Interaction instruction processing method and device, vehicle and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, in particular to an interaction instruction processing method and device, a vehicle and a storage medium. The interaction instruction processing method comprises the steps that when a cloud side instruction reasoning result and an end side instruction reasoning result are received within a preset instruction period, the cloud side instruction reasoning result and the end side instruction reasoning result are received; respectively identifying a first confidence value of the cloud side instruction reasoning result and a second confidence value of the end side instruction reasoning result, and when the first confidence value is greater than the second confidence value, controlling the vehicle to execute corresponding operation based on the cloud side instruction reasoning result corresponding to the first confidence value, otherwise, controlling the vehicle to execute corresponding operation. And controlling the vehicle to execute corresponding operation based on an end-side instruction reasoning result corresponding to the second confidence value. Therefore, the problems of high delay, serious network dependence and the like caused by pure dependence on cloud computing resources when a large model is adopted for text reasoning are solved, reasoning is carried out on text instruction information through cooperation of the cloud side and the end side, instruction reasoning results on the two sides are recognized through the end-cloud arbitration module, and an instruction with a high confidence value is executed.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and particularly relates to an interactive instruction processing method, apparatus, vehicle, and storage medium. Background Art

[0002] With the development of artificial intelligence technology, the vehicle-mounted voice instruction processing method based on rules is difficult to meet the growing user interaction experience requirements, and the text processing technology based on large models has been widely used in fields such as vehicle-mounted systems.

[0003] In related technologies, when introducing a large model for text reasoning, it often relies on the processing power of the cloud, but relying solely on cloud computing will bring problems of latency and network dependence. In addition, only using local devices will also limit the use of large models and reduce the ability to make full use of large models, which urgently needs to be solved. Summary of the Invention

[0004] This application provides an interactive instruction processing method, apparatus, vehicle, and storage medium to solve the problems of high latency and serious network dependence caused by solely relying on cloud computing resources when using a large model for text reasoning.

[0005] The first aspect embodiment of this application provides an interactive instruction processing method, including the following steps:

[0006] Determine whether the cloud-side instruction reasoning result and the end-side instruction reasoning result are received within a preset instruction cycle;

[0007] If the cloud-side instruction reasoning result and the end-side instruction reasoning result are received within the preset instruction cycle, respectively identify the first confidence value of the cloud-side instruction reasoning result and the second confidence value of the end-side instruction reasoning result;

[0008] Determine whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, control the vehicle to perform corresponding operations based on the cloud-side instruction reasoning result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, control the vehicle to perform corresponding operations based on the end-side instruction reasoning result corresponding to the second confidence value.

[0009] According to an embodiment of this application, before determining whether the cloud-side instruction reasoning result and the end-side instruction reasoning result are received within a preset instruction cycle, it further includes:

[0010] Determine whether the voice instruction information of the user is received;

[0011] If the voice command information is received, the voice command information is processed for voice data, and the voice command information after voice data processing is converted into corresponding text command information;

[0012] The text command information is respectively sent to the cloud - side command inference component and the end - side command inference component. The cloud - side command inference component performs cloud - side inference on the text command information to obtain a cloud - side command inference result, and the end - side command inference component performs end - side inference on the text command information to obtain an end - side command inference result.

[0013] According to an embodiment of the present application, after determining whether the cloud - side command inference result and the end - side command inference result are received within a preset command cycle, it further includes:

[0014] If the cloud - side command inference result and the end - side command inference result are not received within the preset command cycle, an instruction response failure prompt is generated.

[0015] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, it further includes:

[0016] If the first confidence value is equal to the second confidence value, the first confidence value and the second confidence value are identified based on a preset instruction execution standard. When the first confidence value meets the preset instruction execution standard, the vehicle is controlled to perform corresponding operations based on the cloud - side command inference result corresponding to the first confidence value, or when the second confidence value meets the preset instruction execution standard, the vehicle is controlled to perform corresponding operations based on the end - side command inference result corresponding to the second confidence value.

