Voice wake-up method and device, electronic equipment and storage medium

Abstract

The invention relates to a voice wake-up method and device, electronic equipment and a storage medium, and the voice wake-up method comprises the steps: collecting a wake-up audio, inputting the wake-up audio into a pre-constructed voice model, and generating a recognition result; collecting video information of personnel in a target vehicle in real time, and obtaining personnel visual information according to the video information; according to the recognition result, whether the first person sending the wake-up audio is matched with a second person recorded in the person visual information or not is judged, and if yes, a wake-up signal is triggered to wake up a voice system of the target vehicle. The method provided by the invention can effectively solve the problem of mistaken wake-up outside the vehicle, and improves the wake-up accuracy and the vehicle safety.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a voice wake-up method, device, electronic device and storage medium. Background Art

[0002] With the continuous development of intelligent vehicle systems, vehicle voice systems for human-computer interaction have gradually become a research and development focus. Vehicle voice systems in standby mode usually use voice wake-up to start voice recognition. By comparing the similarity between the wake-up audio and the wake-up word, if the similarity exceeds the threshold, the wake-up is considered successful. The advantage of voice wake-up is that it has low implementation cost and convenient voice model training.

[0003] However, this voice wake-up method often triggers various false wake-ups, and the wake-up accuracy is relatively low, which seriously affects the user experience. In particular, people outside the vehicle can also wake up the in-vehicle voice system at will, which can easily lead to safety problems. Summary of the invention

[0004] In order to solve the above technical problems, the present disclosure provides a voice wake-up method, device, electronic device and storage medium, which can effectively solve the problem of false wake-up from outside the vehicle and improve the wake-up accuracy and vehicle safety.

[0005] In a first aspect, an embodiment of the present disclosure provides a voice wake-up method, including:

[0006] Collecting wake-up audio and inputting the wake-up audio into a pre-built voice model to generate a recognition result;

[0007] Collecting video information of people inside the target vehicle in real time, and obtaining visual information of the people based on the video information;

[0008] According to the recognition result, it is determined whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person. If so, a wake-up signal is triggered to wake up the voice system of the target vehicle.

[0009] Optionally, the real-time acquisition of video information of people inside the target vehicle and obtaining visual information of the people based on the video information includes:

[0010] Accessing the server, collecting video information of the person inside the target vehicle in real time, and obtaining visual information of the person based on the video information, wherein the visual information of the person is generated by a visual camera device detecting the person inside the target vehicle at a preset frequency and stored in the server;

[0011] The personnel visual information includes the position information of the target person inside the target vehicle and the visual information of the target person.

[0012] Optionally, the server deploys a remote procedure call system, and the access server includes:

[0013] Determine whether the server is a local service, and if so, access it through the loopback address of the server and the preset port of the server; or, if not, access it through the Internet Protocol address of the server and the preset port.

[0014] Optionally, judging, based on the recognition result, whether the first person who sends the wake-up audio matches the second person recorded in the person visual information, and if so, triggering a wake-up signal to wake up the voice system of the target vehicle includes:

[0015] Based on the recognition result and the wake-up audio, determine whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and output a matching result;

[0016] If the matching result is that the first person is the second person recorded in the person visual information, a wake-up signal is triggered, wherein the wake-up signal is used to wake up the voice system inside the target vehicle.

[0017] Optionally, the recognition result includes a wake-up confidence and a current wake-up position.

[0018] Optionally, based on the recognition result and the wake-up audio, determining whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and outputting a matching result includes:

[0019] When the wake-up confidence is greater than a preset threshold, determine whether the personnel visual information includes information of a second person at the current wake-up position. If so, input the wake-up audio and the target visual information of the second person in the personnel visual information into the wake-up model, and output the matching result.

[0020] Optionally, the recognition result also includes wake-up time, and the target visual information includes mouth coordinate information.

[0021] Optionally, inputting the wake-up audio and the target visual information of the second person in the person visual information into the wake-up model, and outputting the matching result includes:

[0022] Determine all mouth coordinate information of the second person within the wake-up time in the person visual information, and generate a mouth coordinate array;

[0023] The wake-up audio and the mouth coordinate array are input into the wake-up model to determine whether the mouth changes of the second person match the wake-up audio, and the matching result is output.

