, the interactive experience of lifting user.
Meanwhile, when the timing of intervalometer is more than Preset Time, control intelligent appliance equipment to enter resting state, can keep away
Exempt from speech recognition system and be constantly in speech recognition state, thus reducing the situation of misrecognition.
Alternatively, as one embodiment of the present of invention, step 130 as shown in Figure 2 may include that
Step 131, when the corresponding position of positional information is in speech recognition area information corresponding speech recognition region
When, determine the identification controlling intelligent appliance equipment to carry out voice signal.
Or, step 132, when the corresponding position of positional information is in speech recognition area information corresponding speech recognition area
When overseas, determine collection voice signal next time.
In above-described embodiment, when sound source position is in speech recognition region, intelligent appliance equipment is controlled to carry out voice
The identification of signal, when outside sound source position being in speech recognition region, collection voice signal next time, can effectively exclude
The interference of dummy order sound source.
Alternatively, as an alternative embodiment of the invention, as shown in Fig. 2 after step 131, audio recognition method
100 can also include:
140, voice signal identifies whether successfully, if it is not, then execution step 132;If so, then execution step 141.
141, the sound source corresponding to voice signal is positioned again, determines the current location information of sound source.
142, according to current location information more new location information, to determine the voice knowledge corresponding to new speech recognition region
Other area information.
143, reset intervalometer, make intervalometer restart timing, and execution step 132.
Specifically, in this embodiment, when voice signal identifies successfully in addition it is also necessary to further to corresponding to voice signal
Sound source positioned again, determine the current location information of sound source, and according to current location information more new location information, with true
The fixed new speech recognition area information corresponding to speech recognition region.Meanwhile, need reset intervalometer so that intervalometer again
Start timing, and gather voice signal next time.
Alternatively, as an alternative embodiment of the invention, as shown in figure 3, before step 110, audio recognition method
100 can also include:
150, gather the first voice signal, the first voice signal is used for making intelligent appliance equipment enter voice from resting state
Identification state.
160, positioned according to the first voice signal, and determined the primary importance letter of sound source corresponding to the first voice signal
Breath.
170, according to primary importance information, determine speech recognition area information.
In above-described embodiment, intelligent appliance equipment is made to enter the first language of speech recognition state from resting state by basis
Message number is positioned, determining speech recognition area information so that according to subsequent acquisition to voice signal positioned
The positional information of sound source and this speech recognition area information, it is determined whether control intelligent appliance equipment to carry out the knowledge of voice signal
Not, the discrimination of speech recognition system, the interactive body of lifting user such that it is able to exclusive PCR voice signal, are effectively improved
Test.
With reference to Fig. 4, audio recognition method provided in an embodiment of the present invention is described in detail.As shown in Figure 4
A kind of audio recognition method 200 include:
210, intelligent appliance equipment is in can not be with the resting state of user mutual, if can collect for inciting somebody to action intelligence
Home appliance wakes up the first voice signal making it into speech recognition state, if having collected the first voice signal, executes
Step 220;Otherwise, intelligent appliance equipment keeps resting state.
220, positioned according to the first voice signal, and determined the primary importance of this sound source corresponding to the first voice signal
Information.Meanwhile, intervalometer can be started and start timing.
230, according to primary importance information, determine speech recognition area information.
240, if collected voice signal in Preset Time, if so, then execution step 250;Otherwise, intelligent appliance
Equipment enters resting state.
For example: in this embodiment, Preset Time can collect for setting intelligent appliance that is to say, that working as 10s
When standby wake-up makes it into the first voice signal of speech recognition state, intervalometer, can be collected if in 10s with timing
Voice signal, then execution step 250;If it exceeds 10s, do not collect voice signal, then intelligent appliance equipment enters dormancy shape
State.250, positioned according to voice signal, determined the positional information of sound source corresponding to this voice signal.
260, according to positional information and speech recognition area information, determine that the corresponding position of this positional information is in voice and knows
In the corresponding speech recognition region of other area information, if so, then execution step 270, otherwise, execution step 240.
270, control intelligent appliance equipment to carry out the identification of this voice signal.
280, determine and this voice signal identified whether successfully, if so, then execute 290;Otherwise, execution step 240.
290, the sound source corresponding to voice signal is positioned again, determines the current location information of sound source.
