Response method and device of voice equipment, electronic equipment and storage medium

Abstract

The invention relates to a voice equipment response method and device, electronic equipment and a storage medium, and the method comprises the steps: responding to a plurality of concurrent voice instructions, creating a voice task for each voice instruction, and determining the priority of each voice task; according to the priorities of the voice tasks, a scheduling strategy based on time slices is adopted to schedule the voice tasks in sequence, when the voice tasks are scheduled, the time slices are distributed to the voice tasks, and the voice tasks are executed in the time slices, so that the voice device responds to the voice instruction. Therefore, under the condition that a plurality of voice instructions are concurrent, the edge nodes can orderly and efficiently process the voice instructions, so that stable and reliable user experience is provided, meanwhile, by reasonably distributing time slices and preferentially processing high-priority tasks, rapid processing of the high-priority tasks can be ensured, time delay is remarkably reduced, and the user experience is improved. Therefore, the voice device can quickly and accurately respond to the voice instruction of the user.

Description

Technical Field

[0001] This application relates to the field of computers, and particularly to a response method, apparatus, electronic device, and storage medium for a voice device. Background Art

[0002] With the development of the Internet of Things, voice devices are increasingly widely used in fields such as smart homes, voice assistants, and intelligent vehicles. These devices have greatly improved the convenience and intelligence level of user interaction through voice recognition technology.

[0003] Currently, traditional voice recognition and response systems mostly rely on cloud computing, which is prone to high latency and bandwidth bottlenecks. To alleviate the limitations of cloud computing, the application of edge computing technology has been proposed in the industry. By sinking computing power to the network edge, that is, the device side or a data center close to the device, edge computing effectively shortens the data processing and response time and reduces the response latency of voice devices.

[0004] However, although existing edge computing technologies can reduce response latency, they lack effective task priority scheduling and resource allocation strategies, which makes it difficult to efficiently process and ensure low-latency responses of voice devices in high-concurrency scenarios, that is, when multiple people issue voice commands simultaneously. Summary of the Invention

[0005] This application provides a response method, apparatus, electronic device, and storage medium for a voice device to solve the technical problem that in the high-concurrency scenario of voice commands in the prior art, it is difficult to efficiently process and ensure low-latency responses of voice devices.

[0006] In a first aspect, this application provides a response method for a voice device, and the method includes:

[0007] In response to multiple concurrent voice commands, create voice tasks for each of the voice commands respectively, and determine the priorities of the respective voice tasks;

[0008] According to the priorities of the voice tasks, sequentially schedule the voice tasks by using a time-slice-based scheduling strategy. Among them, when scheduling the voice tasks, allocate time slices for the voice tasks and execute the voice tasks within the time slices so that the voice device responds to the voice commands.

[0009] In a possible implementation manner, the determining the priorities of the respective voice tasks includes:

[0010] Perform semantic analysis on the voice command to obtain the user intention of the voice command;

[0011] According to the user intention of the voice command, determine the dependency relationships among the multiple voice tasks;

[0012] Determine the priority of each of the voice tasks according to the user intention of the voice instruction and the dependency relationship between the multiple voice tasks.

[0013] In a possible implementation manner, the method further includes:

[0014] Adopt a preset priority dynamic adjustment strategy to adjust the current priorities of the multiple voice tasks to obtain the latest priorities of the multiple voice tasks, so as to execute the step of scheduling and executing the voice tasks in sequence according to the latest priorities based on the priority of the voice tasks and adopting a scheduling strategy based on time slices.

[0015] In a possible implementation manner, the executing the voice task within the time slice includes:

[0016] Within the time slice, allocate processing resources to the voice task according to the priority of the voice task, so as to execute the voice task by using the processing resources.

[0017] In a possible implementation manner, the priority of the voice task includes a first priority. The allocating processing resources to the voice task according to the priority of the voice task within the time slice includes:

[0018] When the priority of the voice task is the first priority, allocate processing resources on an edge node to the voice task within the time slice, so that the edge node executes the voice task.

[0019] In a possible implementation manner, the priority of the voice task includes a second priority. The allocating processing resources to the voice task according to the priority of the voice task within the time slice includes:

[0020] When the priority of the voice task is the second priority, allocate processing resources on an edge node and in the cloud to the voice task within the time slice, so that the edge node and the cloud cooperate to execute the voice task.