[0017] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, it further includes:

[0018] If the first confidence value is greater than the second confidence value, the first confidence value is stored in the target data set, otherwise, the second confidence value is stored in the target data set.

[0019] According to the interaction instruction processing method of the embodiments of the present application, when the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle, the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result are respectively identified. When the first confidence value is greater than the second confidence value, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. Otherwise, the vehicle is controlled to perform corresponding operations based on the end-side instruction inference result corresponding to the second confidence value. Thus, the problems of high latency and severe network dependence caused by solely relying on the computing resources of the cloud when using large models for text inference are solved. The text instruction information is inferred through the cooperation of the cloud side and the end side, and the instruction inference results on both sides are identified by the end-cloud arbitration module, and the instruction with a higher confidence value is executed.

[0020] The second aspect of the embodiments of the present application provides an interaction instruction processing device, including:

[0021] A judgment module, configured to judge whether the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle;

[0022] An identification module, configured to, if the cloud-side instruction inference result and the end-side instruction inference result are received within the preset instruction cycle, respectively identify the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result;

[0023] A control module, configured to judge whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, the vehicle is controlled to perform corresponding operations based on the end-side instruction inference result corresponding to the second confidence value.

[0024] According to an embodiment of the present application, before judging whether the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle, the judgment module is further configured to:

[0025] Judge whether the voice instruction information of the user is received;

[0026] If the voice instruction information is received, the voice instruction information is subjected to voice data processing, and the voice instruction information after voice data processing is converted into corresponding text instruction information;

[0027] Send the text instruction information to the cloud-side instruction inference component and the edge-side instruction inference component respectively. The cloud-side instruction inference component performs cloud-side inference on the text instruction information to obtain a cloud-side instruction inference result, and the edge-side instruction inference component performs edge-side inference on the text instruction information to obtain an edge-side instruction inference result.

[0028] According to an embodiment of the present application, after determining whether the cloud-side instruction inference result and the edge-side instruction inference result are received within a preset instruction cycle, the determination module is further configured to:

[0029] If the cloud-side instruction inference result and the edge-side instruction inference result are not received within the preset instruction cycle, generate a prompt indicating that the instruction response fails.

[0030] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, the control module is further configured to:

[0031] If the first confidence value is equal to the second confidence value, identify the first confidence value and the second confidence value based on a preset instruction execution standard, and when the first confidence value meets the preset instruction execution standard, control the vehicle to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value, or when the second confidence value meets the preset instruction execution standard, control the vehicle to perform corresponding operations based on the edge-side instruction inference result corresponding to the second confidence value.

[0032] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, the control module is further configured to:

[0033] If the first confidence value is greater than the second confidence value, store the first confidence value in the target data set; otherwise, store the second confidence value in the target data set.

[0034] According to the interaction instruction processing device in the embodiment of the present application, when receiving the cloud-side instruction inference result and the end-side instruction inference result within a preset instruction cycle, it respectively identifies the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result, and when the first confidence value is greater than the second confidence value, it controls the vehicle to execute corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value, otherwise, it controls the vehicle to execute corresponding operations based on the end-side instruction inference result corresponding to the second confidence value. Thus, it solves the problems of high latency and severe network dependence caused by simply relying on the computing resources of the cloud when using large models for text inference. By jointly reasoning on text instruction information between the cloud side and the end side, and having the end-cloud arbitration module identify the instruction inference results on both sides and execute the instruction with a higher confidence value.

[0035] The third aspect of the present application provides a vehicle, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the program to implement the interaction instruction processing method as described in the above embodiment.

[0036] The fourth aspect of the present application provides a computer-readable storage medium, and the computer-readable storage medium stores computer instructions for causing the computer to execute the interaction instruction processing method as described in the above embodiment.

[0037] The fifth aspect of the present application provides a computer program product, including a computer program, and the computer program is executed to implement the interaction instruction processing method as described in the above embodiment.

