Robot interaction control method, server, client and robot

Abstract

The invention discloses a robot interaction control method, a server, a client and a robot, and belongs to the technical field of artificial intelligence. The method comprises the following steps: a client receives user behavior information and sends the user behavior information to a server; the server recognizes an interactive scene type according to the user behavior information, prompt information matched with the scene type and the user behavior information are input into a preset language processing model, task information output by the language processing model is obtained, and a control instruction of the robot is generated based on the task information; the control instruction is sent to the robot; the prompt information is used for guiding the language processing model to generate expected output; the robot executes the task according to the control instruction. According to the embodiment of the invention, which scene type the interaction belongs to can be identified, so that the language processing model can generate expected output under the scene type according to the user behavior information, interaction of the robot under different scenes can be realized, and the intelligent effect of the robot is improved.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and in particular relates to a robot interaction control method, server, client and robot. Background Art

[0002] Embodied intelligence refers to the ability of an intelligent system or machine to interact with its environment in real time through perception and interaction. Embodied intelligent systems typically possess the capabilities of perception, cognition, decision-making, and action. They can interact with their environment through sensors and actuators, making decisions and taking actions based on environmental changes.

[0003] Research on embodied intelligence currently focuses on embodied navigation, question-answering, and a variety of object manipulation tasks, including object rearrangement. These tasks overlap with existing disciplines, but also have distinct emphases. Navigation within the embodied intelligence paradigm focuses on achieving navigational goals through interaction, including point targets, object targets, command targets, and voice navigation. This requires the agent to actively explore the physical environment through vision, hearing, and language comprehension. Embodied question-answering is an upgrade to navigation, focusing on exploring and understanding the surrounding environment through interaction, associating language commands with answering specific questions. Embodied rearrangement involves the agent transferring objects in the physical environment from an initial configuration to a target configuration, typically in home environments. This type of task deemphasizes underlying robotics technologies, such as contact interaction and control between the robot and objects, but rather focuses on scene understanding, overall object state perception, and task planning. Robotic object manipulation is a key research area in robotics. Robotic manipulation within the embodied intelligence perspective focuses on learning how to interact and understand, control, and change external states through contact interaction, achieving task transferability, environmental adaptability, and skill scalability.

[0004] It can be seen that each of the above tasks has certain work being done. However, in these works, the models generally only handle a single task, are not unified enough, and lack a model that can handle multiple tasks. Summary of the Invention

[0005] This application aims to solve at least one of the technical problems existing in the prior art. To this end, this application proposes a robot interaction control method, server, client, and robot to enable robot interaction in different scenarios and improve the intelligent effect of the robot.

[0006] In a first aspect, the present application provides a robot interactive control method, characterized by comprising:

[0007] The client receives the user behavior information and sends the user behavior information to the server;

[0008] The server identifies the interaction scenario type based on the user behavior information, inputs prompt information matching the scenario type and the user behavior information into a preset language processing model, obtains task information output by the language processing model, generates control instructions for the robot based on the task information, and sends the control instructions to the robot; the prompt information is used to guide the language processing model to produce the desired output;

[0009] The robot performs tasks according to the control instructions.

[0010] According to the robot interaction control method of the present application, user behavior information is received through the client, and the user behavior information is sent to the server; the server identifies the scene type of the interaction based on the user behavior information, inputs the prompt information matching the scene type and the user behavior information into a preset language processing model, obtains the task information output by the language processing model, generates the robot's control instructions based on the task information; sends the control instructions to the robot; the prompt information is used to guide the language processing model to produce the desired output; the robot performs the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0011] According to one embodiment of the present application, inputting the prompt information matching the scenario type and the user behavior information into a preset language processing model to obtain task information output by the language processing model includes:

[0012] Inputting prompt information matching the scenario type and the user behavior information into a first language processing model to obtain a plurality of subtasks output by the first language processing model; the plurality of subtasks are obtained by decomposing the tasks matching the user behavior information;

[0013] The multiple subtasks are sent to a second language processing model to obtain multiple task information planned by the second language processing model based on the multiple subtasks.

[0014] In this embodiment, the task obtained by analyzing the user behavior information through the first language model is decomposed, which is equivalent to a pre-planned brain, breaking down the task into multi-step subtasks, and then planning the subtasks through the second language processing model. By executing the task in this summarize-then-divide-and-conquer manner, the task can be fully summarized and decomposed, and each subtask can be executed step by step, which can improve the success rate of task completion.

[0015] According to one embodiment of the present application, the multiple subtasks are executed sequentially, and the sending of the multiple subtasks to the second language processing model to obtain multiple task information planned by the second language processing model based on the multiple subtasks includes:

[0016] Sending the first subtask to a second language processing model, obtaining first task information planned by the second language processing model based on the first subtask, generating a first control instruction for the robot based on the first task information; and sending the first control instruction to the robot;

[0017] Upon receiving feedback information that the robot successfully performs the task according to the first control instruction, the second subtask is sent to the second language processing model, the second task information planned by the second language processing model based on the second subtask is obtained, and the second control instruction of the robot is generated based on the second task information; the second control instruction is sent to the robot; wherein, the execution order of the first subtask takes precedence over the execution order of the second subtask.

[0018] In this embodiment, by executing multiple subtasks step by step in sequence, and planning the next subtask through the second language processing model after the previous subtask is completed, it is beneficial to control the execution process of each task and further improve the efficiency of task completion.

[0019] According to one embodiment of the present application, the client provides an information exchange page, and the client receives user behavior information based on the information exchange page.