[0024] Optionally, the method further includes:

[0025] If the wake-up confidence is less than or equal to the preset threshold, the person visual information does not include any user's information, the person visual information does not include the second person's information, and / or the matching result is that the first person is not the second person, the wake-up signal is not triggered.

[0026] In a second aspect, an embodiment of the present disclosure provides a voice wake-up device, including:

[0027] A collection unit, used to collect wake-up audio, and input the wake-up audio into a pre-built voice model to generate a recognition result;

[0028] An acquisition unit, used to collect video information of people inside the target vehicle in real time, and acquire visual information of the people based on the video information;

[0029] A judgment unit is used to judge whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person according to the recognition result, and if so, trigger a wake-up signal to wake up the voice system of the target vehicle.

[0030] In a third aspect, an embodiment of the present disclosure provides an electronic device, including:

[0031] Memory;

[0032] Processor; and

[0033] Computer programs;

[0034] The computer program is stored in the memory and is configured to be executed by the processor to implement the voice wake-up method as described above.

[0035] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the voice wake-up method as described above are implemented.

[0036] The embodiment of the present disclosure provides a voice wake-up method including: collecting wake-up audio, and inputting the wake-up audio into a pre-built voice model to generate a recognition result; collecting video information of people inside the target vehicle in real time, and obtaining visual information of the people based on the video information; judging whether the first person who issued the wake-up audio matches the second person recorded in the visual information of the people based on the recognition result, and if so, triggering a wake-up signal to wake up the voice system of the target vehicle. The method provided by the present disclosure collects video information of people inside the vehicle in real time, obtains visual information of people inside the vehicle, and judges whether the first person who issued the wake-up audio is the second user recorded in the visual information of the people, that is, through the multimodal information of audio and video, it is judged whether the people inside the vehicle issued the wake-up audio, which effectively solves the problem of false wake-up outside the vehicle, and improves the wake-up accuracy, vehicle safety and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0039] Figure 1 A flow chart of a voice wake-up method provided in an embodiment of the present disclosure;

[0040] Figure 2 A flowchart of another voice wake-up method provided by an embodiment of the present disclosure;

[0041] Figure 3 A schematic diagram of the structure of a voice wake-up device provided in an embodiment of the present disclosure;

[0042] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0043] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0044] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0045] At present, pure voice recognition solutions are mostly used to wake up the vehicle's voice system. This solution uses a pure voice wake-up method to compare the similarity between the wake-up audio and the wake-up word. If the similarity exceeds a certain threshold, it is determined that the wake-up is successful. The advantage of this solution is low cost and convenient voice model training. However, the defect of this solution is that it is difficult to solve the problem of false wake-up outside the cabin, that is, people outside the car wake up the voice assistant in the car. Among them, out-of-cabin wake-up refers to people waking up the voice assistant in the car by talking loudly outside the cabin. For example, a user speaks outside the car and wakes up the voice assistant of a vehicle owned by another user. For the vehicle, it is a false wake-up. This problem is likely to threaten the safety of the vehicle and the user, and it is difficult to solve this problem only by training a voice model.

[0046] In view of the above technical problems, the embodiments of the present disclosure provide a voice wake-up method, which adopts a multimodal approach, integrates vision and voice, solves the problem of false wake-up outside the cabin that is difficult to solve with a pure voice solution, and enhances the accuracy of voice wake-up by integrating visual signals. The specific details can be described in detail through at least one of the following embodiments.

[0047] Figure 1 A flow chart of a voice wake-up method provided in an embodiment of the present disclosure is applied to a vehicle, wherein the vehicle includes a client and a server, and both the client and the server deploy a remote procedure call system (gRPC Remote Procedure Calls), wherein the gRPC system is based on HTTP / 2 protocol model transmission, and uses a data description language (Protocol Buffers) as an interface description language. The voice wake-up method can be applied to the client, specifically including the following: Figure 1 The following steps S101 to S103 are shown:

[0048] S101. Collect wake-up audio, and input the wake-up audio into a pre-built voice model to generate a recognition result.