[0038] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. Description of the Drawings

[0039] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0040] Figure 1 It is a flowchart of an interaction instruction processing method provided according to an embodiment of the present application;

[0041] Figure 2 It is a schematic diagram of the overall architecture of end-cloud collaboration according to an embodiment of the present application;

[0042] Figure 3 It is a flowchart of end-cloud collaborative processing according to an embodiment of the present application;

[0043] Figure 4 It is a schematic diagram of end-cloud model inference according to an embodiment of the present application;

[0044] Figure 5 Schematic diagram of end-cloud arbitration and high-confidence instruction caching service according to an embodiment of the present application;

[0045] Figure 6 Example diagram of an interactive instruction processing device according to an embodiment of the present application;

[0046] Figure 7 Schematic structural diagram of a vehicle according to an embodiment of the present application. Detailed implementation manners

[0047] The embodiments of the present application will be described in detail below. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements with the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as a limitation to the present application.

[0048] The interactive instruction processing method, device, vehicle and storage medium of the embodiments of the present application will be described below with reference to the accompanying drawings. In view of the problems of high latency and severe network dependence caused by solely relying on cloud computing resources when using large models for text reasoning in the above-mentioned background art, the present application provides an interactive instruction processing method. In this method, when the cloud-side instruction reasoning result and the end-side instruction reasoning result are received within a preset instruction cycle, the first confidence value of the cloud-side instruction reasoning result and the second confidence value of the end-side instruction reasoning result are respectively identified, and when the first confidence value is greater than the second confidence value, the vehicle is controlled to execute corresponding operations based on the cloud-side instruction reasoning result corresponding to the first confidence value; otherwise, the vehicle is controlled to execute corresponding operations based on the end-side instruction reasoning result corresponding to the second confidence value. Thus, the problems of high latency and severe network dependence caused by solely relying on cloud computing resources when using large models for text reasoning are solved. The text instruction information is inferred through the cooperation of the cloud side and the end side, and the instruction reasoning results on both sides are identified by the end-cloud arbitration module, and the instruction with a higher confidence value is executed.

[0049] Specifically, Figure 1 Flowchart of an interactive instruction processing method provided by an embodiment of the present application.

[0050] As Figure 1 shown, the interactive instruction processing method includes the following steps:

[0051] In step S101, it is judged whether the cloud-side instruction reasoning result and the end-side instruction reasoning result are received within a preset instruction cycle.

[0052] According to an embodiment of the present application, before determining whether the cloud-side instruction inference result and the edge-side instruction inference result are received within a preset instruction cycle, it further includes: determining whether a voice instruction message of the user is received; if the voice instruction message is received, performing voice data processing on the voice instruction message, and converting the voice instruction message after voice data processing into a corresponding text instruction message; sending the text instruction message to a cloud-side instruction inference component and an edge-side instruction inference component respectively, where the cloud-side instruction inference component performs cloud-side inference on the text instruction message to obtain a cloud-side instruction inference result, and the edge-side instruction inference component performs edge-side inference on the text instruction message to obtain an edge-side instruction inference result.

[0053] Among them, the preset instruction cycle can be set by those skilled in the art according to actual test requirements, or obtained through a limited number of computer simulations, and no specific limitation is made here.

[0054] Specifically, as Figures 2 to 4 shown, when a user uses a vehicle and wants to interact with the vehicle to control related functions of the vehicle, the user can first issue a voice instruction message. For example, issue a voice instruction message of "turn on the air conditioner". At this time, it is collected by the cockpit voice assistant. After the voice instruction message issued by the user is collected, data processing is performed on the voice instruction message, including noise reduction, feature extraction, etc. on the voice instruction message, and then the preprocessed voice instruction message is converted into a text instruction message. In order to provide an efficient and reliable text instruction processing service, in the embodiment of the present application, the text instruction message is sent to a cloud-side instruction inference component and an edge-side instruction inference component respectively. The cloud-side instruction inference component performs cloud-side inference on the text instruction message to obtain a cloud-side instruction inference result, and the edge-side instruction inference component performs edge-side inference on the text instruction message to obtain an edge-side instruction inference result. At the same time, the text instruction message is sent to the edge-cloud arbitration module on the edge side to inform the edge-cloud arbitration module that there is a new text instruction message to be processed.