[0021] In a possible implementation manner, the priority of the voice task includes a third priority. The allocating processing resources to the voice task according to the priority of the voice task within the time slice includes:

[0022] When the priority of the voice task is the third priority, allocate batch processing resources on an edge node and / or in the cloud to the voice task within the time slice, so that the edge node and / or the cloud perform batch processing on the voice task.

[0023] In a second aspect, the present application provides a response device for a voice device, the device comprising:

[0024] A priority determination module, configured to create a voice task for each of the plurality of concurrent voice commands, and determine the priority of each of the voice tasks;

[0025] A scheduling module, configured to sequentially schedule and execute the voice tasks according to the priorities of the voice tasks by adopting a time-slice-based scheduling strategy, wherein when scheduling the voice tasks, time slices are allocated to the voice tasks, and the voice tasks are executed within the time slices, so that the voice device responds to the voice commands.

[0026] In a possible implementation manner, the priority determination module includes:

[0027] A semantic analysis unit, configured to perform semantic analysis on the voice command to obtain the user intention of the voice command;

[0028] A dependency analysis unit, configured to determine the dependency relationship between the plurality of voice tasks according to the user intention of the voice command;

[0029] A determination unit, configured to determine the priority of each of the voice tasks according to the user intention of the voice command and the dependency relationship between the plurality of voice tasks.

[0030] In a possible implementation manner, the device further includes:

[0031] An adjustment unit, configured to adjust the current priorities of the plurality of voice tasks by adopting a preset priority dynamic adjustment strategy to obtain the latest priorities of the plurality of voice tasks, so as to perform the step of sequentially scheduling and executing the voice tasks according to the priorities of the voice tasks based on the latest priorities.

[0032] In a possible implementation manner, the scheduling module includes:

[0033] A resource scheduling unit, configured to allocate processing resources to the voice task according to the priority of the voice task within the time slice, so as to execute the voice task by using the processing resources.

[0034] In a possible implementation manner, the priority of the voice task includes a first priority, and the resource scheduling unit is configured to, when the priority of the voice task is the first priority, allocate the processing resources on the edge node to the voice task within the time slice, so that the edge node executes the voice task.

[0035] In a possible implementation, the priority of the voice task includes a second priority. The resource scheduling unit is configured to, when the priority of the voice task is the second priority, allocate processing resources on the edge node and the cloud to the voice task within the time slice, so that the edge node and the cloud cooperate to execute the voice task.

[0036] In a possible implementation, the priority of the voice task includes a third priority. The resource scheduling unit is configured to, when the priority of the voice task is the third priority, allocate batch processing resources on the edge node and / or the cloud to the voice task within the time slice, so that the edge node and / or the cloud perform batch processing on the voice task.

[0037] In a third aspect, the present application provides an electronic device, including: a processor and a memory. The processor is configured to execute the response program of the voice device stored in the memory to implement the voice device response method according to any one of the first aspects.

[0038] In a fourth aspect, the present application provides a storage medium storing one or more programs, and the one or more programs can be executed by one or more processors to implement the voice device response method according to any one of the first aspects.

[0039] The above technical solutions provided by the embodiments of the present application have the following advantages compared with the prior art: In the method provided by the embodiments of the present application, in the case of concurrent multiple voice commands, in response to the concurrent multiple voice commands, voice tasks are respectively created for each voice command, and the priorities of the respective voice tasks are determined. According to the priorities of the voice tasks, a scheduling strategy based on time slices is adopted to schedule and execute the voice tasks in sequence. Among them, when scheduling the voice task, a time slice is allocated to the voice task, and the voice task is executed within the time slice, so that the voice device responds to the voice command, ensuring that in the case of concurrent multiple voice commands, the edge node can process these voice commands orderly and efficiently, thereby providing a stable and reliable user experience. At the same time, by reasonably allocating time slices and preferentially processing high-priority tasks, it is possible to ensure the rapid processing of high-priority tasks, significantly reduce latency, so as to ensure that the voice device can quickly and accurately respond to the user's voice command. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings here are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0041] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0042] One or more embodiments are exemplarily illustrated by the pictures in the corresponding accompanying drawings. These exemplary illustrations do not limit the embodiments. Elements with the same reference numerals in the drawings represent similar elements. Unless otherwise stated, the drawings in the figures do not constitute a scale limitation.