295, according to current location information more new location information, to determine the voice knowledge corresponding to new speech recognition region
Other area information, meanwhile, resets intervalometer, makes intervalometer restart timing, and execution step 240.
For example: in this embodiment, Preset Time can be 10s, and collected voice signal in 6s, and to language
Message number is positioned and is identified, and speech recognition success, now needs replacement intervalometer to make its reclocking, i.e. at next
Whether voice signal can be collected in individual 10s, if can collect, continue and voice signal is positioned and identifies;
If more than 10s regardless of whether collecting, intelligent appliance equipment enters resting state.
In above-described embodiment, the first voice signal crossing wake-up home appliance carries out sound localization, then by language afterwards
Message number realization is interacted with home appliance, and updates the positional information of sound source, only the voice signal of this position is identified,
The interference in dummy order source can effectively be excluded.
It should be understood that in this embodiment, Preset Time can be configured according to actual situation, the embodiment of the present invention pair
This does not do any restriction, and Preset Time is 10s is only to illustrate in order to the technical scheme of the embodiment of the present invention to be described, not to this
Inventive embodiments constitute any restriction.
It should also be understood that in the various embodiments described above of the present invention, the size of the sequence number of above-mentioned each process is not meant to hold
The priority of row order, the execution sequence of each process should be determined with its function and internal logic, and should not be to the embodiment of the present invention
Implementation process constitutes any restriction.
Above in association with Fig. 1 to Fig. 4, a kind of entered based on the audio recognition method of sound localization to provided in an embodiment of the present invention
Go detailed description, with reference to Fig. 5 to a kind of speech recognition system based on sound localization provided in an embodiment of the present invention
It is described in detail.
A kind of speech recognition system 500 based on sound localization as shown in Figure 5, comprising: voice collector 510, control
Device 520 and processor 530.Wherein,
Voice collector 510 is used for, when intelligent appliance equipment enters speech recognition state, gathering voice signal.Controller
520 are used for when intelligent appliance equipment enters speech recognition state, and control timer starts timing.Processor 530 is used for when fixed
When device timing be less than or equal to Preset Time when collect voice signal, then according to voice collector 510 collection voice
Signal is positioned, and determines the positional information of sound source corresponding to voice signal, and according to positional information and the speech recognition prestoring
Area information, it is determined whether control intelligent appliance equipment to carry out the identification of voice signal.
Specifically, in this embodiment, controller 520 is additionally operable to, when the timing of intervalometer is more than Preset Time, control
Intelligent appliance equipment enters resting state.
In above-described embodiment, when intelligent appliance equipment enters speech recognition state, controller control timer starts to count
When, when the timing of intervalometer is less than or equal to Preset Time, voice collector collects voice signal, and processor is according to this
Voice signal is positioned, and determines the positional information of sound source corresponding to voice signal, and according to positional information and the voice prestoring
Identification region information, it is determined whether control intelligent appliance equipment to carry out the identification of voice signal, such that it is able to exclusive PCR voice
Signal, effectively improves the discrimination of speech recognition system, the interactive experience of lifting user.
Meanwhile, when the timing of intervalometer is more than Preset Time, controller controls intelligent appliance equipment to enter resting state,
Speech recognition system can be avoided to be constantly in speech recognition state, thus reducing the situation of misrecognition.
It should be understood that in this embodiment, speech recognition system 500 may correspond to speech recognition according to embodiments of the present invention
Above and other operation of the modules in the executive agent of method, and speech recognition system 500 and/or function are respectively
Realize the corresponding flow process of each method in Fig. 1 to Fig. 4, for sake of simplicity, will not be described here.
It should also be understood that in this embodiment, voice collector 510 can adopt microphone array, is used for meeting plane position
The location requirement put.When intelligent appliance equipment is intelligent refrigerator, voice collector 510 can be horizontal using 2mic or 4mic
Array;When the small household appliances that intelligent appliance equipment is air purifier etc, voice collector 510 can adopt the round battle array of 5mic
Row.
It should be noted that in this embodiment, processor 530 is in the voice signal being gathered according to voice collector 510
Positioned, before determining the positional information of sound source corresponding to voice signal, voice signal can be carried out at the audio frequency of routine
Reason, for example: a/d conversion, noise reduction etc., but the embodiment of the present invention does not limit to this.