[0020] In this embodiment, the client can provide an information exchange page so that the user can communicate with the robot on the information exchange page, and the client can obtain the user's relevant information, so that the user's ideas, needs, etc. can be understood based on the user's relevant information during the interaction process, thereby improving the interactive experience.

[0021] According to one embodiment of the present application, the method further includes:

[0022] The server sends the first visual signal and the third visual signal of the robot to the client;

[0023] The client visually displays the first vision and the third vision of the robot through the information exchange page.

[0024] In this embodiment, the client can display the robot's first visual information and second visual information to the user based on the signal sent by the server. During the interaction process, it helps the user understand the robot's status, environment, etc., provides feedback for the interaction, and further improves the interaction experience.

[0025] According to one embodiment of the present application, the scenario type includes at least one of an instruction compliance scenario, an observation and help scenario, and an interactive question and answer scenario.

[0026] In a second aspect, the present application provides a robot interaction control method, which is applied to a client, comprising:

[0027] Receive user behavior information;

[0028] The user behavior information is sent to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information matching the scenario type and the user behavior information into a preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions.

[0029] According to the robot interaction control method of the present application, the client receives user behavior information; sends the user behavior information to the server so that the server can identify the scene type of the interaction based on the user behavior information, input the prompt information matching the scene type and the user behavior information into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0030] In a third aspect, the present application provides a robot interaction control method, which is applied to a server and includes:

[0031] Identify interaction scenario types based on user behavior information;

[0032] Inputting prompt information matching the scenario type and the user behavior information into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to generate a desired output;

[0033] generating a control instruction for the robot based on the task information;

[0034] The control instruction is sent to the robot so that the robot performs the task according to the control instruction.

[0035] According to the robot interaction control method of the present application, the server can identify the scene type of the interaction based on the user behavior information; input the prompt information matching the scene type and the user behavior information into the preset language processing model to obtain the task information output by the language processing model; the prompt information is used to guide the language processing model to produce the desired output; generate the robot's control instructions based on the task information; send the control instructions to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to based on the user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scene type based on the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0036] In a fourth aspect, the present application provides a robot interaction control method, applied to a robot, comprising:

[0037] Receiving a control instruction sent by a server; wherein the control instruction is generated based on task information, the task information is generated by the server identifying the interaction scenario type based on user behavior information, inputting prompt information matching the scenario type and the user behavior information into a preset language processing model, and obtaining an output based on the language processing model;

[0038] Execute tasks according to the control instructions.

[0039] According to the robot interaction control method of the present application, a control instruction is received from a server; wherein, the control instruction is generated based on task information, the task information is used by the server to identify the scene type of the interaction based on user behavior information, and the prompt information matching the scene type and the user behavior information are input into a preset language processing model, and the output is obtained based on the language processing model; the task is executed according to the control instruction. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0040] In a fifth aspect, the present application provides a client, including:

[0041] A first receiving module is used to receive user behavior information;

[0042] The first sending module is used to send the user behavior information to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information matching the scenario type and the user behavior information into a preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions; the prompt information is used to guide the language processing model to produce the desired output.

[0043] According to the client of the present application, by receiving user behavior information, the user behavior information is sent to the server so that the server can identify the scene type of the interaction based on the user behavior information, input the prompt information matching the scene type and the user behavior information into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of robots in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0044] In a sixth aspect, the present application provides a server, including:

[0045] Identification module, used to identify the interaction scenario type based on user behavior information;

[0046] An input module, configured to input prompt information matching the scenario type and the user behavior information into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to generate a desired output;

[0047] A generating module, configured to generate a control instruction for the robot based on the task information;

[0048] The second sending module is used to send the control instruction to the robot so that the robot performs the task according to the control instruction.

[0049] According to the server of the present application, the scenario type of the interaction is identified based on the user behavior information; prompt information matching the scenario type and the user behavior information are input into a preset language processing model to obtain the task information output by the language processing model; the prompt information is used to guide the language processing model to produce the desired output; a control instruction for the robot is generated based on the task information; and the control instruction is sent to the robot so that the robot performs the task according to the control instruction. The embodiment of the present application can identify which scenario type the interaction belongs to based on the user behavior information, and then extract the prompt information matching the scenario type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scenario type according to the user behavior information. In this way, the interaction of the robot in different scenarios can be realized, and the intelligent effect of the robot can be improved.

[0050] In a seventh aspect, the present application provides a robot, comprising:

[0051] a second receiving module, configured to receive a control instruction sent by a server; wherein the control instruction is generated based on task information, the task information is generated by the server identifying the interaction scenario type based on user behavior information, inputting prompt information matching the scenario type and the user behavior information into a preset language processing model, and obtaining an output based on the language processing model;

[0052] An execution module is used to execute tasks according to the control instructions.

[0053] According to the robot of the present application, the control instruction is received from the server; wherein, the control instruction is generated based on the task information, the task information is used by the server to identify the scene type of the interaction based on the user behavior information, and the prompt information matching the scene type and the user behavior information are input into the preset language processing model, and the output is obtained based on the language processing model; the task is executed according to the control instruction. The embodiment of the present application can identify which scene type the current interaction belongs to through the user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0054] In an eighth aspect, the present application provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the robot interaction control method as described in any one of the first to fourth aspects above is implemented.

[0055] In a ninth aspect, the present application provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the robot interaction control method as described in any one of the first to fourth aspects above.

[0056] In the tenth aspect, the present application provides a chip, which includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the robot interaction control method as described in any one of the first to fourth aspects above.