[0055] Among them, when the edge-side instruction inference component performs edge-side inference on the text instruction message, it can generate a confidence score for each possible result according to the algorithms inside the model (such as probability distribution, softmax function, etc.). This score reflects the model's estimation of the correctness of the result, so as to obtain the edge-side instruction inference result; when the cloud-side instruction inference component performs cloud-side inference on the text instruction message, more complex technologies can be used to evaluate the confidence, such as ensemble learning, context understanding, etc., to improve the accuracy of judgment, so as to obtain the cloud-side instruction inference result.

[0056] Furthermore, after the cloud-side instruction inference result is obtained after processing the text instruction information based on the cloud-side instruction inference component, and the end-side instruction inference result is obtained after processing the text instruction information based on the end-side instruction inference component, the cloud-side instruction inference result and the end-side instruction inference result are sent to the end-cloud arbitration module, which receives them within a preset instruction cycle.

[0057] It should be noted that for some simple text instruction information, its features are relatively obvious and easy to identify. The terminal-side instruction reasoning component can usually process these instructions quickly and accurately. Therefore, the terminal-side instruction reasoning component can directly process them and generate terminal-side instruction reasoning results. For some complex text instruction information, due to the limited resources of the terminal-side instruction reasoning component, it may not be able to achieve the ideal processing effect. Therefore, it is necessary to use the large model reasoning service for more in-depth processing. The cloud-side instruction reasoning component can process both simple text instruction information and complex text instruction information. Therefore, the combination of the terminal and cloud sides can ensure that the processing of text instruction information is both efficient and accurate, thereby improving the user experience.

[0058] In step S102, if a cloud-side instruction inference result and a terminal-side instruction inference result are received within a preset instruction cycle, a first confidence value of the cloud-side instruction inference result and a second confidence value of the terminal-side instruction inference result are respectively identified.

[0059] Specifically, Figure 5 As shown, since a certain time period is required to avoid the user from waiting for too long when processing the user's text instruction information, the embodiment of the present application needs to be received within a preset instruction cycle, that is, after the voice assistant converts the user's voice instruction information into text instruction information and sends it out, the timing of the instruction cycle is started. For example, the start time may be T0 and the end time may be T1. If the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within the preset instruction cycle, that is, the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within the time period of T0-T1, the terminal-cloud arbitration module respectively identifies the first confidence value of the cloud-side instruction reasoning result and the second confidence value of the terminal-side instruction reasoning result, and determines the threshold values ​​of the first confidence value and the second confidence value. Since a high confidence value usually means that the result of the instruction is very reliable, that is, the system has high confidence in its accuracy, these instructions are often stable and accurate in multiple processing. Therefore, the terminal-cloud arbitration module needs to identify the first confidence value of the cloud-side instruction reasoning result and the second confidence value of the terminal-side instruction reasoning result, and select the instruction reasoning result corresponding to the higher confidence value.

[0060] It should be noted that, in order to ensure the real-time performance and response speed of the system, especially in time-sensitive application scenarios (such as in-vehicle environments), the maximum allowable time for the entire processing flow (from receiving an instruction to making a decision and executing it) should not exceed the preset instruction cycle, so as to quickly respond to the user's instructions.

[0061] In step S103, it is determined whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, the vehicle is controlled to perform corresponding operations based on the edge-side instruction inference result corresponding to the second confidence value.

[0062] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, it further includes: if the first confidence value is equal to the second confidence value, the first confidence value and the second confidence value are identified based on a preset instruction execution standard, and when the first confidence value meets the preset instruction execution standard, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value, or when the second confidence value meets the preset instruction execution standard, the vehicle is controlled to perform corresponding operations based on the edge-side instruction inference result corresponding to the second confidence value.