[0043] Figure 1 It is a flowchart of an embodiment of a response method for a voice device provided by an embodiment of the present application;

[0044] Figure 2 It is a flowchart of an embodiment of another response method for a voice device provided by an embodiment of the present application;

[0045] Figure 3 It is a flowchart of an embodiment of another response method for a voice device provided by an embodiment of the present application;

[0046] Figure 4 It is a block diagram of an embodiment of a voice device response system provided by an embodiment of the present application;

[0047] Figure 5 It is a block diagram of an embodiment of a response device for a voice device provided by an embodiment of the present application;

[0048] Figure 6 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts fall within the scope of protection of the present application.

[0050] The following disclosure provides many different embodiments or examples for implementing different structures of the present application. To simplify the disclosure of the present application, the components and settings of specific examples are described below. Of course, they are only examples and are not intended to limit the present application. In addition, the present application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplification and clarity and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0051] To solve the technical problem in the prior art that it is difficult to efficiently process and ensure low-latency response of a voice device in a high-concurrency scenario of voice commands, the present application provides a response method, device, electronic device, and storage medium for a voice device, which can enable an edge node to orderly and efficiently process these voice commands when multiple voice commands are concurrent, thereby providing a stable and reliable user experience. At the same time, by reasonably allocating time slices and preferentially processing high-priority tasks, it can ensure the rapid processing of high-priority tasks and significantly reduce latency to ensure that the voice device can quickly and accurately respond to the user's voice commands.

[0052] Figure 1 It is a flowchart of an embodiment of a response method for a voice device provided by an embodiment of the present application.

[0053] In one embodiment, the method is applied to an edge node in the Internet of Things. As an exemplary application scenario, the aforementioned Internet of Things is a smart home Internet of Things, and smart homes include but are not limited to smart home devices such as smart TVs, smart air conditioners, and smart refrigerators. The edge node is an important role in the smart home Internet of Things and is located between the data source and the cloud. The edge node can perform data processing and analysis locally, reducing dependence on the cloud, reducing latency, and saving bandwidth. For example, the voice commands of the user received by the smart speaker can be recognized at the edge node and corresponding operations can be performed without sending all voice data to the cloud for processing. Since the edge node is close to the user terminal, it can respond to the user's requests and commands faster. This helps to improve the overall response speed and user experience of the smart home system. Further, in a scenario where multiple devices and multiple users concurrently issue voice commands, the edge node faces the problem of effectively managing and scheduling the voice tasks corresponding to these voice commands. The embodiments of the present application aim to solve this problem. As Figure 1 shown, the method includes the following steps:

[0054] Step 101, in response to multiple concurrent voice commands, create voice tasks for each voice command respectively and determine the priority of each voice task.

[0055] In one embodiment, the above-mentioned multiple concurrent voice commands come from different voice devices. For example, smart speakers, smart air conditioners, smart TVs, etc. in the smart home Internet of Things. Exemplarily, when a user issues a voice command to a voice device, the voice device collects a voice signal through a high-performance array microphone. Subsequently, the voice device performs speech recognition processing on the collected voice signal, including but not limited to stages such as identifying the start and end of the voice signal, noise suppression processing, feature extraction processing, pattern recognition, etc., converts the voice signal into a text-form voice command, and then transmits it to the edge node for further processing.

[0056] When the edge node receives multiple concurrent voice commands, it creates an independent voice task for each voice command. These voice tasks usually involve various types, such as turning off / on an emergency indicator light, playing music, log analysis, etc. Further, in a multi-task environment, resources are limited, and the urgency and importance of different tasks may also vary. Therefore, the edge node can determine the priorities of the multiple concurrent voice tasks to decide which voice tasks should be executed first and which voice tasks can be postponed.

[0057] Exemplarily, the determination of voice task priorities is based on multiple factors, including but not limited to the urgency of the task, the complexity of the task, the source of the task, the user's behavior habits, and the current load condition of the system, etc. Complex calculation logic may also be involved in determining the priorities of voice tasks to ensure that the system can both respond to emergency needs and maintain overall performance and user experience.

