CN111897584A

CN111897584A - Wake-up method and device for voice equipment

Info

Publication number: CN111897584A
Application number: CN202010817772.XA
Authority: CN
Inventors: 李路天; 徐俊峰; 周添杰
Original assignee: AI Speech Ltd
Current assignee: AI Speech Ltd
Priority date: 2020-08-14
Filing date: 2020-08-14
Publication date: 2020-11-06
Anticipated expiration: 2040-08-14
Also published as: CN111897584B

Abstract

The invention discloses a wake-up method for voice equipment, which comprises the steps of receiving a first audio when the voice equipment is in a first state, and judging whether the first audio is not lower than a preset first wake-up threshold value or not; if the first audio is not lower than the first awakening threshold, controlling the voice equipment to enter a second state, intercepting a second audio from the first audio, and judging whether the first audio is not lower than a preset second awakening threshold, wherein the second awakening threshold is higher than the first awakening threshold; if the first audio is lower than a preset second awakening threshold value, sending a second audio to a cloud server; the method for waking up the voice equipment solves the problem that the false wake-up rate and the wake-up rate only need to be balanced when only local wake-up is needed because the wake-up model file on the equipment cannot be too large and is limited by the storage and calculation performance of the equipment.

Description

Wake-up method and device for voice equipment

Technical Field

The invention relates to the field of voice awakening, in particular to an awakening method and device for voice equipment.

Background

The voice awakening aims at awakening the intelligent terminal by understanding voice data of a user, so that a first gate leading to artificial intelligence of human beings is opened, and the intelligent terminal is widely applied to various intelligent terminals, such as intelligent wearable equipment, mobile phones, tablet computers, intelligent household appliances and the like.

The inventor finds that when the voice awakening function is realized, the awakening model file on the equipment cannot be too large due to the storage and calculation performance of the equipment, so that the mistaken awakening rate and the awakening rate only need to be balanced when only local awakening is used, the mistaken awakening rate is high when the awakening rate is high, and the awakening rate is low when the mistaken awakening rate is low; and the use of 'on-line awakening' requires that the audio is uploaded to the server in real time, so that the network consumption is increased, and the awakening rate is greatly reduced under the condition of poor network.

Disclosure of Invention

The present invention is directed to a wake-up method and apparatus for a voice device, which solves at least one of the above problems.

In a first aspect, the present invention provides a wake-up method for a voice device, including:

the speech device has a first state, a second state and a third state, the method comprising: when the voice equipment is in the first state, receiving a first audio, and judging whether the first audio is not lower than a preset first awakening threshold value; if the first audio is not lower than the first awakening threshold, controlling the voice equipment to enter a second state, acquiring a second audio from the first audio, wherein the second audio is a callback audio after processing the first audio, and judging whether the first audio is not lower than a preset second awakening threshold, wherein the second awakening threshold is higher than the first awakening threshold; if the first audio is lower than a preset second awakening threshold value, sending the second audio to a cloud server; and receiving a wake-up result returned by the cloud server, and determining whether to wake up the voice equipment based on the wake-up result.

In a second aspect, the present invention provides an apparatus for a speech device, comprising:

the voice frequency receiving module is configured to receive a first audio frequency when the voice device is in the first state, and judge whether the first audio frequency is not lower than a preset first awakening threshold value; the first awakening threshold judging module is configured to judge whether the current audio is not lower than a preset first awakening threshold; the audio processing module is configured to process the current audio to obtain a contrast audio if the current audio is not lower than the first awakening threshold; a second wake-up threshold determination module configured to determine whether the current audio is not lower than a preset second wake-up threshold, where the second wake-up threshold is higher than the first wake-up threshold; the comparison audio output module is configured to output comparison audio to the cloud server if the current audio is lower than a preset second awakening threshold; and the awakening module receives the recognition result returned by the cloud server as an awakening word and outputs an awakening instruction.

In a third aspect, an electronic device is provided, comprising: the apparatus includes at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the steps of the wake-up method for a voice device of any of the embodiments of the present invention.

In a fourth aspect, the present invention also provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, where the computer program includes program instructions, and when the program instructions are executed by a computer, the computer is caused to execute the steps of the wake-up method for a speech device according to any embodiment of the present invention.

According to the embodiment of the invention, the first awakening threshold and the second awakening threshold are set, when the first awakening threshold is reached, local awakening is started to be prepared, then whether the second awakening threshold is reached is judged, and when the second awakening threshold is not reached, the second audio is obtained by intercepting the first audio, and the cloud is used for analyzing the second audio, so that the awakening rate is ensured, meanwhile, the mistaken awakening rate is greatly reduced, and meanwhile, the speed of awakening time is ensured;

the invention sets a first state, a second state and a third state, wherein the first state is in the standby state, the second state is in the standby state when the first audio is intercepted to obtain the second audio and the second audio is sent, and the third state is in the standby state when the device is completely awakened, thereby further reducing the power consumption of the device.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.