[0063] According to an embodiment of the present application, after determining whether the first confidence value is greater than the second confidence value, it further includes: if the first confidence value is greater than the second confidence value, the first confidence value is stored in the target data set, otherwise, the second confidence value is stored in the target data set.

[0064] Among them, the preset instruction execution standard can be set by those skilled in the art according to actual test requirements, or obtained through a limited number of computer simulations, and no specific limitation is made here.

[0065] Specifically, after the edge-cloud arbitration module identifies the first confidence value of the cloud-side instruction inference result and the second confidence value of the edge-side instruction inference result, if it is identified that the first confidence value is greater than the second confidence value, it indicates that the cloud-side instruction inference result is more reliable and performs stably and accurately in multiple processes. Therefore, the cloud-side instruction inference result corresponding to the first confidence value is sent to the edge-side execution system. When the edge-side execution system receives the cloud-side instruction inference result, the cloud-side instruction inference result is then sent to the corresponding execution subunit for execution, and the cloud-side instruction inference result is stored in the target dataset. If it is identified that the first confidence value is less than the second confidence value, it indicates that the edge-side instruction inference result is more reliable and performs stably and accurately in multiple processes. Therefore, the edge-side instruction inference result corresponding to the second confidence value is sent to the edge-side execution system. When the edge-side execution system receives the edge-side instruction inference result, the edge-side instruction inference result is then sent to the corresponding execution subunit for execution, and the edge-side instruction inference result is stored in the target dataset.

[0066] It should be noted that once an instruction is determined to have a high confidence level, the edge-cloud arbitration module will update the relevant information of this instruction (including the instruction itself and its corresponding processing result) to the target dataset on the edge side. This target dataset can be understood as a local storage or database for saving the mapping relationship of known reliable instructions and results. As the system runs, whenever a new high-confidence instruction inference result appears, it will be added to this target dataset. This dynamic update method enables the edge side to continuously learn and accumulate experience, gradually improving its independent processing ability. Thus, when encountering the same or similar instructions subsequently, the edge side can directly search for matching items in the target dataset without having to send them to the cloud side for processing again, which can significantly reduce latency and provide a faster response time, especially important in the case of poor network conditions.

[0067] Furthermore, if the first confidence value is the same as the second confidence value, other factors need to be combined and comprehensively evaluated based on the preset instruction execution criteria.

[0068] Specifically, the first confidence value and the second confidence value can be identified based on the preset instruction execution criteria. When the first confidence value meets the preset instruction execution criteria, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. That is to say, the edge-cloud arbitration module can evaluate based on the current network conditions. If the first confidence value meets the preset instruction execution criteria, that is, the network connection is stable and the latency is low, then it is more inclined to use the result processed by the cloud side. If the second confidence value meets the preset instruction execution criteria, that is, it is detected that the network conditions are poor (such as high latency or unstable connection), then the result processed by the edge side is given priority to avoid processing delays or failures caused by network problems.

[0069] Thus, through a multi-level and multi-dimensional comprehensive evaluation method, the edge-cloud arbitration module can improve the accuracy and reliability of instruction processing as much as possible while ensuring efficiency, providing users with a more smooth and natural human-computer interaction experience.

[0070] According to an embodiment of the present application, after determining whether the cloud-side instruction inference result and the edge-side instruction inference result are received within a preset instruction cycle, it further includes: if the cloud-side instruction inference result and the edge-side instruction inference result are not received within the preset instruction cycle, an instruction response failure prompt is generated.

[0071] Specifically, if within the preset instruction cycle, the edge-cloud arbitration module does not receive the cloud-side instruction inference result and the edge-side instruction inference result, that is, the instruction is empty, it is considered that the instruction processing in the current instruction cycle fails. To avoid the situation where the user waits for a long time without knowing, when the system detects the above situation, it will actively provide the user with an instruction response failure prompt, for example, a prompt related to "Sorry, no valid response was received. Please try to resend the instruction". By providing a clear feedback prompt, the user can know that the current interaction was not successful and has the opportunity to resend the instruction, thereby improving the usability and friendliness of the system.

