Control method, device, apparatus and medium of intelligent device

Abstract

This invention discloses a control method, apparatus, device, and medium for an intelligent device. The method includes: acquiring a dataset of the operating status of all intelligent devices within a target area; classifying all intelligent devices according to their operating status to obtain a first intelligent device set; calling a preset recognition model and configuring the preset recognition model on each intelligent device in the first intelligent device set to obtain a target device; controlling the target device to acquire user request information and acquiring a first recognition result of the target device in response to the user request information; and generating control instructions based on the first recognition result to control the intelligent device corresponding to the user request information. By using the intelligent device as the control center, the computational burden on a single control center is reduced, thus improving the control efficiency of the intelligent device.

Description

Technical Field

[0001] This invention relates to the field of smart home technology, and in particular to a control method, device, equipment and medium for a smart device. Background Technology

[0002] With the rapid development of automation, the Internet of Things, and other smart technologies, smart homes are becoming increasingly convenient for people's lives. However, in current technologies, the process of implementing smart homes often involves one or a fixed number (at least two) control centers performing data calculations to control networked smart devices. Each device receives user commands in a fixed manner. However, using one or a fixed number of control centers for data calculation often results in a large amount of computation, placing a heavy computational burden on a single control center and leading to low efficiency. Meanwhile, other electronic devices with data receiving and processing capabilities remain idle or under low computational load, resulting in resource waste. Using a fixed number (at least two) control centers for data calculation... Calculations show that controlling networked smart devices increases costs and leads to idle control centers when computational load is low. Furthermore, when a single control center performs data calculations to control networked smart devices, it often relies on a single model (algorithm) to identify user intentions, making it susceptible to interference and inaccurate identification. While using a fixed number (greater than or equal to two) of control centers to perform data calculations to control networked smart devices allows for the simultaneous identification of multiple models (algorithms), the differences in performance between existing intention recognition models are not significant when using a single device receiving user commands in a fixed manner. Therefore, using multiple models (algorithms) to identify user intentions is not very meaningful, and inaccurate identification still exists. Moreover, when using a single device to receive user commands in a fixed manner, if the receiving device is a non-portable electronic device, such as a smart speaker, it needs to be placed in a fixed location. Users need to move the device to a location near the smart speaker to control it, causing inconvenience.

[0003] Cooking is an indispensable activity for human survival and development. Throughout history, people have cooked once or more every day. The utensils, ingredients, seasonings, and procedures used in different dishes for a single meal vary greatly. If one person only follows the steps to complete one dish at a time, it often takes a long time to cook a meal. With the continuous progress of society, although there have been significant improvements in cooking utensils, saving some processing time, food preparation still requires manual step-by-step completion, which undoubtedly cannot meet the needs of today's fast-paced lifestyle. With the rapid development of automation, the Internet of Things, and other smart technologies, smart cooking and smart kitchens are becoming increasingly popular. However, when smart cooking and smart kitchen equipment are connected to a local area network to realize smart homes, the same problems arise. Summary of the Invention

[0004] In view of this, the present invention provides a control method, apparatus, device and medium for intelligent devices, which solves the problems of inaccurate control and low control efficiency of intelligent devices in the prior art.

[0005] To achieve one, some, or all of the above objectives, or other objectives, the present invention proposes a control method for an intelligent device, comprising:

[0006] Obtain the operating status dataset of all smart devices within the target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state and off state;

[0007] All smart devices are classified according to their operating status to obtain a first set of smart devices, a second set of smart devices, and a third set of smart devices.

[0008] A preset recognition model is called from the preset model library and configured on the smart devices in the first set of smart devices to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models.

[0009] Control the target device to acquire user request information, and acquire the first recognition result of the target device in response to the user request information;

[0010] Based on the first identification result, control instructions are generated to control the smart device corresponding to the user request information.

[0011] Optionally, the step of obtaining the operational status dataset of all smart devices within the target area includes:

[0012] Divide the user's space into at least two sub-areas according to the building structure;

[0013] Image data for each sub-region is acquired, and the image data is input into the human body recognition model to obtain a second recognition result;

[0014] Based on the second identification result, the sub-region where the user is located is determined as the target region;

[0015] The image data corresponding to the target area is used as the target image data, and the operating status dataset of all smart devices in the target area is obtained based on the target image data.