[0057] In an eleventh aspect, the present application provides a computer program product, comprising a computer program, which, when executed by a processor, implements the robot interaction control method as described in any one of the first to fourth aspects above.

[0058] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become obvious from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments with reference to the accompanying drawings, in which:

[0060] Figure 1 This is an architectural diagram of a robot interactive control system provided by an embodiment of the present application;

[0061] Figure 2This is a schematic diagram of the client interface provided by an embodiment of the present application;

[0062] Figure 3 This is one of the flow charts of the robot interactive control method provided in the embodiment of the present application;

[0063] Figure 4 This is the second flow chart of the robot interactive control method provided in the embodiment of the present application;

[0064] Figure 5 This is the third flow chart of the robot interaction control method provided in the embodiment of the present application;

[0065] Figure 6 This is the fourth flow chart of the robot interaction control method provided in the embodiment of the present application;

[0066] Figure 7 This is a schematic diagram of the structure of the client provided in the embodiment of the present application;

[0067] Figure 8 This is a schematic diagram of the structure of the server provided in the embodiment of the present application;

[0068] Figure 9 is a schematic structural diagram of a robot provided in an embodiment of the present application;

[0069] Figure 10 It is a structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0070] The following will be combined with the accompanying drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field are within the scope of protection of this application.

[0071] The terms "first," "second," and the like in the specification and claims of this application are used to distinguish similar objects, and are not used to describe a specific order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of this application can be implemented in an order other than that illustrated or described herein, and that the objects distinguished by "first," "second," and the like are generally of the same type, and do not limit the number of objects; for example, the first object can be one or more. In addition, the term "and / or" in the specification and claims refers to at least one of the connected objects, and the character " / " generally indicates that the objects connected are in an "or" relationship.

[0072] The following, in conjunction with the accompanying drawings, describes in detail the robot interaction control method, server, client, and robot provided in the embodiments of the present application through specific embodiments and their application scenarios.

[0073] First, the robot interactive control system provided by the embodiment of the present application is introduced. Figure 1 As shown, the robot interaction control system may include a client 100 , a server 200 and a robot 300 .

[0074] In the embodiment of the present application, the client 100 can be any terminal device that can access the network. The terminal device includes but is not limited to a portable communication device such as a mobile phone, a tablet computer, a laptop computer, etc. with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). It should also be understood that in some embodiments, the terminal device may not be a portable communication device, but a desktop computer with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). In some embodiments, the terminal device may also be a smart wearable device, such as a smart watch, a smart bracelet, smart glasses, or a smart head-mounted device.

[0075] In the embodiment of the present application, the robot 300 can be a single robot or multiple robots. If the robot 300 is multiple robots, the multiple robots can be robots of the same type or different types. The robot 300 can also be a virtual robot without a physical body or a robot with a physical body. The robot 300 can also have a driving device for driving the robot 300 to move. The robot 300 can also have a mechanical arm structure, and the mechanical arm is driven by a motor to complete the movement. Of course, the robot 300 can also be a robot of other forms, and the embodiment of the present application is not limited to this.

[0076] In an embodiment of the present application, the user can communicate with the robot 300 through the client 100. For example, the user can communicate with the robot 300 by inputting text or voice in the client 100, or issue instructions, and the robot 300 performs actions based on the user's input information.

[0077] Of course, robot 300 generally cannot directly recognize user input or directly perform task planning. After receiving user input, client 100 can send the user input to server 200. The server parses and calculates the user input, analyzes the user's intent, and converts it into instructions that robot 300 can recognize. The instructions are then sent to robot 300, allowing robot 300 to execute the task according to the instructions.

[0078] The workflow of the robot interactive control system of this application is introduced with a specific scenario example. Figure 2 As shown, the user can input from the interface provided by the client 100, and then these inputs will be sent to the robot 300 in the final environment to perform actions to complete the task. The user and the robot 300 can communicate through natural language on the interface, where Figure 2 The left side displays the first-person and third-person perspectives of robot 300, output from server 200. The right side shows a dialog box containing a user's conversation with robot 300. The user enters information into the dialog box to communicate with the agent. The user's input in the dialog box is transmitted to server 200 via client 100, and server 200 processes the input information based on the large language model.

[0079] Robot 300's action list can include actions like "walk to," "pick up," "place," "open," and "conversation." For example, "walk to the pool" means robot 300 moves to the pool; "pick up the basketball" means robot 300 picks up a basketball and holds it in its robotic hand; "place on the sofa" means placing the object held in its robotic hand on the sofa; "open the refrigerator" means the agent opens the refrigerator door; and "conversation: Do you know where the basketball is?" means robot 300 chooses to communicate with the user. In each round of communication, the client receives a description of the robot's embodied perspective of the environment from server 200 and displays it to the user in the interface.

[0080] The following describes the robot interaction control method provided by the embodiment of the present application. Figure 3 As shown, the robot interaction control method includes: step 310, step 320, step 330, step 340, step 350, step 360 and step 370.

[0081] Step 310: The client receives user behavior information.

[0082] In the embodiment of the present application, the client 100 can be any terminal device that can access the network. The terminal device includes but is not limited to a portable communication device such as a mobile phone, a tablet computer, a laptop computer, etc. with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). It should also be understood that in some embodiments, the terminal device may not be a portable communication device, but a desktop computer with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). In some embodiments, the terminal device may also be a smart wearable device, such as a smart watch, a smart bracelet, smart glasses, or a smart head-mounted device.