[0072] Thus, based on the above implementation manner, the embodiments of the present application can achieve the following beneficial effects:

[0073] (1) Improve processing efficiency and response speed. By the edge-cloud collaborative work, the preliminary processing of text instructions is performed on the edge side, and the powerful computing power of the cloud is used to process complex instructions, which can significantly improve the processing speed; for simple instructions, the edge side can directly process and quickly respond, and for complex instructions, the cloud resources are used to obtain more accurate results.

[0074] (2) By caching high-confidence instructions, when the same or similar instructions are encountered subsequently, they can be directly and quickly processed on the edge side without being repeatedly sent to the cloud, further improving the response speed.

[0075] (3) Enhance system reliability. Dynamically determine the best processing location according to factors such as network conditions and instruction complexity, ensuring that reliable processing results can be provided even in the case of unstable network; by comparing the confidence values of the inference results at both ends to select the optimal solution, the accuracy of decision-making is improved, and the possibility of misjudgment is reduced.

[0076] According to the interaction instruction processing method of the embodiments of the present application, when the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle, the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result are respectively identified. When the first confidence value is greater than the second confidence value, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. Otherwise, the vehicle is controlled to perform corresponding operations based on the end-side instruction inference result corresponding to the second confidence value. Thus, the problems of high latency and serious network dependence caused by simply relying on the computing resources of the cloud when using a large model for text inference are solved. The text instruction information is inferred by the cooperation of the cloud side and the end side, and the instruction inference results on both sides are identified by the end-cloud arbitration module, and the instruction with a higher confidence value is executed.

[0077] Next, an interaction instruction processing device according to an embodiment of the present application will be described with reference to the accompanying drawings.

[0078] Figure 6 It is a block diagram of an interaction instruction processing device according to an embodiment of the present application.

[0079] As Figure 6 shown, the interaction instruction processing device 10 includes: a judgment module 100, an identification module 200, and a control module 300.

[0080] Among them, the judgment module 100 is used to judge whether the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle;

[0081] The identification module 200 is used to, if the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle, respectively identify the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result;

[0082] The control module 300 is used to judge whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, the vehicle is controlled to perform corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, the vehicle is controlled to perform corresponding operations based on the end-side instruction inference result corresponding to the second confidence value.

[0083] According to an embodiment of the present application, before judging whether the cloud-side instruction inference result and the end-side instruction inference result are received within a preset instruction cycle, the judgment module 100 is further used for:

[0084] Judging whether the voice instruction information of the user is received;

[0085] If a voice command message is received, the voice command message is processed for voice data, and the voice command message after voice data processing is converted into a corresponding text command message;

[0086] The text command message is sent to a cloud-side command inference component and an edge-side command inference component respectively. The cloud-side command inference component performs cloud-side inference on the text command message to obtain a cloud-side command inference result, and the edge-side command inference component performs edge-side inference on the text command message to obtain an edge-side command inference result.

[0087] According to an embodiment of the present application, after determining whether the cloud-side command inference result and the edge-side command inference result are received within a preset command cycle, the determination module 100 is further configured to:

[0088] If the cloud-side command inference result and the edge-side command inference result are not received within the preset command cycle, an instruction response failure prompt is generated.

[0089] According to an embodiment of the present application, after determining whether a first confidence value is greater than a second confidence value, the control module 300 is further configured to:

[0090] If the first confidence value is equal to the second confidence value, the first confidence value and the second confidence value are identified based on a preset command execution standard. When the first confidence value meets the preset command execution standard, the vehicle is controlled to perform corresponding operations based on the cloud-side command inference result corresponding to the first confidence value, or when the second confidence value meets the preset command execution standard, the vehicle is controlled to perform corresponding operations based on the edge-side command inference result corresponding to the second confidence value.