[0016] Optionally, the step of using image data corresponding to the target area as target image data and obtaining the operating status dataset of all smart devices within the target area based on the target image data includes:

[0017] The image data corresponding to the target area is used as the target image data, and the type information of each smart device displayed by the target image data is identified;

[0018] Based on the type information of each smart device, the power consumption data of each smart device, and the traffic data of each smart device, the operating status of each smart device is determined, and a dataset of the operating status of all smart devices in the target area is obtained.

[0019] Optionally, the step of calling a preset recognition model from a preset model library and configuring the preset recognition model on a smart device in the first set of smart devices to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition model, includes:

[0020] Obtain the metadata of all preset recognition models in the preset model library. The metadata includes metadata for each preset recognition model, and the metadata includes model feature mapping relationships.

[0021] Based on the model feature mapping relationship, obtain the feature identifier of the feature on which each preset recognition model depends;

[0022] Based on the feature identifier, the feature values ​​of the features that each preset recognition model depends on are obtained to obtain a feature value set;

[0023] Based on the feature set, a preset recognition model is retrieved from the preset model library, and the retrieved preset recognition model is configured on the smart devices in the first set of smart devices to obtain the target device.

[0024] Optionally, the step of retrieving a preset recognition model from a preset model library based on the feature value set, and configuring the retrieved preset recognition model in the smart devices of the first smart device set to obtain the target device includes:

[0025] Based on the feature value set, a preset recognition model is retrieved from the preset model library;

[0026] The number of preset recognition models invoked is counted and used as the first count;

[0027] The number of intelligent devices in the first intelligent device cluster is counted and used as the second number.

[0028] Compare the first number and the second number, and select the smaller number between the first number and the second number as the target number;

[0029] A target number of the preset recognition models are selected by random sampling, and a target number of smart devices to be configured are selected from the first set of smart devices.

[0030] The selected preset recognition model is configured on the selected smart device to be configured according to the preset rules to obtain the target device. The preset rules are configured on any preset recognition model and only on one smart device to be configured.

[0031] Optionally, the step of controlling the target device to obtain user request information and obtaining the recognition result of the target device for the user request information includes:

[0032] Obtain the human-computer interaction rules formulated for the target device, and control the target device to obtain user request information according to the human-computer interaction rules;

[0033] The first identification result of the target device in response to the user request information is obtained. The first identification result includes the smart device to be controlled corresponding to the user request information and the control instructions for the smart device to be controlled.

[0034] Optionally, the method further includes:

[0035] A preset recognition model is called from the preset model library, and the preset recognition model is configured on the smart devices in the second set of smart devices to obtain candidate devices;

[0036] Control the alternative device to acquire user request information, and acquire the third identification result of the alternative device in response to the user request information;

[0037] Based on the first identification result and the third identification result, control instructions are generated to control the smart device corresponding to the user request information.

[0038] On the other hand, embodiments of this application provide a control device for a smart device, the control device comprising:

[0039] The data acquisition module is used to acquire the operating status dataset of all smart devices within the target area. The smart devices are electronic devices with data transmission and reception functions and data calculation functions. The operating status dataset includes the operating status of each smart device, including an on state, a standby state, and a off state.

[0040] The classification module is used to classify all smart devices according to their operating status to obtain a first set of smart devices, a second set of smart devices, and a third set of smart devices.

[0041] The calling module is used to call a preset recognition model from the preset model library and configure the preset recognition model on the smart devices in the first set of smart devices to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models.

[0042] The identification module is used to control the target device to obtain user request information and to obtain a first identification result of the target device in response to the user request information;

[0043] The control module is used to generate control commands based on the first identification result to control the smart device corresponding to the user request information.

[0044] On the other hand, embodiments of this application provide an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, the steps of the control method of the intelligent device described above are performed.

[0045] On the other hand, embodiments of this application provide a computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, and the computer program is executed by a processor to perform the steps of the control method of the intelligent device described above.