[0049] It is understandable that the target vehicle is equipped with an audio collection device for collecting ambient audio within a certain range of the target vehicle. After the wake-up audio is collected by the audio collection device, it is detected whether the wake-up audio includes a wake-up word, for example, the wake-up word is "voice assistant" or the preset name of the voice system, etc. If the wake-up audio includes the wake-up word, the pre-built voice model is awakened, the wake-up audio is used as the input of the voice model, and the recognition result of the wake-up audio is output. It is understandable that the specific structure of the voice model is not limited and can be selected according to user needs.

[0050] S102: Collect video information of people inside the target vehicle in real time, and obtain visual information of the people based on the video information.

[0051] Optionally, in the above S102, the video information of the person inside the target vehicle is collected in real time, and the visual information of the person is obtained according to the video information, which can be specifically implemented by the following steps:

[0052] Access the server, collect video information of people inside the target vehicle in real time, and obtain visual information of the people based on the video information, wherein the visual information of the people is generated by a visual camera device detecting people inside the target vehicle at a preset frequency and stored in the server; wherein the visual information of the people includes the position information of the target person inside the target vehicle and the visual information of the target person.

[0053] It can be understood that, based on the above S101, when it is detected that the voice model is awakened, the grpc server is accessed through the grpc client to obtain the video information of the people inside the target vehicle, and the visual information of the people is obtained based on the video information. The target vehicle is equipped with a visual camera device, which can also be understood as a visual server. The visual camera device is used to detect the visual information of people at all positions inside the target vehicle at a preset frequency, and store the collected visual information of the people to the server for the client to obtain. Storing it in the server can reduce the storage space occupied, and the required video information can be retrieved from the server only when needed. Preferably, the visual camera The imaging device detects visual information of people at a frequency of 5 times per second, wherein the visual information of people includes the position information of the target person inside the target vehicle and the mouth coordinate information of the target person. The position information includes the main driver's seat, the co-driver's seat, the second row on the left and the second row on the right, etc., which can be set according to the internal structure of the target vehicle and is not limited here. The mouth coordinate information can specifically refer to the information of the user's mouth position change, such as the position change information caused by the opening and closing of the mouth, etc., wherein the visual information of people can also include the age of the target person and the number of people at the position. For example, in the scene where an adult holds a child on the left side of the second row, the number of people is 2.

[0054] Wherein, both the server and the client deploy a remote procedure call system.

[0055] Optionally, the above access to the server can be implemented through the following steps:

[0056] Determine whether the server is a local service, and if so, access it through the loopback address of the server and the preset port of the server; or, if not, access it through the Internet Protocol address of the server and the preset port.

[0057] It is understandable that each detection result of the visual camera device will be provided to the grpc server for storage and wait for the client to call. Before calling, the client will determine whether the server is a local service, that is, whether the client and the server are located in the same system. It is understandable that the vehicle will deploy multiple systems, such as the front system and the rear system. In one possible situation, the client is deployed on the front system, and the server is also deployed on the front system. The client and the server are both located in the same system of the front system, which means that the server is a local service. In another possible situation, the client is deployed on the front system, and the server is deployed on the rear system. The client and the server are not located in the same system, which means that the server is not a local service. After completing the judgment, if the server is a local service, the client can access the server through the loopback address and the preset interface of the server. If the server is not a local service, that is, it is located on another system, the client can access it through the Internet Protocol Address (IP address) and the preset port where the server is located.

[0058] S103: Determine, based on the recognition result, whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person; if so, trigger a wake-up signal to wake up the voice system of the target vehicle.

[0059] It can be understood that based on the above S101 and S102, according to the recognition results, a logical judgment is made whether the first person who sends the wake-up audio matches the second person recorded in the person visual information. If the first person and the second person match, that is, the information of the first person is recorded in the person visual information, the wake-up signal is triggered, wherein the wake-up signal is used to wake up the voice system of the target vehicle, and the voice system can be a voice assistant / in-vehicle voice intelligent system inside the target vehicle.