[0091] According to an embodiment of the present application, after determining whether a first confidence value is greater than a second confidence value, the control module 300 is further configured to:

[0092] If the first confidence value is greater than the second confidence value, the first confidence value is stored in the target data set, otherwise, the second confidence value is stored in the target data set.

[0093] According to the interaction instruction processing device of the embodiments of the present application, when receiving the cloud-side instruction inference result and the end-side instruction inference result within a preset instruction cycle, it respectively identifies the first confidence value of the cloud-side instruction inference result and the second confidence value of the end-side instruction inference result, and when the first confidence value is greater than the second confidence value, controls the vehicle to execute corresponding operations based on the cloud-side instruction inference result corresponding to the first confidence value, otherwise, controls the vehicle to execute corresponding operations based on the end-side instruction inference result corresponding to the second confidence value. Thereby, it solves the problems of high latency and serious network dependence caused by simply relying on the computing resources of the cloud when using a large model for text inference. By jointly reasoning on text instruction information by the cloud side and the end side, and the end-cloud arbitration module identifies the instruction inference results on both sides and executes the instruction with a higher confidence value.

[0094] Figure 7 The following is a schematic structural diagram of a vehicle provided by the embodiments of the present application. The vehicle may include:

[0095] A memory 701, a processor 702, and a computer program stored on the memory 701 and executable on the processor 702.

[0096] When the processor 702 executes the program, it implements the interaction instruction processing method provided in the above embodiments.

[0097] Furthermore, the vehicle further includes:

[0098] A communication interface 703, used for communication between the memory 701 and the processor 702.

[0099] The memory 701 is used to store a computer program executable on the processor 702.

[0100] The memory 701 may include a high-speed RAM memory, and may also include non-volatile memory, such as at least one disk memory.

[0101] If the memory 701, the processor 702, and the communication interface 703 are implemented independently, the communication interface 703, the memory 701, and the processor 702 can be interconnected through a bus and complete communication with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation,Figure 7 It is represented by only one thick line, but it does not mean that there is only one bus or one type of bus.

[0102] Optionally, in a specific implementation, if the memory 701, the processor 702, and the communication interface 703 are integrated on a single chip, the memory 701, the processor 702, and the communication interface 703 can communicate with each other through an internal interface.

[0103] The processor 702 may be a central processing unit (CPU for short), or an application specific integrated circuit (ASIC for short), or one or more integrated circuits configured to implement the embodiments of the present application.

[0104] This embodiment also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned interactive instruction processing method is implemented.

[0105] This embodiment also provides a computer program product, including a computer program, and the computer program is executed to implement the interactive instruction processing method of the above embodiment.

[0106] In the description of this specification, the descriptions with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without conflict, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0107] In addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0108] Any process or method description represented in a flowchart or otherwise described herein can be understood to represent a module, segment, or portion of code including one or more executable instructions for implementing custom logical functions or processes. The scope of the preferred embodiments of this application includes additional implementations where functions may be executed not in the order shown or discussed, including in a substantially simultaneous manner according to the functions involved or in a reverse order, which should be understood by those skilled in the art to which the embodiments of this application pertain.

[0109] The logic and / or steps represented in a flowchart or otherwise described herein, for example, can be considered a sequenced list of executable instructions for implementing logical functions and can be embodied specifically in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection portion (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which a program can be printed, as the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then storing it in a computer memory.

[0110] It should be understood that the various parts of this application can be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0111] Those of ordinary skill in the art can understand that all or part of the steps carried out in implementing the above-described embodiment methods can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0112] In addition, in each of the embodiments of the present application, each functional unit can be integrated in a processing module, or each unit can exist physically alone, or two or more units can be integrated in a module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0113] The above-mentioned storage medium can be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for processing interactive instructions, characterized in that: The following steps are involved: Determine whether the cloud-side command inference result and the terminal-side command inference result are received within a preset command cycle; If the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within the preset instruction cycle, a first confidence value of the cloud-side instruction reasoning result and a second confidence value of the terminal-side instruction reasoning result are respectively identified; Determine whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, control the vehicle to perform a corresponding operation based on the cloud-side instruction inference result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, control the vehicle to perform a corresponding operation based on the terminal-side instruction inference result corresponding to the second confidence value.

