Body surface appendage form control method of artificial intelligence toy, electronic equipment, readable storage medium and computer program product

Abstract

The invention discloses a body surface appendage form control method of an artificial intelligence toy, electronic equipment, a readable storage medium and a computer program product, and belongs to the technical field of intelligent terminals. The method comprises the following steps: acquiring target situation data, wherein the target situation data comprises at least one of the following data: interaction state data of a user, environment state data of the user, interaction state data of an artificial intelligence toy and environment state data of the artificial intelligence toy; corresponding control reference data are determined according to the target situation data, and the control reference data are used for indicating the body surface appendage form corresponding to the target situation data; and controlling the shape of the body surface appendage of the artificial intelligence toy based on the control reference data.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of intelligent terminal, in particular to a body surface accessory form control method of an artificial intelligence toy, an electronic device, a readable storage medium and a computer program product. BACKGROUND

[0002] Real pets, such as cats and dogs, birds, fur and feather state can change, and the mood of the pet can be fed back through the change of the fur, for example, anger will make the hair stand on end. The existing artificial intelligence toys generally collect through audio and video, and feed back the mood of the pet through the change of the pet's voice, expression and action to respond to the user's action, expression and manner.

[0003] At present, some artificial intelligence toys also have body surface accessories, such as fur, hair or feathers, but they are fixed and cannot move, and there is a certain gap with real pets. The existing artificial intelligence toys cannot feed back the inner world of the pet through the change of the fur, simulate the interaction between the real pet and the user, and give the user a response. SUMMARY

[0004] Embodiments of the present application provide a body surface accessory form control method of an artificial intelligence toy, an electronic device, a readable storage medium and a computer program product, which can at least solve the problem that the existing artificial intelligence toy cannot feed back the inner world of the pet through the change of the fur, simulate the interaction between the real pet and the user, and give the user a response.

[0005] In order to solve the above technical problems, the present application is implemented as follows: In a first aspect, a body surface accessory form control method of an artificial intelligence toy is provided, comprising: obtaining target emotional state data, the target emotional state data comprising at least one of the following: user interaction state data, user environment state data, artificial intelligence toy interaction state data and artificial intelligence toy environment state data; determining corresponding control reference data according to the target emotional state data, wherein the control reference data is used to indicate the body surface accessory form corresponding to the target emotional state data; controlling the body surface accessory form of the artificial intelligence toy based on the control reference data.

[0006] In a second aspect, an electronic device is provided, which comprises a processor and a memory, and the memory stores at least one computer program, which is loaded and executed by the processor to implement the above-mentioned body surface accessory form control method of an artificial intelligence toy.

[0007] In a third aspect, a readable storage medium is provided, and the readable storage medium stores at least one computer program. The computer program is loaded and executed by a processor to implement the body surface accessory shape control method of the artificial intelligence toy.

[0008] In a fourth aspect, a computer program product is provided, and the computer program product includes at least one computer program. The computer program is loaded and executed by a processor to implement the body surface accessory shape control method of the artificial intelligence toy provided in various optional implementation manners.

[0009] The body surface accessory shape control method of the artificial intelligence toy, the electronic device, the readable storage medium, and the computer program product provided by the embodiments of the present application can collect target state data of a user and an artificial intelligence toy in real time, identify changing emotions of the user and the artificial intelligence toy in an interaction process, and then control the body surface accessory shape to change states according to the emotions, so as to realize various changes and shapes of the body surface accessory shape of the artificial intelligence toy, meet emotional needs of user interaction and state changes of the artificial intelligence toy, and improve user interaction experience.

[0010] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0011] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present application and, together with the specification, serve to explain the principles of the present application.