[0046] Implementing the embodiments of the present invention will have the following beneficial effects:

[0047] By acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state, and off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from a preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information. Using smart devices as control centers reduces the computational burden on a single control center and improves the control efficiency of smart devices. When the first set of smart devices includes at least two devices, these devices can receive user commands (voice commands, gesture commands, text commands, etc.) in different ways and use different preset recognition models to identify the user's intent. This ensures the accuracy of user intent recognition. Furthermore, since different smart devices are located in different places, they can obtain user commands with different characteristics. Identifying these different user commands using different preset recognition models further ensures the accuracy of user intent recognition. Simultaneously, the location of the device receiving user commands is not fixed, allowing users to conveniently control any smart device in the smart home network from any area covered by the smart home, thus improving the user experience. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] in:

[0050] Figure 1 This is a flowchart of a control method for an intelligent device provided in an embodiment of this application;

[0051] Figure 2 This is a flowchart of another control method for an intelligent device provided in an embodiment of this application;

[0052] Figure 3This is a schematic diagram of the structure of a control device for an intelligent device provided in an embodiment of this application;

[0053] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application;

[0054] Figure 5 This is a schematic diagram of the structure of a computer-readable storage medium provided in an embodiment of this application. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] like Figure 1 As shown, this application provides a control method for an intelligent device, including:

[0057] S101. Obtain the operating status dataset of all smart devices within the target area, wherein the smart device is an electronic device with data transmission and reception function and data calculation function, and the operating status dataset includes the operating status of each smart device, including on state, standby state and off state;

[0058] For example, the user's location is taken as the target area, and the operating status dataset of all smart devices in the user's location is obtained. The smart devices are electronic devices with data transmission and reception functions and data calculation functions, such as smart TVs, which can receive display data (programs) sent by the server, send request data to the server, and parse instructions to adjust the display content. The operating status dataset includes the operating status of each smart device, including on state, standby state, and off state.

[0059] S102. Classify all smart devices according to their operating states to obtain a first set of smart devices, a second set of smart devices, and a third set of smart devices.

[0060] For example, all smart devices are classified according to their operating states to obtain a set of smart devices in the on state (first smart device set), a set of smart devices in the standby state (second smart device set), and a set of smart devices in the off state (third smart device set).

[0061] S103. Call a preset recognition model from the preset model library and configure the preset recognition model on the smart devices in the first smart device set to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition model.

[0062] For example, the preset model library is stored in the cloud to reduce the resource consumption of the local area network. A call request is sent to the cloud, and the preset recognition model is extracted from the preset model library according to the call request. The preset recognition model is configured in the smart device in the first smart device set (the device set of smart devices in the enabled state) to obtain the target device.

[0063] For example, the preset model library includes at least two preset recognition models with different algorithms, as well as preset recognition models with the same algorithm but different model training methods.

[0064] S104. Control the target device to obtain user request information, and obtain the first identification result of the target device for the user request information;

[0065] For example, when controlling the target device to obtain user request information, the method of controlling the target device to obtain user request information is also different because the types of information that the target device can receive are different. For example, when the target device is a monitoring device, the monitoring device is controlled to obtain the user's image (posture image, lip image) as user request information. When the target device is an audio device, the audio device is controlled to obtain the user's voice data as user request information.

[0066] For example, since different target devices acquire user request information in different data formats, different intent recognition models are needed to process the user request information. For instance, when acquiring a user's image (gesture image, lip-sync image) as user request information, the target device obtains the recognition result for the user request information through image recognition. When acquiring a user's voice data as user request information, the target device obtains the recognition result for the user request information through voice recognition.

[0067] S105. Generate control instructions based on the first identification result to control the smart device corresponding to the user request information.

[0068] For example, the first identification result (user intent) is "start cooking". Since the smart device associated with "cooking" is a smart cooking device, the smart cooking device is determined to be the smart device corresponding to the user request information. "Start" is used as control information to generate control commands. Here, the smart device as the target device is an electronic device with data transmission and reception functions and data calculation functions. That is, in this embodiment, the smart device as the target device needs to perform data acquisition (equivalent to a device receiving user commands) and data calculation (equivalent to a control center), while the smart device (the smart device corresponding to the user request information) is the controlled device. It does not need to directly receive user commands but indirectly receives user commands by receiving control commands. The smart device controls the smart device according to the user commands.