Fig. 1 is a flowchart of a wake-up method for a voice device according to an embodiment of the present invention;

fig. 2 is a flowchart of another wake-up method for a voice device according to an embodiment of the present invention;

fig. 3 is a flowchart of a wake-up method for a voice device according to an embodiment of the present invention;

fig. 4 is a flowchart of another wake-up method for a voice device according to an embodiment of the present invention;

fig. 5 is a block diagram of a wake-up method and apparatus for a voice device according to an embodiment of the present invention;

fig. 6 is a block diagram of another wake-up method apparatus for a voice device according to an embodiment of the present invention;

fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1, a flowchart of an embodiment of a wake-up method for a voice device according to the present application is shown, and the wake-up method for a voice device according to the present embodiment can be applied to various electronic devices.

As shown in fig. 1, in step 101, the voice device is in a first state, and whether a received first audio is not lower than a preset first wake-up threshold is received;

in step 102, if the first audio is not lower than the first awakening threshold, controlling the voice device to enter a second state, intercepting a second audio from the first audio, and determining whether the first audio is not lower than a preset second awakening threshold, wherein the second awakening threshold is higher than the first awakening threshold;

in step 103, if the first audio is lower than a preset second awakening threshold, sending the second audio to a cloud server;

in step 104, a wake-up result returned by the cloud server is received, and whether to wake up the voice device is determined based on the wake-up result.

In this embodiment, in step 101, continuously receiving an audio, and determining whether the received audio reaches a first wake-up threshold; in step 102, when the received audio reaches a first awakening threshold, the device enters a second state, a part of the audio is intercepted from the audio reaching the first awakening threshold, the intercepted audio is judged, and whether the intercepted audio reaches a second awakening threshold is judged; in step 103, the intercepted audio does not reach the second awakening threshold value, and the intercepted audio is transmitted to the cloud server; in step 104, when the cloud server returns information, whether to wake up the voice device is determined according to the returned information.

This embodiment divides into three state with speech equipment, then after the standby is the speech information who receives and reaches first threshold value, just get into the state to speech information, if speech information does not reach the second threshold value, intercepting part speech information sends to the high in the clouds and handles, do so, improve the awakening rate and reduced the mistake awakening rate, intercepting part information makes the high in the clouds processing procedure more convenient more energy saving, this embodiment is when guaranteeing the awakening rate, reduce the mistake awakening rate, guarantee to awaken fast and energy saving awakening, furthermore, the second audio frequency is the audio frequency of duration for 1 second, furthermore, the second audio frequency is three routes gsc audio frequencies and one route original audio frequency, because the angle of awakening has not been calculated at this moment, so it is awakening the audio frequency not to know which way audio frequency is, so several ways of audio frequency upload further guarantee the accuracy of awakening simultaneously.

With further reference to fig. 2, a flow chart of an embodiment of the wake-up method for a speech device of the present application is shown, which is mainly a flow chart of steps further defined for the additional flow of fig. 1.

After judging whether the first audio is not lower than a preset second awakening threshold value, the method further comprises the following steps,

in step 202, if the first audio frequency is not lower than a preset second wake-up threshold, a wake-up instruction is output to control the voice device to enter a third state.

In this embodiment, when the first audio is not lower than the preset second wake-up threshold, the device is directly woken up, and the process of cloud processing is removed, so that energy is further saved.

With further reference to fig. 3, a flow of an embodiment of the wake-up method for a voice device of the present application is shown, and the flow chart mainly further defines the flow chart of fig. 1.

In step 301, when the voice device is in a standby state, receiving a first audio, and determining whether the first audio is not lower than eighty percent of a preset first wake-up threshold;

in step 302, if the received first audio is not lower than eighty percent of the preset first wake-up threshold, the voice device enters the first state.

The method comprises the steps of adding a step similar to pre-awakening before the voice equipment processes audio to trigger the voice equipment to enter the first state, wherein the threshold value entering the first state can be allocated according to requirements, the threshold value entering the first state is set to be eighty percent of the first awakening threshold value, awakening success and accuracy can be improved, and the threshold value entering the first state is obtained.

The following description is provided to enable those skilled in the art to better understand the present disclosure by describing some of the problems encountered by the inventors in implementing the present disclosure and by describing one particular embodiment of the finally identified solution.