[0012] Figure 1 A flowchart of a body surface accessory shape control method of an artificial intelligence toy provided by an exemplary embodiment of the present application is shown; Figure 2 A flowchart of determining corresponding control reference data according to target state data provided by an exemplary embodiment of the present application is shown; Figure 3 A schematic diagram of an emotion and body surface accessory shape mapping relationship provided by an exemplary embodiment of the present application is shown; Figure 4 A schematic diagram of a matching relationship between an adjustment fine parameter and an accessory adjustment mode provided by an exemplary embodiment of the present application is shown; Figure 5 A structural schematic diagram flowchart of a static control adjustment module L6 provided by an exemplary embodiment of the present application is shown; Figure 6 A schematic diagram of controlling a body surface accessory shape of an artificial intelligence toy based on a body surface accessory electrostatic control adjustment mode by calling a static control adjustment module L6 provided by an exemplary embodiment of the present application is shown. Figure 7 A structural schematic diagram of the magnetic control adjustment module L7 provided by an example embodiment of the present application is shown. Figure 8 A structural schematic diagram of the electric control adjustment module L8 provided by an example embodiment of the present application is shown. Figure 9 A schematic diagram of the electric control adjustment mode of the body surface appendage of the artificial intelligence toy by calling the static control adjustment module L8 to control the form of the body surface appendage of the artificial intelligence toy is shown. Figure 10 A structural schematic diagram of the comb control adjustment module L9 provided by an example embodiment of the present application is shown. Figure 11 A flowchart of constructing an emotion and body surface appendage form learning model provided by an example embodiment of the present application is shown. Figure 12 A block diagram of the body surface appendage form control device of the artificial intelligence toy provided by an example embodiment of the present application is shown. Figure 13 A block diagram of another body surface appendage form control device of the artificial intelligence toy provided by an example embodiment of the present application is shown. Figure 14 A structural block diagram of an electronic device provided by an example embodiment of the present application is shown. DETAILED DESCRIPTION

[0013] The example embodiments will be described in detail herein, with examples shown in the drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following example embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of apparatuses and methods consistent with some aspects of the present application as detailed in the appended claims.

[0014] To solve the problem that the existing artificial intelligence toy cannot feedback the inner world of the pet through the change of the fur, simulate the interaction between the real pet and the user based on the existing artificial intelligence toy, an artificial intelligence toy body surface appendage form control method is provided in the embodiments of the present application.

[0015] The execution subject of the method can be various types of artificial intelligence toys, or can be an application or application (Application, APP) installed on the artificial intelligence toy. The artificial intelligence toy has a body surface attachment such as fur, hair or feathers, such as cat, dog, bear, bird and other pet toys with fur or feathers. During interaction with the user, the artificial intelligence toy can feedback the current inner world of the artificial intelligence toy, such as joy, gentleness, fear, worry, sadness, etc. through the state change of the surface fur, which can greatly enhance the interaction experience between the artificial intelligence toy and the user. The fur in different positions of the artificial intelligence toy can present different shapes and states, for example, under different moods and external interaction conditions, it can be erected, soft, partially rolled, and partially soft; under different interaction emotions, different shapes and patterns are presented; under different seasons or environmental temperatures, different fur lengths, patterns, and different patterns can be formed, and different shapes and patterns can be programmed to meet the emotional needs of user interaction and the state change of the artificial intelligence toy itself, and improve the user interaction experience.

[0016] The following will be combined with Figures 1 to 11 The body surface attachment shape control method of the artificial intelligence toy provided in the embodiments of the present application will be explained and described in detail. These embodiments are only used to explain the present application and do not constitute a limitation on the present application.

[0017] Figure 1 A flow chart of the body surface attachment shape control method of the artificial intelligence toy according to an example embodiment of the present application is shown. As Figure 1 The body surface attachment shape control method of the artificial intelligence toy mainly includes the following steps (S101-S103): S101, obtaining target emotional data, the target emotional data including at least one of the following: user interaction state data, user environment state data, artificial intelligence toy interaction state data and artificial intelligence toy environment state data; In this embodiment, the target emotional data is the emotional data collected in real time when the current user and the artificial intelligence toy are interacting. The target emotional data mainly includes the interaction state data of the user and the artificial intelligence toy during interaction and the environment state data of the environment they are in. By collecting the target emotional data of the user and the artificial intelligence toy in real time, the corresponding emotional label can be matched, so as to identify the emotional change of the user or the artificial intelligence toy, and then the body surface attachment shape can be controlled to change state according to the emotional change.

[0018] In some embodiments, the interaction state data of the user and the artificial intelligence toy includes at least one of the following: action state information, voice state information, facial expression information, heartbeat state information and custom state information.