[0069] For example, when a user wants to control a smart cooking device to cook, the operating status of all smart devices near the user is obtained, resulting in an operating status set. For instance, if the user is in the living room and the smart devices nearby are a smart TV, a smart speaker, and a smart monitoring device, with the smart TV in an on state, the smart speaker in a standby state, and the smart monitoring device in an on state, then the operating status set is "on, standby, on". All smart devices are categorized according to their operating status, resulting in a first smart device set (smart TV, smart monitoring device), a second smart device set (smart speaker), and a third smart device set (empty). A preset recognition model is called from the preset model library and configured on the smart TV and the smart monitoring device. The configured smart TV and the configured smart monitoring device are then used as target devices, and the configured smart TV is controlled to obtain the user's request information. At this point, the user request information can be key input by the user via a button. The configured intelligent monitoring device acquires the user request information, which can also be the user's image information. A first intent recognition model, trained based on the key input and target intent, identifies the user's key input to obtain a first intent (control the intelligent cooking device to perform cooking). A second intent recognition model, trained based on the image information and target intent, identifies the user's image information to obtain a second intent (control the intelligent cooking device to perform cooking). The target intent generated based on the first and second intents is used as the first recognition result. A control command is generated based on the first recognition result to control the intelligent device corresponding to the user request information. The control command includes the intelligent equipment (intelligent cooking device) to be controlled and the instruction content (perform cooking) for the intelligent equipment (intelligent cooking device). Regardless of the user's location, any intelligent device can be controlled. That is, the user can control the intelligent cooking device from the living room. When the intelligent cooking device and the user are not in the same area, cross-area control is possible.

[0070] For example, when at least two target devices are identified, if the target devices are located in different local area networks, the first intent or the second intent is transmitted through the data channel between the different local area networks to realize the process of generating a target intent based on the first intent and the second intent.

[0071] By acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state, and off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from a preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information. Using smart devices as control centers reduces the computational burden on a single control center and improves the control efficiency of smart devices. When the first set of smart devices includes at least two devices, these devices can receive user commands (voice commands, gesture commands, text commands, etc.) in different ways and use different preset recognition models to identify the user's intent. This ensures the accuracy of user intent recognition. Furthermore, since different smart devices are located in different places, they can obtain user commands with different characteristics. Identifying these different user commands using different preset recognition models further ensures the accuracy of user intent recognition. Simultaneously, the location of the device receiving user commands is not fixed, allowing users to conveniently control any smart device in the smart home network from any area covered by the smart home, thus improving the user experience.

[0072] In one possible implementation, the step of acquiring the operational status dataset of all smart devices within the target area includes:

[0073] Divide the user's space into at least two sub-areas according to the building structure;

[0074] Image data for each sub-region is acquired, and the image data is input into the human body recognition model to obtain a second recognition result;

[0075] Based on the second identification result, the sub-region where the user is located is determined as the target region;

[0076] The image data corresponding to the target area is used as the target image data, and the operating status dataset of all smart devices in the target area is obtained based on the target image data.

[0077] For example, the process of dividing a user's space into at least two sub-areas according to the building structure, such as dividing the user's house (the user's space) into sub-spaces such as living room, bedroom, and kitchen according to the building structure.

[0078] For example, image data of each sub-region can be acquired by a camera device and input into a human body recognition model to obtain a second recognition result; alternatively, infrared image data of each sub-region can be acquired by an infrared sensor device and input into a human body recognition model to obtain a second recognition result.

[0079] For example, the construction process of the human body recognition model is as follows: construct a feature interaction network, a local feature enhancement network, and a local feature optimization network, with a residual neural network as the backbone network, followed by the feature interaction network, the local feature enhancement network, and the local feature optimization network in sequence, to obtain the human body recognition model;

[0080] The training process of the human body recognition model is as follows: A human body training dataset is acquired, and training samples from the dataset are input into the human body recognition model; multiple stage features are output through the multiple stage networks of the residual neural network; the multiple stage features are fused using the feature interaction network to obtain fused features, and the multiple stage features and the fused features are interacted to obtain total interaction features; the total interaction features are processed using the local feature enhancement network to obtain sample features and multiple local features; based on the sample features and multiple local features, the human body recognition model is trained using a loss function.