[0083] In the embodiments of the present application, user behavior information may be information related to user behavior, such as text information or voice information input by the user, or user gestures or motion information, etc. Of course, user behavior information may also be other information related to user behavior, which is not limited in the embodiments of the present application.

[0084] In some embodiments, the user can input text information, voice information, etc. to the client through a keyboard, touch, etc., and the client can receive the information input by the user. The client can also use a camera to capture user gestures, movements, and other information.

[0085] Step 320: Send the user behavior information to the server.

[0086] Step 330: The server identifies the interaction scenario type based on the user behavior information.

[0087] In an embodiment of the present application, the types of interaction scenarios between the user and the robot may include instruction compliance scenarios, observation and help scenarios, and interactive question and answer scenarios.

[0088] In a command-obeying scenario, the user sends a command to the robot, which then completes the command. For example, if the user issues the command "put the teddy bear into the pool," the robot will act based on the user's command and perform the task of "putting the teddy bear into the pool."

[0089] In the observation and assistance scenario, the robot, based on theory of mind, can autonomously reason and generate actions based on the current scene and user behavior information. Specifically, the robot can analyze observations of human behavior and speech, engage in dialogue with the human, align on consensus, and ultimately generate a plan. For example, scenario observation: You are in the living room 1 and see a basketball 1, a baseball bat 2, and a table 328. Character observation: A person has just returned from playing basketball and put the basketball on the ground. Character dialogue: You must be exhausted from playing basketball today. Robot: Playing basketball does require a lot of energy. Now that you've put the basketball down, is there anything I can help you with? Character dialogue: Yes, I really want to put it away, but I'm so tired. Robot: Oh, I see. You want me to put the basketball back in the basketball bag, right? Character dialogue: No, I usually put it on the basketball hoop in the living room. Robot: Okay, I understand. The robot's next task is to put the basketball back on the basketball hoop in the living room.

[0090] In interactive question-and-answer scenarios, users ask the robot a question about a specific scenario, and the agent explores and interacts with the scenario to answer the question. For example, a user might ask the robot how many knives are in a house. The robot will then search different rooms, potentially asking questions to the user. Finally, when the robot feels its search is complete, it will generate an answer.

[0091] Of course, the interaction scenario type can also be other scenarios, such as navigation scenarios, object manipulation scenarios, object rearrangement scenarios, etc. In different scenarios, the robot's logical emphasis is different. For example, in the navigation scenario, the focus is on completing the navigation goal through interaction; in the interactive question-and-answer scenario, the focus is on exploring and understanding the surrounding environment through interaction, and associating language instructions with answering specific questions. In the embodiment of the present application, the server can identify the interaction scenario based on the user's behavior information, thereby helping the robot to interact with the user in a targeted manner based on the user's behavior information.

[0092] The server can analyze the user behavior information. For example, if the user behavior information is voice or text information, the server can identify the meaning of the information. If the information is identified as a question sentence, the interaction scene type can be determined as an interactive question-and-answer scene. If the information is identified as an action instruction, the protection scene type can be determined as an instruction-obedience scene. If a character is observed to have performed a specific action and the information is identified as a declarative sentence, the scene type can be determined as an observation and help scene. Of course, the server can also determine the interaction scene type based on whether the user behavior information contains preset information. For example, if the user's behavior information contains "information 1", the interaction scene type can be determined as an interactive question-and-answer scene; if the user's behavior information contains "information 2", the interaction scene type can be determined as an instruction-obedience scene; if the user's behavior information contains "information 3", the interaction scene type can be determined as an observation and help scene.

[0093] Step 340: Input the prompt information and user behavior information that match the scenario type into a preset language processing model to obtain task information output by the language processing model.

[0094] In the embodiment of the present application, the language processing model may be a natural language processing model obtained through pre-training. The speech processing model may be a model such as ChatGPT, Wenxinyiyan, or BERT (Bidirectional Encoder Representations from Transformers).

[0095] In an embodiment of the present application, prompt information can be used to guide the language processing model to produce the desired output. The prompt information can be a text or an instruction. When using a language processing model, the prompt information can help the model understand the requirements of the task and generate a corresponding response. For example, in the field of natural language generation, if you want a language processing model to generate a text describing summer, such as for the input information "Please write a description about summer", the language processing model will generate a description about summer based on the input information. However, the range of descriptions about summer is very wide, and the language processing model may randomly generate a description about summer. If the description required by the user is described from the perspective of an elementary school student, you can set a prompt information "If you are an elementary school student, please write a description about summer", and then input this prompt information into the model. Based on this prompt, the model will generate a description about summer from the perspective of an elementary school student.

[0096] Based on the description of step 330, it can be seen that the robot's logical emphasis is different in different scenarios, so different prompt information can be set in advance for different scenario types so that the language processing model can output the expected information in different scenario types.

[0097] For example, in a command-obeying scenario, you might set a prompt message: You are an autonomous intelligent robot capable of finding and retrieving objects for humans. Your task is to provide assistance to humans. The following information will be provided: Character Dialogue: The human's response to your commands, assistance, or commands. Scene Observation: The room you are in and the items you currently see (for example, cabinet_328_link indicates a cabinet, numbered 328; white_cup_3_link indicates a white cup, numbered 3), as well as the visible rooms (for example, living_room_1, where living_room is the room type and 1 is its number). Action Feedback: Whether the action was successful. At each step, you can perform the following actions, and are limited to the following: Robot Thinking: Based on your previous suggestions and human instructions, propose a rational thought to plan your action. Use the following actions, and you must begin your action with "Robot Action:": trobot.pick("a"): Use this action to pick up item "a". trobot.place("a"): Use this action to place the item in your hand on item "a". trobot.chat(chat_content): Use this action to communicate with humans. Note: You should only issue one valid thought and action under the current observation.