[0019] For example, the action state information can be a user's touching action on the artificial intelligence toy, or a non-contact action such as a gesture action, a body action, etc. The action state information can also be an action of the artificial intelligence toy, such as body shaking, tail shaking, etc. In some embodiments, the touching or pressing action can be achieved by a contact sensor of the artificial intelligence toy, such as a capacitive or resistive sensor, by detecting the change in the resistance or capacitance value of a body surface appendage such as fur. The non-contact action information can be collected by a general camera on the artificial intelligence toy, such as an RGB camera, a depth camera, an ultrasonic sensor, a millimeter wave sensor, etc.

[0020] For example, the voice state information can be a conversation between the user and the artificial intelligence toy, a conversation between the user and another person, a user's own monologue, and a self-talk of the artificial intelligence toy. In some embodiments, the voice state information can be collected by a microphone or microphone array on the artificial intelligence toy.

[0021] For example, the facial expression information can be a facial expression and a dynamic feature of an expression change when the user interacts with the artificial intelligence toy. For example, a happy smile, an angry frown, a nervous expression, a relaxed expression, etc. The expression change of the key parts of the face includes: eyes, such as eye opening degree, pupil change, mouth, such as opening and closing, left and right corner change, chin, nose, such as arching, nasal muscle change, nostril change, eyebrow, such as frowning, raising eyebrows, cheek change, etc. In some embodiments, the facial expression change of the user can be determined by comparing the features of the key parts of the current face image of the user obtained by the shooting and image processing unit built in the artificial intelligence toy with the pre-stored feature library.

[0022] For example, the heartbeat state information can be the user's heartbeat and breathing state information when the user interacts with the artificial intelligence toy, such as heartbeat acceleration, heartbeat relaxation, heart rate, heart rate variability, heart cycle, etc. The user's heartbeat parameters can be collected by an Ultra Wide Band (UWB) chip built in the artificial intelligence toy. The UWB radar technology is a kind of communication technology using a bandwidth of 1.5 GHz or more without carrier wave, which transmits data by using non-sine wave narrow pulse of nanosecond to picosecond, and can achieve millimeter level accuracy. When the radar module transmits electromagnetic waves to detect the human body, the fluctuation of the human chest and the beating of the heart will cause different time delays of the electromagnetic waves. The module can measure the accurate distance change value according to the different time delays of the echo, and use the related signal processing method to extract the parameters related to the heart and lung, such as breathing and heart rate, to realize the function of measuring the human body breathing and heartbeat.

[0023] For example, the custom state information can be a special instruction for expressing the user's emotion, such as the user inputting the instruction to the artificial intelligence toy through voice or text: calling ****I am not happy, Mimi Mimi I am very angry****, etc. to express the user's emotion, so as to directly determine the corresponding emotion label according to the special instruction, and realize the change of the body appendage form of the artificial intelligence toy. In some embodiments, the custom state information can be matched to the custom form of the body appendage form, for example, the fluff presents the shape of a heart.

[0024] The environmental state data indicates the environmental state in which the artificial intelligence toy or the user is currently located, such as temperature, humidity, light, illumination, wind power, season, air flow quality, etc. The environmental state data can be collected by sensors built-in the pet, such as temperature sensor, light sensor, wind power sensor, etc.

[0025] In the embodiments of the present application, by collecting various interaction state data and environmental state data, sufficient sample data is ensured to be collected to match more accurate emotions.

[0026] S102, determining corresponding control reference data according to the target emotional state data, wherein the control reference data is used to indicate the body appendage form corresponding to the target emotional state data; In some embodiments, determining the corresponding control reference data according to the target emotional state data includes: extracting the key feature parameters of one or more information in the target emotional state data, such as the action state information, the voice state information, the facial expression information, the heartbeat state information, the custom state information and the environmental state data, and identifying the corresponding emotion label based on the above key feature parameters. In the embodiments, the emotion label is used to indicate the emotional state of the user or the artificial intelligence toy; and the body appendage form corresponding to the emotion label is determined as the corresponding control reference data according to the preset emotion and body appendage form mapping relationship.