[0058] Step 102: According to the priorities of the voice tasks, adopt a time-slice-based scheduling strategy to schedule the voice tasks in sequence. Among them, when scheduling a voice task, allocate a time slice for the voice task, and execute the voice task within the time slice to enable the voice device to respond to the voice command.

[0059] The time-slice-based scheduling strategy means that: divide time into multiple time periods of a certain length, and only allow one task to execute within each time period. This scheduling strategy allows the system to effectively manage multiple tasks, ensuring that each task can obtain a certain amount of processing time, and at the same time can prevent a single task from occupying the processor resources for a long time, resulting in other tasks being unable to be executed in a timely manner.

[0060] In step 102, according to the priority of the voice task, the scheduling strategy based on time slices means that the execution entity of the embodiment of the present application will, according to the priority of the voice task, allocate an independent time slice for each voice task. Among them, the voice task with a higher priority will be allocated a time slice earlier in time (optionally, the voice task with a higher priority will be allocated a longer time slice), while the voice task with a lower priority will be allocated a time slice later in time to ensure that the voice task with a higher priority is executed first, so as to ensure the fast processing of high-priority tasks and significantly reduce latency. Within the allocated time slice, the execution entity of the embodiment of the present application will focus on executing the voice task and perform corresponding processing operations to make the voice device respond to the voice instruction, such as playing music, turning on / off the light, etc., until the time slice ends or the task execution is completed. If the voice task is not completed within the allocated time slice, it may be suspended temporarily and wait for the next available time slice to continue execution.

[0061] It can be seen that, according to the priority of the voice task, adopting the scheduling strategy based on time slices to schedule voice tasks in sequence can realize the orderly execution of voice tasks according to the priority of the tasks, so as to ensure that in a multi-voice task concurrent environment, the edge node can provide a stable and reliable voice task execution mechanism, thereby ensuring that the voice device can quickly and accurately respond to the user's voice instruction.

[0062] The technical solution provided by the embodiment of the present application, in the case of multiple concurrent voice instructions, in response to the multiple concurrent voice instructions, creates voice tasks for each voice instruction respectively, determines the priorities of the respective voice tasks, and according to the priorities of the voice tasks, adopts a scheduling strategy based on time slices to schedule and execute the voice tasks in sequence. Among them, when scheduling the voice tasks, time slices are allocated for the voice tasks, and the voice tasks are executed within the time slices to make the voice device respond to the voice instructions, ensuring that in the case of multiple concurrent voice instructions, the edge node can process these voice instructions orderly and efficiently, thereby providing a stable and reliable user experience. At the same time, by reasonably allocating time slices and preferentially processing high-priority tasks, it can ensure the fast processing of high-priority tasks and significantly reduce the latency, so as to ensure that the voice device can quickly and accurately respond to the user's voice instructions.

[0063] Figure 2 It is a flowchart of an embodiment of another response method of a voice device provided by an embodiment of the present application. Figure 2 The process shown is Figure 1 On the basis of the process shown, it describes how to determine the priorities of the respective voice tasks. As Figure 2 shown, it includes the following steps:

[0064] Step 201, perform semantic analysis on the voice instruction to obtain the user intention of the voice instruction.

[0065] As an optional implementation, the voice command can be input into a trained semantic analysis model, such as a BERT model or a GPT model. This semantic model is based on deep learning natural language processing technology and can understand and parse complex text content. Then, through the processing of the semantic analysis model, the user intention behind the voice command can be identified. These user intentions can be classified into different functional categories, such as network configuration, device control, data processing, and other functional categories for subsequent processing and response.

[0066] Step 202: Determine the dependency relationship between multiple voice tasks according to the user intention of the voice command.

[0067] In step 202, determining the dependency relationship between multiple voice tasks includes determining the order of execution of multiple voice tasks (i.e., which tasks need to be executed first, which tasks can be executed in parallel or subsequently) and the dependency between tasks (i.e., whether the execution of one task depends on the completion of another task). For example, the system initialization task needs to be executed before the device startup task, and the two cannot be executed in parallel.

[0068] Step 203: Determine the priority of each voice task according to the user intention of the voice command and the dependency relationship between multiple voice tasks.