[0081] For example, the construction process of the human body recognition model is as follows: construct a feature interaction network, a local feature enhancement network, and a local feature optimization network, with a residual neural network as the backbone network, followed by the feature interaction network, the local feature enhancement network, and the local feature optimization network in sequence, to obtain the human body recognition model;

[0082] In one possible implementation, the step of using image data corresponding to the target region as target image data and obtaining a dataset of the operating status of all smart devices within the target region based on the target image data includes:

[0083] The image data corresponding to the target area is used as the target image data, and the type information of each smart device displayed by the target image data is identified;

[0084] Based on the type information of each smart device, the power consumption data of each smart device, and the traffic data of each smart device, the operating status of each smart device is determined, and a dataset of the operating status of all smart devices in the target area is obtained.

[0085] For example, the image data corresponding to the target area is used as the target image data, and the target image data is input into the device recognition model to obtain all the smart devices existing in the target area, and then the operating status dataset of all smart devices in the target area is obtained. The model algorithm and structure of the device recognition model are the same as those of the human body recognition model, but the device recognition model is trained by device features and device names.

[0086] For example, the image data corresponding to the target area is used as the target image data, and the target image data is input into the device recognition model to obtain all the smart devices existing in the target area, and then the operating status dataset of all smart devices in the target area is obtained. The model algorithm and structure of the device recognition model are the same as those of the human body recognition model, but the device recognition model is trained by device features and device names.

[0087] For example, power consumption data of each smart device is acquired through a non-invasive measurement device, and traffic data of each smart device is acquired through a network traffic statistics method. Then, based on the type information of the smart device, the power consumption characteristics of the smart device under different operating states (on, standby, and off) are determined. The operating state of the smart device is determined through the power consumption data and power consumption characteristics. The reasonableness of the determined operating state is verified through the traffic data of the smart device. For example, the operating state of the smart TV is determined to be on based on the power consumption data and power consumption characteristics, but the traffic data of the smart TV is zero, which is obviously unreasonable and the operating state of the smart TV needs to be re-evaluated.

[0088] In one possible implementation, the step of calling a preset recognition model from a preset model library and configuring the preset recognition model on a smart device in the first set of smart devices to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models, includes:

[0089] Obtain the metadata of all preset recognition models in the preset model library. The metadata includes the metadata of each preset recognition model, and the metadata includes the model feature mapping relationship.

[0090] Based on the model feature mapping relationship, obtain the feature identifier of the feature on which each preset recognition model depends;

[0091] Based on the feature identifier, the feature values ​​of the features that each preset recognition model depends on are obtained to obtain a feature value set;

[0092] Based on the feature set, a preset recognition model is retrieved from the preset model library, and the retrieved preset recognition model is configured on the smart devices in the first set of smart devices to obtain the target device.

[0093] For example, such as Figure 2 As shown, the invocation process includes: in response to a model invocation request, obtaining the model's metadata, which includes model feature mapping relationships; specifically, in response to a model invocation request, obtaining the metadata of the model invoked in the request. More specifically, the model invocation request contains the model identifier of the invoked model; based on this identifier, obtaining the corresponding model's metadata, which includes model feature mapping relationships, is used to identify the features required by the current version of the model, i.e., the features the model depends on. The model identifier here can be the model's name or model number, and can be used to uniquely identify the model.

[0094] Based on the model feature mapping relationship, obtain the feature identifiers of the features that the model depends on; specifically, obtain the feature mapping relationship of the model and obtain the feature identifiers of the features that the model depends on.

[0095] Based on the feature identifier, the feature values ​​of the features the model depends on are obtained. Specifically, based on the feature identifier and the metadata of the feature corresponding to the feature identifier, the feature value corresponding to the feature identifier is obtained. The metadata of the feature includes the storage information of the feature value. Based on this storage information, the required feature value can be obtained. The feature metadata includes: feature name, dimension, data type, default value, and feature pool storage information. The feature metadata can be feature identifier, carrier object type, feature name, feature data format, feature storage medium information, etc.

[0096] The model is invoked based on the feature values ​​of the features it depends on; specifically, the feature values ​​of the features it depends on are obtained, the model service is invoked, and the model calculation results are returned.