[0060] Optionally, in the above S103, based on the recognition result, it is determined whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person, and if so, a wake-up signal is triggered to wake up the voice system of the target vehicle, which can be specifically implemented by the following steps:

[0061] Based on the recognition result and the wake-up audio, determine whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and output a matching result;

[0062] If the matching result is that the first person is the second person recorded in the person visual information, a wake-up signal is triggered, wherein the wake-up signal is used to wake up the voice system inside the target vehicle.

[0063] It is understandable that the recognition result and wake-up audio output by the voice model are used as the input of the pre-built wake-up model, wherein the wake-up model is used to determine whether the first person matches the person recorded in the person visual information, and output the matching result. Specifically, the internal structure of the wake-up model is not limited, and it is sufficient to realize the logical judgment function described in this embodiment. Among them, the matching results include two matching results: the first person is the second person recorded in the person visual information and the first person is not the second person. If the second person matching the first person is recorded in the person visual information, the wake-up signal is triggered to wake up the voice assistant. It is understandable that the person visual information records the information of all people in the target vehicle. In one possible situation, the target vehicle does not include any person, that is, there is no one in the car. In another possible situation, the target vehicle includes at least one person. The wake-up model is used to determine a second person matching the first person among the at least one person recorded in the person visual information.

[0064] The recognition result includes the wake-up confidence, the wake-up time and the current wake-up position.

[0065] Optionally, based on the recognition result and the wake-up audio, judging whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and outputting a matching result, can be specifically implemented through the following steps:

[0066] When the wake-up confidence is greater than a preset threshold, determine whether the personnel visual information includes information of a second person at the current wake-up position. If so, input the wake-up audio and the target visual information of the second person in the personnel visual information into the wake-up model, and output the matching result.

[0067] It can be understood that the recognition results output by the voice model include wake-up confidence, wake-up time and current wake-up position. The wake-up time refers to the duration of the collected wake-up audio, for example 5s. The current wake-up position is the position of the first person who sends the wake-up audio inside the target vehicle predicted by the voice model, which can be represented by the seat number. For example, in one possible scenario, the first person sends the wake-up audio outside the target vehicle near the main driver's seat. In this case, the current wake-up position recognized by the voice pattern is the main driver's seat inside the target vehicle, but the first person sends the wake-up audio outside the vehicle. In this case, false wake-up is very likely to occur based only on the voice model. Therefore, first determine whether the wake-up confidence is greater than the preset threshold. If the wake-up confidence is greater than the preset threshold, determine whether the personnel visual information includes information about at least one person, that is, determine whether there is a person inside the vehicle. If the personnel visual information includes information about at least one person, further determine whether there is information about a second person located at the current wake-up position in the personnel visual information, that is, determine whether the first person located at the current wake-up position predicted by the voice model is actually inside the vehicle. Based on the above example, the current wake-up position recognized by the voice model is the main driving seat, and determine whether the personnel visual information includes information about the second person located at the main driving seat. If the personnel visual information includes information about the second person located at the main driving seat, that is, there is a second person who matches the position of the first person, that is, there is indeed a person in the main driving seat, then the wake-up audio and the mouth coordinate information of the second person in the personnel visual information are input into the wake-up model to determine whether the second person actually issued the wake-up audio, and output the matching result.

[0068] Wherein, the recognition result also includes the wake-up time, and the target visual information includes the mouth coordinate information.

[0069] Optionally, the step of inputting the wake-up audio and the target visual information of the second person in the visual information of the person into the wake-up model and outputting the matching result may be specifically implemented through the following steps:

[0070] Determine all mouth coordinate information of the second person within the wake-up time in the visual information of the person, and generate a mouth coordinate array; input the wake-up audio and the mouth coordinate array into the wake-up model to determine whether the mouth changes of the second person match the wake-up audio, and output the matching result.