[0098] For example, in the Observe and Help scenario, you can set a prompt message: You are a robot that understands "theory of mind." Your task is to analyze the human behavior and speech you observe, communicate with the human, align your consensus, and ultimately generate a plan: Your plan will help another robot act in the environment to assist the human. That robot can perform the following actions: robot.pick("a"): Use this action to pick up item "a". robot.place("a"): Use this action to place the item in its hand on item "a". robot.chat(chat_content): Use this action to communicate with the human. The following external information will be obtained: Character Observation: You will be informed of what the human is currently doing. Character Dialogue: The human's speech. Scene Observation: The room you are in and the items you currently see (for example, cabinet_328_link indicates that this is a cabinet with the number 328), as well as the visible rooms (for example, living_room_1, where living_room is the room type and 1 is its number). Note: You only need to generate the content of \">-ToM:\". The rest comes from the outside world. Generate it once and wait for the next round of information. Here is an example for you to learn: Scene Observation: You are in living room 1. You see a bedroom 2, a sofa_8_link, a black_baseballbat_4_link, a dresser_82_link, a cup_75_link, a diningroom 1, a teddybear_205_link, a bedroom 1, a brown_baseballbat_2_link, and a cabinet_328_link. Character Observation: A person just came back from playing basketball and put the basketball on the ground. Character Dialogue: Playing basketball today is really tiring. ToM: Playing basketball does require a lot of physical strength. Well, you put down the basketball. Is there anything I can help you with? Character Dialogue: Yes, I really want to put this basketball away, but I'm just too tired. ToM: Oh, I see. You want me to put the basketball back in the basketball bag, right? Character dialogue: No, I usually put the basketball on the basketball hoop in the living room. ToM: OK, I understand. The next task is to put the basketball on the basketball hoop in the living room. If you are ready to start a new task, please reply "Start".

[0099] Similarly, for interactive question-and-answer scenarios or other scenarios, matching prompt information can also be pre-set.

[0100] In an embodiment of the present application, task information is information output by a language processing model based on analysis of user behavior information and guidance of prompt information. The task information may include information related to completing the task, which may be the position and posture of the machine, the actions and instructions of the robot to complete the task, the position of the target object, and other information. For example, if the user behavior information is: pick up "basketball_1", the language processing model will output an instruction action, such as: robot.pick(basketball_1), which means that the instruction action is to pick up basketball_1, and the object corresponding to the instruction action. Then, the object information is obtained, and the target position and posture of the robot, as well as the type of action, are calculated. The code will use regular expressions to extract the action, which is "pick up", and the object ID, which is basketball_1. Then, the position and posture of basketball_1 are obtained, and the target position and posture of the robot's manipulator that grabs basketball_1 are calculated. For example, here it is to grab above the basketball, and the calculation is based on the posture of the basketball and a certain height is added to the z-axis.

[0101] Step 350: Generate control instructions for the robot based on the task information.

[0102] After the language processing model outputs the task information, if you want the robot to perform the task, you also need to plan the robot's path and posture. Based on the robot's position and posture in the task information, the position and posture of the target object, and other information, you need to calculate the path the robot moves to the target object and how the robot adjusts its posture to pick up the target object.

[0103] Based on the robot's current position, the server can use a path planning algorithm (such as the A* algorithm or Dijkstra algorithm) to calculate the optimal path for the robot to reach the target object. Path planning involves calculating the robot's posture as it moves along the path, including information such as position, velocity, and acceleration, to ensure it can successfully reach the target object. Based on the results of path planning and posture calculation, control instructions are generated. These control instructions are used to control the robot's driver or actuator, ensuring the robot moves along the planned path and maintains the correct posture.

[0104] Step 360: Send the control instruction to the robot.

[0105] Step 370: The robot performs the task according to the control instruction.

[0106] According to the robot interaction control method of the present application, user behavior information is received through the client and sent to the server; the server identifies the scene type of the interaction based on the user behavior information, inputs the prompt information and user behavior information that match the scene type into a preset language processing model, obtains the task information output by the language processing model, generates the robot's control instructions based on the task information; sends the control instructions to the robot; the prompt information is used to guide the language processing model to produce the desired output; the robot executes the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information that matches the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scene type according to the user behavior information. In this way, the robot can interact in different scenes and improve the intelligent effect of the robot.

[0107] In some embodiments, prompt information and user behavior information matching the scenario type are input into a preset language processing model to obtain task information output by the language processing model, including:

[0108] Inputting prompt information and user behavior information matching the scenario type into the first language processing model to obtain multiple subtasks output by the first language processing model; the multiple subtasks are obtained by decomposing the task based on the matching of the user behavior information;

[0109] The multiple subtasks are sent to the second language processing model to obtain multiple task information planned by the second language processing model based on the multiple subtasks.

[0110] Traditional task-based approaches typically execute a task in a sequential order: receive instruction -> action a -> action b -> action c. For example, for the task "put a teddy bear into a pool," the traditional approach is to follow the environment and the task description. If the teddy bear isn't in sight, the team moves around in various directions, searching for it, and proceed step by step.