[0097] In one possible implementation, the step of retrieving a preset recognition model from a preset model library based on the feature value set, and configuring the retrieved preset recognition model on a smart device in the first set of smart devices to obtain the target device includes:

[0098] Based on the feature value set, a preset recognition model is retrieved from the preset model library;

[0099] The number of preset recognition models invoked is counted and used as the first count;

[0100] The number of intelligent devices in the first intelligent device cluster is counted and used as the second number.

[0101] Compare the first number and the second number, and select the smaller number between the first number and the second number as the target number;

[0102] A target number of the preset recognition models are selected by random sampling, and a target number of smart devices to be configured are selected from the first set of smart devices.

[0103] The selected preset recognition model is configured on the selected smart device to be configured according to the preset rules to obtain the target device. The preset rules are configured on any preset recognition model and only on one smart device to be configured.

[0104] For example, in real-world applications, the number of preset recognition models and the number of smart devices in the first set of smart devices are often different. For instance, if the number of preset recognition models is 5 and the number of smart devices in the first set of smart devices is 8, then 5 is taken as the target number. Five smart devices to be configured are selected from the first set of smart devices by random sampling. The five preset recognition models are configured on the five smart devices to be configured without repetition. That is, the preset rule is that any preset recognition model is configured on only one smart device to be configured, thus obtaining the target device.

[0105] In one possible implementation, the step of controlling the target device to acquire user request information and acquiring a first identification result of the target device in response to the user request information includes:

[0106] Obtain the human-computer interaction rules formulated for the target device, and control the target device to obtain user request information according to the human-computer interaction rules;

[0107] The first identification result of the target device in response to the user request information is obtained. The first identification result includes the smart device to be controlled corresponding to the user request information and the control instructions for the smart device to be controlled.

[0108] For example, since different target devices acquire user request information in different data formats, for instance, in the living room, the monitoring device in the living room is taken as the target device. Since the function of the monitoring device is to acquire images, the human-computer interaction rules are formulated to select the smart device to be controlled based on the user's first specific gesture (posture) and to control the smart device to be controlled based on the user's second specific gesture (posture).

[0109] In one possible implementation, the method further includes:

[0110] A preset recognition model is called from the preset model library, and the preset recognition model is configured on the smart devices in the second set of smart devices to obtain candidate devices;

[0111] Control the alternative device to acquire user request information, and acquire the third identification result of the alternative device in response to the user request information;

[0112] Based on the first identification result and the third identification result, control instructions are generated to control the smart device corresponding to the user request information.

[0113] For example, to improve the accuracy of recognition, the preset recognition model is configured on the smart devices in the second set of smart devices (the set of devices in standby mode) to obtain candidate devices, and then control instructions are generated through more user request information with differences to control the smart device corresponding to the user request information.

[0114] In one possible implementation, such as Figure 3 As shown in the figure, this application embodiment provides a control device for a smart device, the control device comprising:

[0115] The data acquisition module 201 is used to acquire the operating status dataset of all smart devices in the target area. The smart devices are electronic devices with data transmission and reception functions and data calculation functions. The operating status dataset includes the operating status of each smart device, including an on state, a standby state, and a off state.

[0116] The classification module 202 is used to classify all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set.

[0117] The calling module 203 is used to call a preset recognition model from the preset model library and configure the preset recognition model on the smart devices in the first smart device set to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition model.

[0118] The identification module 204 is used to control the target device to obtain user request information and to obtain the first identification result of the target device for the user request information;

[0119] The control module 205 is used to generate control commands based on the first identification result to control the smart device corresponding to the user request information.

[0120] By acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state, and off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from a preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information. Using smart devices as control centers reduces the computational burden on a single control center and improves the control efficiency of smart devices. When the first set of smart devices includes at least two devices, these devices can receive user commands (voice commands, gesture commands, text commands, etc.) in different ways and use different preset recognition models to identify the user's intent. This ensures the accuracy of user intent recognition. Furthermore, since different smart devices are located in different places, they can obtain user commands with different characteristics. Identifying these different user commands using different preset recognition models further ensures the accuracy of user intent recognition. Simultaneously, the location of the device receiving user commands is not fixed, allowing users to conveniently control any smart device in the smart home network from any area covered by the smart home, thus improving the user experience.