[0071] It is understandable that all the mouth coordinate information of the second person during the wake-up time is determined from a large amount of visual information of the person obtained in real time, and a mouth coordinate array is generated, that is, all the mouth coordinate information of the second person at the current wake-up position during the wake-up time is obtained, so that the wake-up model can analyze the mouth changes of the occupant. The second person can be understood as the occupant, and the mouth coordinate array can also be understood as the person's mouth position coordinate array. Subsequently, the wake-up model combines the wake-up audio and the mouth coordinate array of the second person sent in during the wake-up time to determine whether the mouth changes of the second person match the wake-up audio, that is, whether the second person actually sends the wake-up audio, that is, comprehensively judge whether the occupant has given visual wake-up information. Visual wake-up information refers to determining that the user actually sends the wake-up audio based on the user's mouth change information, so as to give the voice assistant a correct wake-up signal. If the wake-up model determines that the second person in the car has indeed given the wake-up audio, it means that the first person and the second person are the same person, and it is a real wake-up signal given by the person in the car, rather than a false wake-up command given by the person outside the car, which effectively improves the wake-up accuracy.

[0072] Optionally, the method further includes:

[0073] If the wake-up confidence is less than or equal to the preset threshold, the person visual information does not include any user's information, the person visual information does not include the second person's information, and / or the matching result is that the first person is not the second person, the wake-up signal is not triggered.

[0074] It is understandable that the wake-up signal is triggered only when the match is successful. The situations in which the wake-up signal is not triggered include at least one of the following: the wake-up confidence is less than or equal to the preset threshold, the personnel visual information does not include any user's information (that is, there is no one in the car), the personnel visual information does not include the information of the second person (that is, there is no one in the vehicle at the current wake-up position), and the matching result is that the first person is not the second person (the second person at the current wake-up position in the vehicle did not emit any audio or did not emit the wake-up audio within the period of receiving the wake-up audio). Based on the above examples, the scenario in which the personnel visual information does not include the information of the second person is: the current wake-up position identified by the voice pattern is the main driver's seat, and the personnel visual information does not include the information of the second person located in the main driver's seat, that is, there is no person in the main driver's seat; the scenario in which the matching result is that the first person is not the second person is: the personnel visual information includes the information of the second person located in the main driver's seat, that is, there is someone in the main driver's seat, recorded as the second person / occupier, but the second person's mouth did not change during the wake-up period (that is, no audio was emitted) or the mouth changes did not match the mouth changes that emitted the wake-up audio (that is, no wake-up audio was emitted), indicating that the person in the car was not actually awakened, and may have been mistakenly awakened by someone outside the car. In this case, the wake-up signal is not triggered.

[0075] The voice wake-up method provided in the embodiment of the present disclosure adopts an audio and video multi-modal approach to solve the problem of false wake-up outside the cabin, integrates visual algorithms and voice algorithms, solves the problem that it is difficult for the voice algorithm to handle false wake-up outside the cabin alone, ensures the safety of the vehicle, and provides user experience.

[0076] Based on the above embodiments, Figure 2 A flowchart of another voice wake-up method provided by an embodiment of the present disclosure, specifically including the following steps: Figure 2 The following process is shown:

[0077] Visual personnel occupancy detection is performed through a visual camera device, and the obtained personnel occupancy information is stored in the grpc server; the wake-up audio is collected through an audio collection device, the wake-up audio includes a wake-up word to wake up the voice model, and the wake-up audio is recognized by the voice model, and the recognition result is output, the recognition result includes the current wake-up position, wake-up confidence and wake-up time, and the grpc client obtains the personnel occupancy information by accessing the grpc server, and the personnel occupancy information includes the location information of the occupant and the mouth coordinate information of the occupant; through the wake-up decision, it is judged whether the wake-up confidence exceeds the preset threshold and whether there is a person at the current wake-up position. If the wake-up confidence does not exceed the preset threshold and / or there is no person at the current wake-up position, the wake-up fails and the wake-up signal is not triggered. If the wake-up confidence exceeds the preset threshold and there is a person at the current wake-up position, the pre-built wake-up model is continued to comprehensively judge whether the voice wake-up and visual wake-up reach the threshold according to the wake-up audio and the mouth coordinate array of the occupant. If both the visual wake-up and the voice wake-up reach the threshold, the wake-up is successful and the wake-up signal is triggered. If either the visual wake-up or the voice wake-up does not reach the threshold, the wake-up fails and the wake-up signal is not triggered.