[0111] In this embodiment, a pre-planned brain can break down tasks into multiple subtasks. For example, the task "Put a teddy bear in a pool" can be broken down into: 1. Find the teddy bear and walk to it; 2. Pick up the teddy bear; 3. Find the pool and walk to it; 4. Place the teddy bear in the pool. At each step, the description of the current subtask is incorporated. For example, if the robot is currently in the second step of the subtask, which is picking up the teddy bear, this gives it additional information: the current task is to pick up the teddy bear.

[0112] In this embodiment, different language processing models can be used to handle different requirements. For example, a first language processing model can be used to handle task decomposition requirements. In this case, the prompt information can also include text or sentences related to task decomposition to guide the first language processing model to output the multiple subtasks obtained by decomposition. A second language processing model can be used to handle task planning requirements. In this case, prompt information can also be provided for the second language processing model to guide the second language processing model to output subtask planning, thereby obtaining task information corresponding to the subtasks.

[0113] In this embodiment, the task obtained by analyzing the user behavior information through the first language model is decomposed, which is equivalent to a pre-planned brain, breaking down the task into multi-step subtasks, and then planning the subtasks through the second language processing model. By executing the task in this summarize-then-divide-and-conquer manner, the task can be fully summarized and decomposed, and each subtask can be executed step by step, which can improve the success rate of task completion.

[0114] In some embodiments, multiple subtasks are executed sequentially and sent to the second language processing model to obtain multiple task information planned by the second language processing model based on the multiple subtasks, including:

[0115] Sending the first subtask to the second language processing model, obtaining first task information planned by the second language processing model based on the first subtask, generating a first control instruction for the robot based on the first task information; and sending the first control instruction to the robot;

[0116] Upon receiving feedback information that the robot successfully performs the task according to the first control instruction, the second subtask is sent to the second language processing model, the second task information planned by the second language processing model based on the second subtask is obtained, and the second control instruction of the robot is generated based on the second task information; the second control instruction is sent to the robot; wherein, the execution order of the first subtask takes precedence over the execution order of the second subtask.

[0117] In this embodiment, the multiple subtasks obtained by decomposition can be executed in a sequential order. For example, the task "Put the teddy bear in the pool" can be decomposed into: 1. Find the teddy bear and walk to it; 2. Pick up the teddy bear; 3. Find the pool and walk to it; 4. Place the teddy bear in the pool. Executing steps 1, 2, 3, and 4 in sequence conforms to the logic of task completion.

[0118] Furthermore, the second language processing model can first plan a subtask to be executed. The robot then executes the subtask based on the control instructions corresponding to the task information output by the second language processing model. Upon completion of the subtask, the second language processing model will provide feedback indicating that the subtask is complete. After receiving this subtask completion information, the second language processing model will begin planning the next subtask, and so on. For example, after the robot finds a teddy bear and walks to it, it will output a "subtask completed" signal after completing this subtask. If this signal is received, it will extract the next subtask, which in this case is picking up the teddy bear, and input it into the second language processing model, allowing it to continue planning. This process continues in this manner. Eventually, when all subtasks are completed, the program will automatically end.

[0119] In this embodiment, by executing multiple subtasks step by step in sequence, and planning the next subtask through the second language processing model after the previous subtask is completed, it is beneficial to control the execution process of each task and further improve the efficiency of task completion.

[0120] In some embodiments, the client provides an information exchange page, and the client receives user behavior information based on the information exchange page.

[0121] Specifically, such as Figure 2 As shown, the client can provide an information exchange page, and the user can enter information in the dialog box of the information exchange page, and the client can receive the input information. Furthermore, the dialog box of the information exchange page can also display the information fed back by the robot.

[0122] In this embodiment, the client can provide an information exchange page so that the user can communicate with the robot on the information exchange page, and the client can obtain the user's relevant information, so that the user's ideas, needs, etc. can be understood based on the user's relevant information during the interaction process, thereby improving the interactive experience.

[0123] In some embodiments, the method further comprises:

[0124] The server sends the robot's first vision and third vision signals to the client;

[0125] The client visualizes the robot's first and third vision through the information exchange page.

[0126] Among them, the third vision can be a top-down vision, side vision or vision from other angles of the robot, which is not limited in the embodiments of the present application.

[0127] In this embodiment, the client can display the robot's first visual information and second visual information to the user based on the signal sent by the server. During the interaction process, it helps the user understand the robot's status, environment, etc., provides feedback for the interaction, and further improves the interaction experience.

[0128] The following describes the robot interactive control method provided by the embodiment of the present application by taking the client as an example of the execution subject. The client can be any terminal device that can access the network. The terminal device includes but is not limited to a portable communication device such as a mobile phone, a tablet computer, a laptop computer, etc. with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). It should also be understood that in some embodiments, the terminal device may not be a portable communication device, but a desktop computer with a touch-sensitive surface (e.g., a touch screen display and / or a touchpad). In some embodiments, the terminal device may also be a smart wearable device, such as a smart watch, a smart bracelet, smart glasses, or a smart head-mounted device.

[0129] like Figure 4 As shown, the robot interaction control method includes: step 410 and step 420.

[0130] Step 410: Receive user behavior information.

[0131] In embodiments of the present application, the client may provide a display interface, and the user may operate the display interface, for example, by touching a touchpad or touch screen, or by using a keyboard or mouse. The user may operate on the client's display interface, such as inputting text, voice, and other information. The client may recognize the user's operations on the interface and thereby receive user behavior information.

[0132] Step 420: Send the user behavior information to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information and user behavior information matching the scenario type into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions.