[0121] In one possible implementation, such as Figure 4As shown, this application embodiment provides an electronic device 300, including: a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following: acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the dataset of operating status includes the operating status of each smart device, including an on state, a standby state, and a off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from a preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; and generating control instructions based on the first recognition result to control the smart device corresponding to the user request information.

[0122] By acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state, and off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from the preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information. Using smart devices as control centers reduces the computational burden on a single control center and improves the control efficiency of smart devices. When the first set of smart devices includes at least two devices, these devices can receive user commands (voice commands, gesture commands, text commands, etc.) in different ways and use different preset recognition models to identify the user's intent. This ensures the accuracy of user intent recognition. Furthermore, since different smart devices are located in different places, they can obtain user commands with different characteristics. Identifying these different user commands using different preset recognition models further ensures the accuracy of user intent recognition. Simultaneously, the location of the device receiving user commands is not fixed, allowing users to conveniently control any smart device in the smart home network from any area covered by the smart home, thus improving the user experience.

[0123] In one possible implementation, such as Figure 5As shown, this application embodiment provides a computer-readable storage medium 400, on which a computer program 411 is stored. When executed by a processor, the computer program 411 performs the following: acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the dataset of operating status includes the operating status of each smart device, including an on state, a standby state, and a off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from a preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information.

[0124] By acquiring a dataset of the operating status of all smart devices within a target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state, and off state; classifying all smart devices according to the operating status to obtain a first smart device set, a second smart device set, and a third smart device set; calling a preset recognition model from the preset model library and configuring the preset recognition model on the smart devices in the first smart device set to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models; controlling the target device to acquire user request information and acquiring a first recognition result of the target device for the user request information; generating control instructions based on the first recognition result to control the smart device corresponding to the user request information. Using smart devices as control centers reduces the computational burden on a single control center and improves the control efficiency of smart devices. When the first set of smart devices includes at least two devices, these devices can receive user commands (voice commands, gesture commands, text commands, etc.) in different ways and use different preset recognition models to identify the user's intent. This ensures the accuracy of user intent recognition. Furthermore, since different smart devices are located in different places, they can obtain user commands with different characteristics. Identifying these different user commands using different preset recognition models further ensures the accuracy of user intent recognition. Simultaneously, the location of the device receiving user commands is not fixed, allowing users to conveniently control any smart device in the smart home network from any area covered by the smart home, thus improving the user experience.

[0125] The computer storage medium of this invention can be any combination of one or more computer-readable media. A computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. For example, a computer-readable storage medium can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0126] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0127] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, etc., or any suitable combination thereof.

[0128] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages—such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0129] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of multiple computing devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computing device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. Thus, the present invention is not limited to any particular combination of hardware and software.

[0130] Note that the above description is merely a preferred embodiment of the present invention and the technical principles employed. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions can be made without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention, the scope of which is determined by the scope of the appended claims.

[0131] The above description discloses only preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made in accordance with the claims of the present invention are still within the scope of the present invention.

Claims

1. A control method for an intelligent device, characterized in that, include: Obtain the operating status dataset of all smart devices within the target area, wherein the smart devices are electronic devices with data transmission and reception functions and data calculation functions, and the operating status dataset includes the operating status of each smart device, including on state, standby state and off state; All smart devices are classified according to their operating states to obtain a first set of smart devices that are in the on state, a second set of smart devices that are in the standby state, and a third set of smart devices that are in the off state. A preset recognition model is called from the preset model library and configured in the smart devices in the first set of smart devices to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models. When the first set of smart devices includes at least two smart devices, user instructions are received in different ways through at least two smart devices, and the user's intent is recognized through different preset recognition models. Control the target device to acquire user request information, and acquire the first recognition result of the target device in response to the user request information; Based on the first identification result, control instructions are generated to control the smart device corresponding to the user request information.

2. The control method for the intelligent device as described in claim 1, characterized in that, The step of obtaining the operational status dataset of all smart devices within the target area includes: Divide the user's space into at least two sub-areas according to the building structure; Image data for each sub-region is acquired, and the image data is input into the human body recognition model to obtain a second recognition result; Based on the second identification result, the sub-region where the user is located is determined as the target region; The image data corresponding to the target area is used as the target image data, and the operating status dataset of all smart devices in the target area is obtained based on the target image data.