[0069] According to the user intention of the voice command, the urgency and importance of the voice task can be judged. For example, tasks related to user safety may be regarded as high-priority, while some non-critical tasks may be given lower priorities. For example, if a task is used for emergency repair of the device, although its computational complexity is high, its priority still needs to be increased.

[0070] In addition to considering the urgency and importance of the task, step 203 also considers the impact of the dependency relationship between tasks on the priority. For example, if a task is a prerequisite for another high-priority task, then even if it is not urgent itself, it needs to be executed first.

[0071] Then, according to the user intention of the voice command and the dependency relationship between multiple voice tasks, the priority of each voice task can be accurately determined, which helps to ensure that multiple voice tasks are scheduled and executed in the correct order and priority to meet user needs.

[0072] Figure 2 The shown process, by parsing the voice command, identifying the user intention, analyzing the task dependency relationship, and determining the task priority, helps to ensure that the system can correctly and efficiently respond to the user's voice command.

[0073] In one embodiment, the application can be Figure 2The priority determined by the shown process is regarded as the initial priority, and the scheduling and execution of voice tasks are carried out according to this initial priority. Further, during the actual execution process, the priority of voice tasks can be dynamically adjusted according to the actual situation, which is usually to ensure the efficiency of task execution, maximize resource utilization, and handle some unpredictable factors. For example, if a high-priority voice task encounters problems or delays, it may be necessary to temporarily increase the priority of other related tasks to ensure the stability and performance of the overall system. Another example is that when network resources are scarce, commands related to communication are preferentially executed to avoid affecting data transmission.

[0074] Based on this, a preset priority dynamic adjustment strategy is adopted to adjust the current priorities of multiple voice tasks to obtain the latest priorities of multiple voice tasks, so as to execute the steps of scheduling and executing voice tasks in sequence according to the priorities of voice tasks based on the latest priorities.

[0075] Exemplarily, the preset priority dynamic adjustment strategy is designed based on the AHP (Analytic Hierarchy Process) algorithm or the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) algorithm.

[0076] Exemplarily, the preset priority dynamic adjustment strategy can also be adjusted in real time using online learning algorithms (such as FTRL, Lifelong Learning), enabling the system to adapt to the needs of different users and optimize the priority processing of tasks. For example, by learning information such as user behavior, device status, and network conditions in real time, the system can continuously update the priority dynamic adjustment strategy to adapt to external changes and improve the intelligence and personalization capabilities of the system. For example, the system can learn in real time how to adjust the response speed according to different user operation habits, or how to adjust the data transmission strategy according to changes in network bandwidth. Another example is that based on the user's historical behavior and operation habits, the system can intelligently predict and optimize the priority of tasks. For instance, if the user frequently operates the air conditioning system during a certain period, the system will schedule tasks related to the air conditioning to a higher priority, while postponing other less urgent tasks (such as network updates).

[0077] Figure 3 This is a flowchart of an embodiment of another response method for a voice device provided by an embodiment of the present application. Figure 3 The shown process is based on Figure 1 On the basis of the shown process, it includes the following steps:

[0078] Step 301: In response to multiple concurrent voice commands, create voice tasks for each voice command respectively and determine the priorities of the respective voice tasks.

[0079] For a detailed description of step 301, please refer to Figure 1 the relevant description in the illustrated embodiments, which will not be elaborated here.

[0080] Step 302: According to the priorities of the voice tasks, adopt a time-slice-based scheduling strategy to schedule and execute the voice tasks in sequence. Among them, when scheduling a voice task, allocate a time slice for the voice task. Within this time slice, allocate processing resources for the voice task according to the priority of the voice task, so as to execute the voice task by using the processing resources.

[0081] In one embodiment, in combination with the time-slice technology, adopt a latency-aware hierarchical processing mechanism to divide the voice tasks into three levels: immediate response, delay-tolerant, and batch processing, so as to effectively balance the priorities of the tasks and the latency requirements and optimize the use of resources.

[0082] Specifically, the immediate response layer means that high-priority tasks are directly processed at the edge node to ensure the response speed of high-priority tasks. The delay-tolerant layer means that medium-priority tasks are processed jointly by the edge node and the cloud to save the computing resources of the edge node. The batch processing layer means that batch processing is performed during low-load periods.

[0083] It can be seen that through the cooperation of the time-slice technology and the hierarchical processing mechanism, it is possible to ensure that tasks at each level can be processed in a timely and efficient manner.