[0133] According to the robot interaction control method of the present application, the client receives user behavior information and sends the user behavior information to the server so that the server can identify the scene type of the interaction based on the user behavior information, input the prompt information matching the scene type and the user behavior information into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify the scene type to which the current interaction belongs through the user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0134] The following describes the robot interaction control method provided in the embodiment of the present application by taking the server as an example of the execution entity.

[0135] like Figure 5 As shown, the robot interaction control method includes: step 510, step 520, step 530 and step 540.

[0136] Step 510: Identify the interaction scenario type based on the user behavior information.

[0137] Step 520: Input the prompt information and user behavior information that match the scenario type into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to generate the desired output.

[0138] Step 530: Generate control instructions for the robot based on the task information.

[0139] Step 540: Send the control instruction to the robot so that the robot performs the task according to the control instruction.

[0140] According to the robot interaction control method of the present application, the server can identify the scene type of the interaction based on the user behavior information; input the prompt information and user behavior information that match the scene type into the preset language processing model to obtain the task information output by the language processing model; the prompt information is used to guide the language processing model to produce the desired output; generate the robot's control instructions based on the task information; and send the control instructions to the robot so that the robot can perform the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to based on the user behavior information, and then extract the prompt information that matches the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scene type based on the user behavior information. In this way, the robot can interact in different scenes and improve the intelligent effect of the robot.

[0141] The following describes the robot interaction control method provided in the embodiment of the present application by taking a robot as an example of the execution subject.

[0142] like Figure 6 As shown, the robot interaction control method includes: step 610 and step 620.

[0143] Step 610: Receive the control instruction sent by the server; wherein the control instruction is generated based on the task information, and the task information is used by the server to identify the interaction scenario type based on the user behavior information, and the prompt information and user behavior information matching the scenario type are input into the preset language processing model, and the output is obtained based on the language processing model.

[0144] Step 620: Execute the task according to the control instruction.

[0145] According to the robot interaction control method of the present application, a control instruction is received from a server; wherein the control instruction is generated based on task information, the task information is used by the server to identify the scene type of the interaction based on user behavior information, and the prompt information and user behavior information matching the scene type are input into a preset language processing model, and the output is obtained based on the language processing model; the task is executed according to the control instruction. The embodiment of the present application can identify which scene type the current interaction belongs to through user behavior information, and then extract the prompt information matching the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the robot can interact in different scenes and improve the intelligent effect of the robot.

[0146] The embodiment of the present application also provides a client.

[0147] like Figure 7 As shown, the client includes:

[0148] A first receiving module 710 is configured to receive user behavior information;

[0149] The first sending module 720 is used to send user behavior information to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information and user behavior information matching the scenario type into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions; the prompt information is used to guide the language processing model to produce the desired output.

[0150] According to the client of the present application, by receiving user behavior information, the user behavior information is sent to the server so that the server can identify the scene type of the interaction based on the user behavior information, input the prompt information and user behavior information that match the scene type into the preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify which scene type the current interaction belongs to through the user behavior information, and then extract the prompt information that matches the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the interaction of the robot in different scenes can be realized, and the intelligent effect of the robot can be improved.

[0151] The embodiment of the present application also provides a server.

[0152] like Figure 8 As shown, the server includes:

[0153] Identification module 810, for identifying the type of interaction scenario based on user behavior information;

[0154] Input module 820 is used to input prompt information and user behavior information matching the scenario type into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to produce the desired output;

[0155] A generating module 830 is used to generate control instructions for the robot based on the task information;

[0156] The second sending module 840 is used to send the control instruction to the robot so that the robot can perform the task according to the control instruction.

[0157] According to the server of the present application, the scenario type of the interaction is identified based on user behavior information; prompt information matching the scenario type and user behavior information are input into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to produce the desired output; control instructions for the robot are generated based on the task information; and the control instructions are sent to the robot so that the robot performs the task according to the control instructions. The embodiment of the present application can identify the scenario type of the current interaction based on user behavior information, and then extract prompt information matching the scenario type. Based on the prompt information, the language processing model can be guided so that the language processing model can produce the desired output under the scenario type based on the user behavior information. In this way, the interaction of the robot in different scenarios can be realized, and the intelligent effect of the robot can be improved.

[0158] An embodiment of the present application also provides a robot.

[0159] like Figure 9 As shown, the robot includes:

[0160] A second receiving module 910 is configured to receive a control instruction sent by the server; wherein the control instruction is generated based on task information. The server identifies the interaction scenario type based on user behavior information, inputs prompt information and user behavior information matching the scenario type into a preset language processing model, and outputs the result based on the language processing model;

[0161] The execution module 920 is used to execute tasks according to the control instructions.

[0162] According to the robot of the present application, the robot receives control instructions sent by the server; wherein the control instructions are generated based on task information, and the task information is used by the server to identify the scene type of the interaction based on the user behavior information, and the prompt information and user behavior information that match the scene type are input into the preset language processing model, and the output is obtained based on the language processing model; and the task is executed according to the control instruction. The embodiment of the present application can identify which scene type the current interaction belongs to through the user behavior information, and then extract the prompt information that matches the scene type. Based on the prompt information, the language processing model can be guided so that the language processing model can generate the expected output under the scene type according to the user behavior information. In this way, the robot can interact in different scenes and improve the intelligent effect of the robot.

[0163] In some embodiments, as Figure 10As shown, an embodiment of the present application further provides an electronic device 1000, including a processor 1001, a memory 1002, and a computer program stored in the memory 1002 and executable on the processor 1001. When the program is executed by the processor 1001, each process of the above-mentioned robot interaction control method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be described here.