The inventors discovered the defects of these similar techniques in the process of implementing the present invention:

the wake-up model file on the device cannot be too large due to the limited storage and calculation performance of the device, which results in that only the "local wake-up" is needed to balance the false wake-up rate and the wake-up rate, the false wake-up rate is high when the wake-up rate is high, and the wake-up rate is low when the false wake-up rate is low. If the online awakening is used, the audio needs to be uploaded to the server in real time, the network consumption is increased, and the awakening rate is greatly reduced under the condition of poor network.

The inventors have found in the course of carrying out the invention why the reason is not easily imaginable:

the method is limited by the storage and calculation performance of the equipment, the awakening model file on the equipment cannot be too large, the voice awakening can reach a certain awakening rate, and the common use requirement can be met by keeping a low false awakening rate, namely, the false awakening rate and the awakening rate are balanced.

The technical problem existing in the prior art is solved through the following scheme:

and the local awakening and the server awakening are simultaneously used, so that the effect of complementary advantages can be realized.

The process is described in detail:

refer to FIG. 4

Step 1: the threshold value for waking up is set to be a lower value than the threshold value for waking up, and the wake-up rate is increased. The high threshold is used for comparison after wake-up.

Step 2: after the awakening engine is started, the audio is continuously input, and 4 paths of audio of the last 1 second are stored at the moment

And step 3: after the pre-awakening is detected, 4 paths of cache audio are respectively uploaded,

and 4, step 4: after the awakening is detected, if the awakening threshold value is higher than the set high threshold value, the awakening is successful, and the operation of the uploading server is ignored. If the preset high threshold value is not reached, sending the audio cached between the pre-awakening and the awakening to the server and sending an audio end mark,

and 5: the 4-path audio indicates that the awakening is successful as long as the 1-path recognized word is the awakening word.

The 4-path cloud identification is used, and the scheme is tested to delay by 100 and 300ms compared with the scheme which only uses local wakeup.

In the process of implementing the present application, the inventors have also tried the following schemes:

the 4-path cloud identification is used, and the scheme is tested to delay by 100 and 300ms compared with the scheme which only uses local wakeup. The alternative is to directly perform server wake-up processing on the 4-way audio, detect whether the audio can be woken up, and do not change other logic.

The disadvantages are as follows: at present, the server does not have the function, so the implementation cannot be realized.

With further reference to fig. 5, there is shown a block diagram of a wake-up unit for a voice device according to the present invention.

As shown in fig. 5, the wake-up recognition processing apparatus 500 includes an audio receiving module 510, configured to receive a first audio when the voice device is in the first state, and determine whether the first audio is not lower than a preset first wake-up threshold;

a first wake-up threshold determination module 520 configured to determine whether the current audio frequency is not lower than a preset first wake-up threshold;

an audio processing module 530, configured to control the voice device to enter a second state, and configured to process the current audio to obtain a contrast audio if the current audio is not lower than the first wake-up threshold;

a second wake-up threshold determining module 540, configured to determine whether the current audio is not lower than a preset second wake-up threshold, where the second wake-up threshold is higher than the first wake-up threshold;

a comparison audio output module 540 configured to output a comparison audio to the cloud server if the current audio is lower than a preset second wake-up threshold;

the wake-up module 560 receives the recognition result returned by the cloud server as a wake-up word, and outputs a wake-up instruction, and further includes an activation module configured to issue an activate audio processing instruction if the current audio is not lower than a preset first wake-up threshold.

With further reference to fig. 6, a block diagram of an embodiment of the wake-up apparatus for a speech device of the present application is shown, which is mainly the block diagram further defined in fig. 5.

As shown in fig. 6, the first wake-up threshold determining module 520 is configured to determine whether the current audio frequency is not lower than a preset first wake-up threshold;

the activation module 570 is configured to activate the device to enter the second state if the determination result of the first wake-up threshold determination module 520 is not lower than the preset first wake-up threshold.

It should be understood that the modules depicted in fig. 5 and 6 correspond to various steps in the methods described with reference to fig. 1, 2, 3, and 4. Thus, the operations and features described above for the method and the corresponding technical effects are also applicable to the modules in fig. 5 and 6, and are not described again here.

It should be noted that the relevant functional modules may also be implemented by a hardware processor, for example, the wake-up module may be implemented by a processor, which is not described herein again.