[0084] Accordingly, in step 302, the priorities of the voice tasks include a first priority. The specific implementation of allocating processing resources for the voice task according to the priority of the voice task within the time slice includes: when the priority of the voice task is the first priority, allocate the processing resources on the edge node for the voice task within the time slice, so that the edge node executes the voice task. Among them, the first priority is a high priority.

[0085] The priorities of the voice tasks include a second priority. The specific implementation of allocating processing resources for the voice task according to the priority of the voice task within the time slice includes: when the priority of the voice task is the second priority, allocate the processing resources on the edge node and the cloud for the voice task within the time slice, so that the edge node and the cloud jointly execute the voice task. Among them, the second priority is a medium priority.

[0086] The priorities of the voice tasks include a third priority. The specific implementation of allocating processing resources for the voice task according to the priority of the voice task within the

[0087] When the priority of the voice task is the third priority, batch processing resources on the edge node and / or in the cloud are allocated to the voice task within the time slice for batch processing of the voice task by the edge node and / or in the cloud.

[0088] Figure 3 The process shown allocates processing resources to a voice task according to the priority of the voice task to execute the voice task using the processing resources, providing a multi-task processing mechanism that coordinates the time slice technology and the hierarchical processing mechanism. The tasks are divided into three levels: immediate response, delay tolerance, and batch processing. For tasks at different levels, processing resources are allocated differentially, which can effectively balance the priority of the tasks and the delay requirements, optimize the use of resources, and ensure that tasks at all levels can be processed in a timely and efficient manner.

[0089] Figure 4 This is a block diagram of an embodiment of a voice device response system provided by an embodiment of the present application. As Figure 4 shown, the system includes a voice recognition module, an edge node, a cloud server, and a privacy protection module.

[0090] Among them, the voice recognition module is the voice device. The voice recognition module is used for capturing and preprocessing voice commands in the Figure 4 shown system, including a voice recognition unit, a 5G communication module, and an edge computing unit.

[0091] Specifically, when the user issues a voice command to the voice recognition module, that is, the voice device, the device collects the voice signal through a high-performance array microphone, and the sampling rate is set to 16 kHz or higher to ensure clear voice data is captured. To reduce the interference of background noise, the device applies voice activity detection (VAD) technology to accurately identify the start and end of the voice signal. Subsequently, the signal undergoes noise suppression processing (such as Wiener filtering) to ensure the clarity of the voice. In the feature extraction stage, the device uses the Mel frequency cepstral coefficient algorithm to convert the voice signal into a feature vector and inputs it into a deep neural network or a convolutional neural network for pattern recognition, and identifies and converts it into a text-form voice command. This process ensures a high accuracy rate of voice recognition, and the recognition result will be transmitted to the edge node for further processing.

[0092] The edge node in Figure 4In the system shown, to execute the method provided in the embodiments of the present application, first, through semantic analysis of voice commands (such as using the BERT model or GPT model), priorities are assigned to each voice command, and according to the priorities of tasks, a scheduling strategy based on time slices is adopted to schedule multiple voice tasks sequentially. Exemplarily, emergency tasks (such as "turn off the emergency lights") are assigned to time slices with higher priorities and are processed preferentially. Through time slices, it can be ensured that high-priority tasks are preferentially executed within the specified time. In addition, during this process, the edge node can also dynamically adjust the policy based on priorities, such as the AHP or TOPSIS algorithm, dynamically adjust the sorting of tasks, that is, priorities, and perform scheduling in the task queue of the edge node.

[0093] Among them, the edge node updates the processing model of the priority dynamic adjustment policy in real time through a distributed learning framework (such as Horovod or Ray). Based on online learning algorithms (such as FTRL, Lifelong Learning), the system can adjust the priority dynamic adjustment policy based on real-time data, adapt to the needs of different users, and optimize the priority processing of tasks.

[0094] The edge node also adopts a latency-aware hierarchical processing mechanism in coordination with time slices. Specifically, the edge node divides voice tasks into three levels according to the latency requirements of voice tasks:

[0095] 1. Immediate response layer: High-priority tasks (such as "turn on the lights") are directly processed at the edge node and use independent time slices.