[0164] It should be noted that the electronic devices in the embodiments of the present application include the above-mentioned mobile electronic devices and non-mobile electronic devices.

[0165] An embodiment of the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the above-mentioned robot interaction control method embodiment are implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0166] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0167] An embodiment of the present application also provides a computer program product, including a computer program, which implements the above-mentioned robot interaction control method when executed by a processor.

[0168] The processor is the processor in the electronic device in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory ROM, a random access memory RAM, a magnetic disk or an optical disk.

[0169] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, which is coupled to the processor. The processor is used to run programs or instructions to implement the various processes of the above-mentioned robot interaction control method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0170] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0171] It should be noted that, in this article, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the statement "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or device comprising the element. In addition, it should be noted that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in the opposite order according to the functions involved. For example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0172] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including a number of instructions for enabling a terminal (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0173] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of this application, ordinary technicians in this field can also make many forms without departing from the purpose of this application and the scope of protection of the claims, all of which are within the protection of this application.

[0174] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0175] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and intent of the present application, and that the scope of the present application is defined by the claims and their equivalents.

Claims

1. A robot interactive control method, characterized in that: include: The client receives the user behavior information and sends the user behavior information to the server; The server identifies the interaction scenario type based on the user behavior information, inputs prompt information matching the scenario type and the user behavior information into a preset language processing model, obtains task information output by the language processing model, generates control instructions for the robot based on the task information, and sends the control instructions to the robot; the prompt information is used to guide the language processing model to produce the desired output; The robot performs tasks according to the control instructions.

2. The method according to claim 1, characterized in that Inputting the prompt information matching the scenario type and the user behavior information into a preset language processing model to obtain task information output by the language processing model includes: Inputting prompt information matching the scenario type and the user behavior information into a first language processing model to obtain a plurality of subtasks output by the first language processing model; the plurality of subtasks are obtained by decomposing the tasks matching the user behavior information; The multiple subtasks are sent to a second language processing model to obtain multiple task information planned by the second language processing model based on the multiple subtasks.

3. The method according to claim 2, wherein the plurality of subtasks are executed sequentially, and wherein the sending of the plurality of subtasks to the second language processing model to obtain a plurality of task information planned by the second language processing model based on the plurality of subtasks comprises: Sending the first subtask to a second language processing model, obtaining first task information planned by the second language processing model based on the first subtask, and generating a first control instruction for the robot based on the first task information; sending the first control instruction to the robot; Upon receiving feedback information that the robot successfully performs the task according to the first control instruction, the second subtask is sent to the second language processing model, the second task information planned by the second language processing model based on the second subtask is obtained, and the second control instruction of the robot is generated based on the second task information; the second control instruction is sent to the robot; wherein, the execution order of the first subtask takes precedence over the execution order of the second subtask.

4. The method according to claim 1, wherein The client provides an information exchange page, and the client receives user behavior information based on the information exchange page.

5. The method according to claim 4, characterized in that The method further comprises: The server sends the first visual signal and the third visual signal of the robot to the client; The client visually displays the first vision and the third vision of the robot through the information exchange page.

6. The method according to claim 1, characterized in that The scenario type includes at least one of an instruction compliance scenario, an observation and help scenario, and an interactive question and answer scenario.

7. A robot interactive control method, characterized in that: Applied to the client, including: Receive user behavior information; The user behavior information is sent to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information matching the scenario type and the user behavior information into a preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions.

8. A robot interactive control method, characterized in that: Applicable to servers, including: Identify interaction scenario types based on user behavior information; Inputting prompt information matching the scenario type and the user behavior information into a preset language processing model to obtain task information output by the language processing model; the prompt information is used to guide the language processing model to generate a desired output; generating a control instruction for the robot based on the task information; The control instruction is sent to the robot so that the robot performs the task according to the control instruction.

9. A robot interactive control method, characterized in that: Applications in robots include: Receiving a control instruction sent by a server; wherein the control instruction is generated based on task information, the task information is generated by the server identifying the interaction scenario type based on user behavior information, inputting prompt information matching the scenario type and the user behavior information into a preset language processing model, and obtaining an output based on the language processing model; Execute tasks according to the control instructions.

10. A client, characterized in that: include: A first receiving module, configured to receive user behavior information; The first sending module is used to send the user behavior information to the server so that the server can identify the interaction scenario type based on the user behavior information, input the prompt information matching the scenario type and the user behavior information into a preset language processing model, obtain the task information output by the language processing model, generate the robot's control instructions based on the task information, and send the control instructions to the robot so that the robot performs the task according to the control instructions; the prompt information is used to guide the language processing model to produce the desired output.

11. A server, characterized in that: include: Identification module, used to identify the interaction scenario type based on user behavior information; An input module, configured to input prompt information matching the scenario type and the user behavior information into a preset language processing model, and obtain task information output by the language processing model; The prompt information is used to guide the language processing model to generate a desired output; A generating module, configured to generate a control instruction for the robot based on the task information; The second sending module is used to send the control instruction to the robot so that the robot performs the task according to the control instruction.

12. A robot, characterized in that: include: a second receiving module, configured to receive a control instruction sent by a server; wherein the control instruction is generated based on task information, the task information is generated by the server identifying the interaction scenario type based on user behavior information, inputting prompt information matching the scenario type and the user behavior information into a preset language processing model, and obtaining an output based on the language processing model; An execution module is used to execute tasks according to the control instructions.

Images