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 7, the electronic device includes: one or more processors 710 and a memory 720, one processor 710 being illustrated in fig. 7. The apparatus for the wake-up method of the voice apparatus may further include: an input device 730 and an output device 740. The processor 710, the memory 720, the input device 730, and the output device 740 may be connected by a bus or other means, such as the bus connection in fig. 7. The memory 720 is a non-volatile computer-readable storage medium as described above. The processor 710 executes various functional applications of the server and data processing by executing nonvolatile software programs, instructions and modules stored in the memory 720, namely, implements the wake-up method for the voice device of the above-described method embodiments. The input device 730 may receive input numeric or character information and generate key signal inputs related to user settings and function control of a wake-up device for a voice apparatus. The output device 740 may include a display device such as a display screen.

The product can execute the method provided by the embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method. For technical details that are not described in detail in this embodiment, reference may be made to the method provided by the embodiment of the present invention.

As an embodiment, the electronic device is applied to a wake-up apparatus for a voice device, and is used for a client, and includes: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor.

And other electronic devices with data interaction functions.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.

The electronic device of the embodiments of the present application exists in various forms, including but not limited to:

(1) a mobile communication device: such devices are characterized by mobile communications capabilities and are primarily targeted at providing voice, data communications. Such terminals include smart phones, multimedia phones, functional phones, and low-end phones, among others.

(2) Ultra mobile personal computer device: the equipment belongs to the category of personal computers, has calculation and processing functions and generally has the characteristic of mobile internet access. Such terminals include: PDA, MID, and UMPC devices, etc.

(3) A portable entertainment device: such devices can display and play multimedia content. The devices comprise audio and video players, handheld game consoles, electronic books, intelligent toys and portable vehicle-mounted navigation devices.

(4) The server is similar to a general computer architecture, but has higher requirements on processing capability, stability, reliability, safety, expandability, manageability and the like because of the need of providing highly reliable services.

(5) And other electronic devices with data interaction functions.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A wake-up method for a voice device, wherein the voice device has a first state, a second state, and a third state, the method comprising:

when the voice equipment is in the first state, receiving a first audio, and judging whether the first audio is not lower than a preset first awakening threshold value;

if the first audio is not lower than the first awakening threshold, controlling the voice equipment to enter a second state, acquiring a second audio from the first audio, wherein the second audio is a callback audio after processing the first audio, and judging whether the first audio is not lower than a preset second awakening threshold, wherein the second awakening threshold is higher than the first awakening threshold;

if the first audio is lower than a preset second awakening threshold value, sending the second audio to a cloud server;

and receiving a wake-up result returned by the cloud server, and determining whether to wake up the voice equipment based on the wake-up result.

2. The method of claim 1, wherein after determining whether the first audio is not below a preset second wake threshold, further comprising:

and if the first audio is not lower than a preset second awakening threshold value, outputting an awakening instruction and controlling the voice equipment to enter a third state.

3. The method of claim 1, wherein when the voice device is in the first state, receiving a first audio, and before determining whether the first audio is not below a preset first wake-up threshold, the method further comprises:

when the voice equipment is in a standby state, receiving a first audio frequency, and judging whether the first audio frequency is not lower than eighty percent of a preset first awakening threshold value;

and if the received first audio is not lower than eighty percent of a preset first awakening threshold value, the voice equipment enters the first state.

4. The method of claim 1, wherein the second audio is audio of duration 1 second.

5. The method of claim 4, wherein the second audio is three gsc pieces of audio and one piece of original audio.

6. A wake-up apparatus for a voice device, comprising:

the audio receiving module is configured to receive a first audio when the voice device is in the first state, and judge whether the first audio is not lower than a preset first awakening threshold;

the first awakening threshold judging module is configured to judge whether the current audio is not lower than a preset first awakening threshold;

the audio processing module controls the voice equipment to enter a second state and is configured to process the current audio to obtain a contrast audio;

a second wake-up threshold determination module configured to determine whether the current audio is not lower than a preset second wake-up threshold, where the second wake-up threshold is higher than the first wake-up threshold;

the comparison audio output module is configured to output comparison audio to the cloud server if the current audio is lower than a preset second awakening threshold;

and the awakening module receives the recognition result returned by the cloud server as an awakening word and outputs an awakening instruction.

7. The apparatus of claim 6, further comprising: and the activation module is configured to send out an audio activation processing instruction if the current audio is not lower than a preset first awakening threshold.

8. The apparatus of claim 6, wherein the audio processing module is further configured to receive audio and to output four ways of contrast audio for the last 1 second continuously.

9. An electronic device, comprising: at least one processor, and a memory communicatively coupled to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the steps of the method of any one of claims 1 to 5.

10. A storage medium having stored thereon a computer program, characterized in that the program, when being executed by a processor, is adapted to carry out the steps of the method of any one of claims 1 to 5.