[0027] In other embodiments, the corresponding control reference data is determined according to the target emotional state data based on a pre-constructed large model. Figure 2 A flowchart for determining the corresponding control reference data according to the target emotional state data is shown in an example embodiment of the present application. As shown in Figure 2 The flow of determining the corresponding control reference data according to the target emotional state data mainly includes the following steps (S201-S203): S201, inputting the target emotional state data into a pre-constructed emotion and body appendage form learning model; S202, taking the output result of the emotion and body appendage form learning model as the emotion label corresponding to the target emotional state data; wherein the emotion label is used to indicate the emotional state of the user or the artificial intelligence toy; In some embodiments, the target modal data collected in step S101 is input into a pre-constructed emotion and body appendage morphology learning model. The learning model can then obtain emotion labels corresponding to the target modal data, identifying changes in the user's or AI toy's emotions. This emotion and body appendage morphology learning model is trained using historical modal data and the emotional states of the user or AI toy as training data.

[0028] For example, emotion labels include: (A) positive emotional states, such as happiness, joy, happiness, excitement, bliss, pleasure, satisfaction, anticipation, etc.; (B) negative emotions, such as unhappiness, sadness, sadness, heartbreak, loss, frustration, pain, helplessness, anxiety, etc.; (C) fear emotions, such as fear, terror, dread, panic, tension, unease, etc.; (D) worry emotions, such as worry, unease, etc.; (E) shy emotions, such as shyness, shame, embarrassment, etc.; (F) anger emotions, such as anger, resentment, rage, etc.; (G) disgust emotions, such as dislike, annoyance, disgust, etc.; (H) trust emotions, such as trust, dependence, closeness, security, reliability, etc.; (I) anticipation emotions, such as anticipation, dreams, longing, yearning, etc.; (J) liking emotions, such as liking, fondness, admiration, etc.; (K) resentment emotions, such as resentment, hatred, blame, etc.; (L) nostalgia emotions, such as longing, nostalgia, reminiscence, recollection, reminiscence, etc. The emotions of users or AI toys may not be singular but complex and multifaceted. Therefore, the emotion labels can be combinations of the emotion labels shown above. For example, they can be combined in sequence, such as first being happy and then sad, like crying tears of joy. The final fluff pattern could be a fluff pattern of happiness followed by a fluff pattern of sadness, forming a dynamic and changing effect.

[0029] S203. Based on the preset mapping relationship between emotions and body surface appendages, determine the body surface appendage morphology corresponding to the emotion label as the corresponding control reference data.

[0030] In some embodiments, the type of morphology of body surface appendages includes at least one of the following: fixed morphology, dynamic morphology, and custom morphology. A fixed morphology refers to maintaining a certain shape. For example, a fixed morphology may include: X1. Compliant morphology, such as compliant, flat, or fluffy; X2. Upright morphology, such as upright, raised, forward-leaning, or backward-leaning; X3. Curly morphology, such as curly, rolled, or sheep curls; X4. Soft, drooping morphology, such as soft or drooping; X5. Messy morphology, such as messy, disordered, or tangled. A dynamic morphology refers to a variable morphology. For example, it may include flowing, breathable undulations, dynamic frizz, dynamic shaking, wavy undulations, a change from upright to flat, or a change from disordered to compliant. A custom morphology refers to defining the local morphology of body surface appendages through programming, forming specific letters, patterns, or shape rules. For example, custom-defined spots, stripes, hearts, smiley faces, etc., with different shapes.

[0031] In some embodiments, the method provided by the embodiments of the present application further comprises: setting the shape of the body surface accessory of the artificial intelligence toy through the programmable control unit to obtain the various types and shapes of the body surface accessory shape described above. In some embodiments, the programmable control unit comprises: a programmable electrostatic villi circuit unit, a programmable magnetic control villi circuit unit, a programmable electric control villi circuit unit, and a programmable villi comb control circuit unit.

[0032] For example, the surface of the artificial intelligence toy is provided with an electrostatic sensing surface, and a static control adjustment module L6 is arranged inside the toy. The static control adjustment module L6 calls the programmable electrostatic villi circuit unit to control the body surface accessory of the artificial intelligence toy to change in different electric fields, for example, the villi can change in different directions, angles, patterns, and flexibility.

[0033] For example, the surface of the artificial intelligence toy is provided with a magnetic control sensing surface, and a magnetic response coating containing magnetic materials or particles is added to the body surface accessory. A magnetic control adjustment module L7 is arranged inside the toy. The magnetic control adjustment module L7 calls the programmable magnetic control villi circuit unit to control the body surface accessory of the artificial intelligence toy to form a specified state, for example, to control the villi to vibrate or undulate in a wave pattern to simulate the state change of the artificial intelligence toy being nervous or happy.