[0096] 2. Delay-tolerant layer: Medium-priority tasks (such as "play music") are processed jointly by the edge node and the cloud.

[0097] 3. Batch processing layer: Low-priority tasks (such as voice log analysis) are postponed for batch processing. The privacy protection module is Figure 4 responsible for privacy protection and verifiable computing in the system shown. Among them, the privacy protection module provides transparency and traceability of the data processing process through zero-knowledge proof (ZKP) and blockchain technology, realizes verifiable computing and privacy protection of voice data, and ensures that user data is not leaked during the entire processing process. At the same time, federated learning technology is used to further enhance the privacy and security of data. In addition, the entire system adopts data isolation and encryption strategies within time slices to ensure that the voice data of each user is independently processed within the slice.

[0098] Figure 4 The system shown can achieve the following beneficial effects:

[0099] 1. Improve system response speed: By combining time slicing technology and a task priority scheduling mechanism based on voice features, it can be dynamically adjusted according to the urgency and complexity of voice tasks to ensure the rapid processing of high-priority voice tasks and significantly reduce latency.

[0100] 2. Optimize latency and resource allocation: Adopt a latency-aware hierarchical processing mechanism to divide voice tasks into immediate response, delay-tolerant, and batch processing layers, effectively balancing the priority and latency requirements of voice tasks and optimizing resource utilization.

[0101] 3. Strengthen privacy protection: Introduce zero-knowledge proof (ZKP) and blockchain technology to achieve verifiable computing and privacy protection of voice data, ensuring that user data is not leaked during the entire processing process.

[0102] 4. Improve the personalized experience: Through a real-time distributed learning mechanism, the system can dynamically adjust the task processing strategy according to the user's voice command habits, enhancing the system's adaptability and personalized response ability.

[0103] Figure 5 The following is a block diagram of an embodiment of a response device for a voice device provided in an embodiment of the present application. As Figure 5 shown, the device includes:

[0104] A priority determination module 51, configured to create voice tasks for each of the concurrent multiple voice commands and determine the priorities of the respective voice tasks.

[0105] A scheduling module 52, configured to sequentially schedule and execute the voice tasks according to the priorities of the voice tasks. When scheduling the voice tasks, time slices are allocated to the voice tasks, and the voice tasks are executed within the time slices so that the voice device responds to the voice commands.

[0106] In a possible implementation manner, the priority determination module 51 includes:

[0107] A semantic analysis unit, configured to perform semantic analysis on the voice command to obtain the user intention of the voice command.

[0108] A dependency analysis unit, configured to determine the dependency relationship between the multiple voice tasks according to the user intention of the voice command.

[0109] A determination unit, configured to determine the priorities of the respective voice tasks according to the user intention of the voice command and the dependency relationship between the multiple voice tasks.

[0110] In a possible implementation manner, the device further includes:

[0111] An adjustment unit, configured to adopt a preset priority dynamic adjustment strategy to adjust the current priorities of multiple said voice tasks, so as to obtain the latest priorities of multiple said voice tasks, and based on the latest priorities, perform the step of scheduling and executing the voice tasks in sequence according to the scheduling strategy based on time slices according to the priorities of the voice tasks.

[0112] In a possible implementation manner, the scheduling module 52 includes:

[0113] A resource scheduling unit, configured to allocate processing resources for the voice tasks according to the priorities of the voice tasks within the time slice, so as to execute the voice tasks by using the processing resources.

[0114] In a possible implementation manner, the priority of the voice task includes a first priority. The resource scheduling unit is configured to, when the priority of the voice task is the first priority, allocate processing resources on an edge node for the voice task within the time slice, so that the edge node executes the voice task.

[0115] In a possible implementation manner, the priority of the voice task includes a second priority. The resource scheduling unit is configured to, when the priority of the voice task is the second priority, allocate processing resources on an edge node and in the cloud for the voice task within the time slice, so that the edge node and the cloud cooperate to execute the voice task.

[0116] In a possible implementation manner, the priority of the voice task includes a third priority. The resource scheduling unit is configured to, when the priority of the voice task is the third priority, allocate batch processing resources on an edge node and / or in the cloud for the voice task within the time slice, so that the edge node and / or the cloud perform batch processing on the voice task.