Alternatively, as one embodiment of the present of invention, processor 530 is specifically at the corresponding position of positional information
When in speech recognition area information corresponding speech recognition region, determine the knowledge controlling intelligent appliance equipment to carry out voice signal
Not.Or, when outside the corresponding position of positional information being in speech recognition area information corresponding speech recognition region, determination is adopted
Collection voice signal next time.
Alternatively, as one embodiment of the present of invention, voice collector 510 is additionally operable to when voice signal recognition failures
When, collection voice signal next time.
Or, processor 530 is additionally operable to, when voice signal identifies successfully, the sound source corresponding to voice signal be entered again
Row positioning, determines the current location information of sound source, and according to current location information more new location information, to determine new voice knowledge
Speech recognition area information corresponding to other region.Controller 520 is additionally operable to reset intervalometer, makes intervalometer restart to count
When.Voice collector is additionally operable to gather voice signal next time.
Alternatively, as an alternative embodiment of the invention, voice collector 510 is additionally operable to gather the first voice signal,
First voice signal is used for making intelligent appliance equipment enter speech recognition state from resting state.Processor 530 is additionally operable to basis
First voice signal is positioned, and determines the primary importance information of sound source corresponding to the first voice signal, and according to first
Confidence ceases, and determines speech recognition area information.
The embodiment of the present invention also provides a kind of intelligent appliance equipment, and this intelligent appliance equipment is included as above-mentioned any embodiment
In the speech recognition system 500 based on sound localization.Specifically, intelligent appliance equipment can include intelligent refrigerator, Intelligent air
Mediation Intelligent air purifier, or it is also possible to include other intelligent appliances, the embodiment of the present invention does not do any limit to this
Fixed.
In above-described embodiment, when intelligent appliance equipment enters speech recognition state, according to speech recognition system collection
Voice signal is positioned, and determines the positional information of sound source corresponding to voice signal, and according to positional information and the voice prestoring
Identification region information, it is determined whether control intelligent appliance equipment to carry out the identification of voice signal, such that it is able to exclusive PCR voice
Signal, effectively improves the discrimination of speech recognition system, the interactive experience of lifting user.
Those skilled in the art can be understood that, for convenience of description and succinctly, foregoing description be
The specific work process of system, device and unit, may be referred to the corresponding process in preceding method embodiment, will not be described here.
It should be understood that disclosed system, apparatus and method in several embodiments provided herein, permissible
Realize by another way.For example, device embodiment described above is only schematically, for example, the division of unit,
It is only a kind of division of logic function, actual can have other dividing mode when realizing, and for example multiple units or assembly are permissible
In conjunction with or be desirably integrated into another system, or some features can be ignored, or does not execute.
The unit illustrating as separating component can be or may not be physically separate, show as unit
Part can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple networks
On unit.The mesh to realize embodiment of the present invention scheme for some or all of unit therein can be selected according to the actual needs
's.
In addition, can be integrated in a processing unit in each functional unit in each embodiment of the present invention it is also possible to
It is that unit is individually physically present or two or more units are integrated in a unit.Above-mentioned integrated
Unit both can be to be realized in the form of hardware, it would however also be possible to employ the form of SFU software functional unit is realized.
If integrated unit realized using in the form of SFU software functional unit and as independent production marketing or use when, can
To be stored in a computer read/write memory medium.Based on such understanding, technical scheme substantially or
Say the part that prior art is contributed, or all or part of this technical scheme can be embodied in the form of software product
Out, this computer software product is stored in a storage medium, including some instructions with so that a computer equipment
(can be personal computer, server, or network equipment etc.) executes all or part of each embodiment method of the present invention
Step.And aforesaid storage medium includes: u disk, portable hard drive, read only memory (rom, read-only memory), random
Access memorizer (ram, random access memory), magnetic disc or CD etc. are various can be with Jie of store program codes
Matter.
More than, the specific embodiment of the only present invention, but protection scope of the present invention is not limited thereto, any it is familiar with
Those skilled in the art the invention discloses technical scope in, various equivalent modifications or replacement can be readily occurred in,
These modifications or replacement all should be included within the scope of the present invention.Therefore, protection scope of the present invention should be wanted with right
The protection domain asked is defined.