[0078] Figure 3 The structure diagram of the voice wake-up device provided by the embodiment of the present disclosure is as follows. The voice wake-up device provided by the embodiment of the present disclosure can execute the processing flow provided by the above-mentioned voice wake-up method embodiment and is applied to vehicles, such as Figure 3 As shown, the voice wake-up device 300 includes a collection unit 301, an acquisition unit 302 and a judgment unit 303, wherein:

[0079] The collection unit 301 is used to collect the wake-up audio and input the wake-up audio into a pre-built speech model to generate a recognition result;

[0080] The acquisition unit 302 is used to collect video information of people inside the target vehicle in real time, and obtain visual information of the people based on the video information;

[0081] The judgment unit 303 is used to judge whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person according to the recognition result, and if so, trigger a wake-up signal to wake up the voice system of the target vehicle.

[0082] Optionally, the acquiring unit 302 is used to:

[0083] Accessing the server, collecting video information of the person inside the target vehicle in real time, and obtaining visual information of the person based on the video information, wherein the visual information of the person is generated by a visual camera device detecting the person inside the target vehicle at a preset frequency and stored in the server;

[0084] The personnel visual information includes the position information of the target person inside the target vehicle and the visual information of the target person.

[0085] Wherein, the server deploys a remote procedure call system.

[0086] Optionally, the acquiring unit 302 is used to:

[0087] Determine whether the server is a local service, and if so, access it through the loopback address of the server and the preset port of the server; or, if not, access it through the Internet Protocol address of the server and the preset port.

[0088] Optionally, the judging unit 303 is used to:

[0089] Based on the recognition result and the wake-up audio, determine whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and output a matching result;

[0090] If the matching result is that the first person is the second person recorded in the person visual information, a wake-up signal is triggered, wherein the wake-up signal is used to wake up the voice system inside the target vehicle.

[0091] The recognition result includes the wake-up confidence and the current wake-up position.

[0092] Optionally, the judging unit 303 is used to:

[0093] When the wake-up confidence is greater than a preset threshold, determine whether the personnel visual information includes information of a second person at the current wake-up position. If so, input the wake-up audio and the target visual information of the second person in the personnel visual information into the wake-up model, and output the matching result.

[0094] Wherein, the recognition result also includes the wake-up time, and the target visual information includes the mouth coordinate information.

[0095] Optionally, the judging unit 303 is used to:

[0096] Determine all mouth coordinate information of the second person within the wake-up time in the person visual information, and generate a mouth coordinate array;

[0097] The wake-up audio and the mouth coordinate array are input into the wake-up model to determine whether the mouth changes of the second person match the wake-up audio, and the matching result is output.

[0098] Optionally, the judging unit 303 is further configured to:

[0099] If the wake-up confidence is less than or equal to the preset threshold, the person visual information does not include any user's information, the person visual information does not include the second person's information, and / or the matching result is that the first person is not the second person, the wake-up signal is not triggered.

[0100] Figure 3 The voice wake-up device of the illustrated embodiment can be used to execute the technical solution of the above-mentioned method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.

[0101] Figure 4 This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. Figure 4 , which shows a schematic diagram of the structure of an electronic device 400 suitable for implementing the embodiment of the present disclosure. The electronic device 400 in the embodiment of the present disclosure may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle terminals (such as vehicle navigation terminals), wearable electronic devices, etc., and fixed terminals such as digital TVs, desktop computers, smart home devices, etc. Figure 4 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0102] like Figure 4As shown, the electronic device 400 may include a processing device (e.g., a central processing unit, a graphics processor, etc.) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage device 408 to a random access memory (RAM) 403 to implement the voice wake-up method of the embodiment described in the present disclosure. In RAM 403, various programs and data required for the operation of the electronic device 400 are also stored. The processing device 401, ROM 402, and RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0103] Typically, the following devices may be connected to the I / O interface 405: an input device 406 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 407 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 408 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 409. The communication device 409 may allow the electronic device 400 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 400 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0104] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program contains a program code for executing the method shown in the flowchart, thereby implementing the voice wake-up method as described above. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 409, or installed from the storage device 408, or installed from the ROM 402. When the computer program is executed by the processing device 401, the above-mentioned functions defined in the method of the embodiment of the present disclosure are executed.