2. The method according to claim 1, characterized in that Before determining whether the cloud-side instruction inference result and the terminal-side instruction inference result are received within a preset instruction cycle, the method further includes: Determine whether the user's voice command information is received; If the voice command information is received, performing voice data processing on the voice command information, and converting the voice command information after the voice data processing into corresponding text command information; The text instruction information is sent to a cloud-side instruction reasoning component and a terminal-side instruction reasoning component respectively. The cloud-side instruction reasoning component performs cloud-side reasoning on the text instruction information to obtain a cloud-side instruction reasoning result, and the terminal-side instruction reasoning component performs terminal-side reasoning on the text instruction information to obtain a terminal-side instruction reasoning result.

3. The method according to claim 1, characterized in that After determining whether the cloud-side instruction inference result and the terminal-side instruction inference result are received within a preset instruction cycle, the method further includes: If the cloud-side instruction inference result and the terminal-side instruction inference result are not received within the preset instruction cycle, a command response failure prompt is generated.

4. The method according to claim 1, characterized in that: After determining whether the first confidence value is greater than the second confidence value, the method further includes: If the first confidence value is equal to the second confidence value, the first confidence value and the second confidence value are identified based on a preset instruction execution standard, and when the first confidence value meets the preset instruction execution standard, the vehicle is controlled to perform a corresponding operation based on the cloud-side instruction reasoning result corresponding to the first confidence value, or, when the second confidence value meets the preset instruction execution standard, the vehicle is controlled to perform a corresponding operation based on the terminal-side instruction reasoning result corresponding to the second confidence value.

5. The method according to claim 4, characterized in that After determining whether the first confidence value is greater than the second confidence value, the method further includes: If the first confidence value is greater than the second confidence value, the first confidence value is stored in the target data set; otherwise, the second confidence value is stored in the target data set.

6. An interactive instruction processing device, characterized in that: include: A judgment module, used to judge whether the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within a preset instruction cycle; an identification module, configured to respectively identify a first confidence value of the cloud-side instruction reasoning result and a second confidence value of the terminal-side instruction reasoning result if the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within the preset instruction cycle; A control module is used to determine whether the first confidence value is greater than the second confidence value. If the first confidence value is greater than the second confidence value, the vehicle is controlled to perform a corresponding operation based on the cloud-side instruction reasoning result corresponding to the first confidence value. If the first confidence value is less than the second confidence value, the vehicle is controlled to perform a corresponding operation based on the terminal-side instruction reasoning result corresponding to the second confidence value.

7. The device according to claim 6, characterized in that Before determining whether the cloud-side instruction inference result and the terminal-side instruction inference result are received within a preset instruction cycle, the determination module is further configured to: Determine whether the user's voice command information is received; If the voice command information is received, performing voice data processing on the voice command information, and converting the voice command information after the voice data processing into corresponding text command information; The text instruction information is sent to a cloud-side instruction reasoning component and a terminal-side instruction reasoning component respectively. The cloud-side instruction reasoning component performs cloud-side reasoning on the text instruction information to obtain a cloud-side instruction reasoning result, and the terminal-side instruction reasoning component performs terminal-side reasoning on the text instruction information to obtain a terminal-side instruction reasoning result.

8. The device according to claim 6, characterized in that After determining whether the cloud-side instruction reasoning result and the terminal-side instruction reasoning result are received within a preset instruction cycle, the determination module is further configured to: If the cloud-side instruction inference result and the terminal-side instruction inference result are not received within the preset instruction cycle, a command response failure prompt is generated.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the interactive instruction processing method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the interactive instruction processing method as described in any one of claims 1 to 5.

Images