3. The control method for the intelligent device as described in claim 2, characterized in that, The step of using image data corresponding to the target area as target image data and obtaining the operating status dataset of all smart devices within the target area based on the target image data includes: The image data corresponding to the target area is used as the target image data, and the type information of each smart device displayed by the target image data is identified; Based on the type information of each smart device, the power consumption data of each smart device, and the traffic data of each smart device, the operating status of each smart device is determined, and a dataset of the operating status of all smart devices in the target area is obtained.

4. The control method for the intelligent device as described in claim 1, characterized in that, The step of calling a preset recognition model from a preset model library and configuring the preset recognition model on a smart device in the first set of smart devices to obtain a target device, wherein the preset recognition model is an intent recognition model, and the preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models, includes: Obtain the metadata of all preset recognition models in the preset model library. The metadata includes the metadata of each preset recognition model, and the metadata includes the model feature mapping relationship. Based on the model feature mapping relationship, obtain the feature identifier of the feature on which each preset recognition model depends; Based on the feature identifier, the feature values ​​of the features that each preset recognition model depends on are obtained to obtain a feature value set; Based on the feature set, a preset recognition model is retrieved from the preset model library, and the retrieved preset recognition model is configured on the smart devices in the first set of smart devices to obtain the target device.

5. The control method for the intelligent device as described in claim 4, characterized in that, The step of retrieving a preset recognition model from a preset model library based on the feature value set, and configuring the retrieved preset recognition model on a smart device in the first set of smart devices to obtain the target device includes: Based on the feature value set, a preset recognition model is retrieved from the preset model library; The number of preset recognition models invoked is counted and used as the first count; The number of intelligent devices in the first intelligent device cluster is counted and used as the second number. Compare the first number and the second number, and select the smaller number between the first number and the second number as the target number; A target number of preset recognition models are selected by random sampling, and a target number of smart devices to be configured are selected from the first set of smart devices. The selected preset recognition model is configured on the selected smart device to be configured according to the preset rules to obtain the target device. The preset rules are configured on any preset recognition model and only on one smart device to be configured.

6. The control method for the intelligent device as described in claim 1, characterized in that, The step of controlling the target device to acquire user request information and acquiring a first identification result of the target device in response to the user request information includes: Obtain the human-computer interaction rules formulated for the target device, and control the target device to obtain user request information according to the human-computer interaction rules; The first identification result of the target device in response to the user request information is obtained. The first identification result includes the smart device to be controlled corresponding to the user request information and the control instructions for the smart device to be controlled.

7. The control method for the intelligent device as described in claim 1, characterized in that, The method further includes: A preset recognition model is called from the preset model library, and the preset recognition model is configured on the smart devices in the second set of smart devices to obtain candidate devices; Control the alternative device to acquire user request information, and acquire the third identification result of the alternative device in response to the user request information; Based on the first identification result and the third identification result, control instructions are generated to control the smart device corresponding to the user request information.

8. A control device for an intelligent device, characterized in that, The control device includes: The data acquisition module is used to acquire the operating status dataset of all smart devices within the target area. The smart devices are electronic devices with data transmission and reception functions and data calculation functions. The operating status dataset includes the operating status of each smart device, including an on state, a standby state, and a off state. The classification module is used to classify all smart devices according to their operating states to obtain a first set of smart devices that are in the on state, a second set of smart devices that are in the standby state, and a third set of smart devices that are in the off state. The calling module is used to call a preset recognition model from the preset model library and configure the preset recognition model on the smart devices in the first set of smart devices to obtain the target device. The preset recognition model is an intent recognition model. The preset model library includes at least two preset recognition models with different algorithms and feature values ​​corresponding to the preset recognition models. When the first set of smart devices includes at least two smart devices, user instructions are received through at least two smart devices in different ways, and the user's intent is recognized through different preset recognition models. The identification module is used to control the target device to obtain user request information and to obtain a first identification result of the target device in response to the user request information; The control module is used to generate control commands based on the first identification result to control the smart device corresponding to the user request information.

9. An electronic device, characterized in that, include: The device includes a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device is running, the processor communicates with the memory via the bus. When the machine-readable instructions are executed by the processor, they perform the steps of the control method of the intelligent device as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, performs the steps of the control method for the intelligent device as described in any one of claims 1 to 7.

Images