[0117] As Figure 6 shown, an embodiment of the present application provides an electronic device, including a processor 111, a communication interface 112, a memory 113, and a communication bus 114. Among them, the processor 111, the communication interface 112, and the memory 113 complete communication with each other through the communication bus 114.

[0118] The memory 113 is used to store a computer program.

[0119] In an embodiment of the present application, when the processor 111 executes the program stored on the memory 113, it implements the response method of the voice device provided in any one of the foregoing method embodiments, including:

[0120] In response to multiple concurrent voice commands, create voice tasks for each of the voice commands respectively, and determine the priorities of the respective voice tasks;

[0121] According to the priorities of the voice tasks, schedule the voice tasks in sequence using a time-slice-based scheduling strategy. Among them, when scheduling the voice tasks, allocate time slices for the voice tasks, and execute the voice tasks within the time slices, so that the voice device responds to the voice commands.

[0122] An embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the response method of the voice device provided in any of the foregoing method embodiments are implemented.

[0123] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the essence of the above technical solution, or the part that contributes to the related technology, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0125] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" as used herein may also include the plural forms. The terms "include", "comprise", "contain", and "have" are inclusive and thus specify the presence of the stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or combinations thereof. The method steps, processes, and operations described herein are not to be construed as necessarily requiring them to be executed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that alternative or additional steps may be used.

[0126] The above are only specific embodiments of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown herein, but rather will be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A response method of a voice device, characterized in that: The method comprises: In response to multiple concurrent voice commands, create a voice task for each of the voice commands, and determine the priority of each of the voice tasks; According to the priority of the voice task, the voice tasks are scheduled in sequence using a time slice-based scheduling strategy, wherein when scheduling the voice task, a time slice is allocated to the voice task, and the voice task is executed within the time slice so that the voice device responds to the voice command.

2. The method according to claim 1, characterized in that: Determining the priority of each of the voice tasks includes: Performing semantic analysis on the voice command to obtain the user intention of the voice command; Determining dependencies between the plurality of voice tasks according to a user intention of the voice command; The priority of each of the voice tasks is determined according to the user intention of the voice instruction and the dependency relationship between the multiple voice tasks.

3. The method according to claim 1 or 2, characterized in that: The method further comprises: A preset priority dynamic adjustment strategy is adopted to adjust the current priorities of the multiple voice tasks to obtain the latest priorities of the multiple voice tasks, and the steps of scheduling and executing the voice tasks in sequence according to the priorities of the voice tasks using a scheduling strategy based on time slicing are executed based on the latest priorities.

4. The method according to claim 1, characterized in that The performing the speech task within the time slice comprises: In the time slice, processing resources are allocated to the voice task according to the priority of the voice task, so as to execute the voice task by utilizing the processing resources.

5. The method according to claim 4, characterized in that The priority of the voice task includes a first priority, and within the time slice, allocating processing resources to the voice task according to the priority of the voice task includes: When the priority of the voice task is the first priority, a processing resource on an edge node is allocated to the voice task within the time slice so that the edge node executes the voice task.

6. The method according to claim 4, characterized in that The priority of the voice task includes a second priority, and within the time slice, allocating processing resources to the voice task according to the priority of the voice task includes: When the priority of the voice task is the second priority, processing resources on the edge node and the cloud are allocated to the voice task within the time slice, so that the edge node and the cloud collaborate to execute the voice task.

7. The method according to claim 4, characterized in that The priority of the voice task includes a third priority, and within the time slice, allocating processing resources to the voice task according to the priority of the voice task includes: When the priority of the voice task is the third priority, batch processing resources on the edge node and / or the cloud are allocated to the voice task within the time slice, so that the edge node and / or the cloud performs batch processing on the voice task.

8. A response device for a voice device, characterized in that: The device comprises: A priority determination module, for creating a voice task for each of the voice commands in response to a plurality of concurrent voice commands, and determining a priority of each of the voice tasks; A scheduling module is used to schedule and execute the voice tasks in sequence according to the priority of the voice tasks using a time slice-based scheduling strategy, wherein when scheduling the voice tasks, a time slice is allocated to the voice tasks, and the voice tasks are executed within the time slice so that the voice device responds to the voice command.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the processor is used to execute a response program of a voice device stored in the memory to implement the response method of a voice device according to any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the response method of the voice device according to any one of claims 1 to 7.

Images