[0105] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0106] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0107] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0108] Optionally, when the above one or more programs are executed by the electronic device, the electronic device may also execute other steps described in the above embodiments.

[0109] Computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0110] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0111] The units involved in the embodiments described in the present disclosure may be implemented by software or hardware, wherein the name of a unit does not, in some cases, limit the unit itself.

[0112] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0113] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0114] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or gateway that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also includes elements inherent to such process, method, article or gateway. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or gateway that includes the elements.

[0115] The above description is only a specific embodiment of the present disclosure, so that those skilled in the art can understand or implement the present disclosure. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to the embodiments described herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A voice wake-up method, characterized in that: include: Collecting wake-up audio and inputting the wake-up audio into a pre-built voice model to generate a recognition result; Collecting video information of people inside the target vehicle in real time, and obtaining visual information of the people based on the video information; According to the recognition result, it is determined whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person. If so, a wake-up signal is triggered to wake up the voice system of the target vehicle.

2. The method according to claim 1, characterized in that The method of determining, based on the recognition result, whether the first person who emits the wake-up audio matches the second person recorded in the person visual information, and if so, triggering a wake-up signal to wake up the voice system of the target vehicle includes: Based on the recognition result and the wake-up audio, determine whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and output a matching result; If the matching result is that the first person is the second person recorded in the person visual information, a wake-up signal is triggered, wherein the wake-up signal is used to wake up the voice system inside the target vehicle.

3. The method according to claim 2, characterized in that The recognition result includes a wake-up confidence and a current wake-up position. Based on the recognition result and the wake-up audio, judging whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person through a pre-built wake-up model, and outputting a matching result includes: When the wake-up confidence is greater than a preset threshold, determine whether the personnel visual information includes information of a second person at the current wake-up position. If so, input the wake-up audio and the target visual information of the second person in the personnel visual information into the wake-up model, and output the matching result.

4. The method according to claim 3, characterized in that The recognition result also includes the wake-up time, the target visual information includes the mouth coordinate information, and the inputting of the wake-up audio and the target visual information of the second person in the person visual information into the wake-up model and the outputting of the matching result includes: Determine all mouth coordinate information of the second person within the wake-up time in the person visual information, and generate a mouth coordinate array; The wake-up audio and the mouth coordinate array are input into the wake-up model to determine whether the mouth changes of the second person match the wake-up audio, and the matching result is output.

5. The method according to claim 3, characterized in that: The method further comprises: If the wake-up confidence is less than or equal to the preset threshold, the person visual information does not include any user's information, the person visual information does not include the second person's information, and / or the matching result is that the first person is not the second person, the wake-up signal is not triggered.

6. The method according to claim 1, characterized in that The real-time collection of video information of people inside the target vehicle and obtaining visual information of the people based on the video information include: Accessing the server, collecting video information of the person inside the target vehicle in real time, and obtaining visual information of the person based on the video information, wherein the visual information of the person is generated by a visual camera device detecting the person inside the target vehicle at a preset frequency and stored in the server; The personnel visual information includes the position information of the target person inside the target vehicle and the visual information of the target person.

7. The method according to claim 6, characterized in that The server deploys a remote procedure call system, and the access server includes: Determine whether the server is a local service, and if so, access it through the loopback address of the server and the preset port of the server; or, if not, access it through the Internet Protocol address of the server and the preset port.

8. A voice wake-up device, characterized in that: include: A collection unit, used to collect wake-up audio, and input the wake-up audio into a pre-built voice model to generate a recognition result; An acquisition unit, used to collect video information of people inside the target vehicle in real time, and acquire visual information of the people based on the video information; A judgment unit is used to judge whether the first person who sends the wake-up audio matches the second person recorded in the visual information of the person according to the recognition result, and if so, trigger a wake-up signal to wake up the voice system of the target vehicle.

9. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the voice wake-up method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the voice wake-up method as claimed in any one of claims 1 to 7 are implemented.

Images