[0034] For example, the body surface accessory of the artificial intelligence toy is coated with a conductive coating, and a current sensing surface is arranged on the surface of the body surface accessory of the artificial intelligence toy. An electric control adjustment module L8 is arranged inside the toy. The electric control adjustment module L8 calls the programmable electric control villi circuit unit to realize various shapes through the coating, such as vertical, messy, flat, curly, fluffy, etc.

[0035] For example, a villi comb structure is arranged on the surface of the body surface accessory of the artificial intelligence toy, and a comb control adjustment module L9 and a micro motor controller are arranged inside the toy. The comb control adjustment module L9 calls the programmable villi comb control circuit unit to move the grooming components at the root of the villi through the motor to achieve the effect of gradual grooming, making the villi soft or flat.

[0036] In this embodiment, each adjustment module can control the body surface accessory shape of the artificial intelligence toy to present various emotion labels corresponding to the body surface accessory shape by calling the programmable control unit described above, so as to present the change and various shapes of the body surface accessory shape.

[0037] In some embodiments, the emotion label is mapped and matched with the body surface accessory shape, and the mapping relationship between the emotion label and the body surface accessory shape corresponding to the emotion and the body surface accessory shape is pre-constructed. In some embodiments, the mapping relationship can be configured by algorithm or configured by the operator himself. For example,Figure 3 An emotion and body appendage shape mapping relationship diagram is shown in an example embodiment of the present application. As shown, different emotion labels can overlap with different body appendage shapes, such as the disgust emotion and the expectation emotion can both correspond to the dynamic hair shape, and the worry emotion can correspond to the curly shape and the messy shape. It can be understood that, Figure 3 The emotion and body appendage shape mapping relationship shown is only an example for illustration, and is only used to explain the present application, and does not constitute a limitation on the present application. Figure 3 The emotion and body appendage shape mapping relationship shown is only an example for illustration, and is only used to explain the present application, and does not constitute a limitation on the present application.

[0038] In this embodiment, after determining the corresponding emotion label according to the target emotion data, the body appendage shape corresponding to the emotion label can be determined as the control reference data based on the emotion and body appendage shape mapping relationship, and the body appendage shape of the artificial intelligence toy is controlled and adjusted.

[0039] S103, controlling the body appendage shape of the artificial intelligence toy based on the control reference data.

[0040] In some embodiments, controlling the body appendage shape of the artificial intelligence toy based on the control reference data includes: matching a corresponding artificial intelligence toy body appendage adjustment mode according to the control reference data, and controlling the body appendage shape of the artificial intelligence toy based on the artificial intelligence toy body appendage adjustment mode.

[0041] In some embodiments, the artificial intelligence toy body appendage adjustment mode includes at least one of the following: a body appendage electrostatic control adjustment mode, a body appendage magnetic control adjustment mode, a body appendage electric control adjustment mode, and a body appendage comb control adjustment mode. In this embodiment, the body appendage electrostatic control adjustment mode is used to display the body appendage shape after electrostatic induction through the electrostatic induction surface or electrostatic coating of the artificial intelligence toy; the body appendage magnetic control adjustment mode is used to display the body appendage shape after magnetic induction through the magnetic induction surface or magnetic attraction coating of the artificial intelligence toy; the body appendage electric control adjustment mode is used to display the body appendage shape after electric control induction through the electric field induction surface or conductive coating of the artificial intelligence toy; and the body appendage comb control adjustment mode is used to display the combing shape through the combing structure and circuit of the artificial intelligence toy.

[0042] In some embodiments, matching the corresponding artificial intelligence toy specific surface accessory adjustment mode according to the control reference data comprises: determining the adjustment mode priority according to the control reference data; and matching the corresponding artificial intelligence toy specific surface accessory adjustment mode based on the adjustment mode priority. In this embodiment, the user's emotion during interaction with the artificial intelligence toy is constantly changing, and therefore the form of the surface accessory of the artificial intelligence toy also needs to correspond to a variety of changes and states, or a series of combinations of fur changes, and the priority of the above four adjustment modes and the combination mode, order, etc. can be determined according to different emotion labels, and are used one by one until the emotional needs of the user interaction and the pet's own state change are met.

[0043] In an application example, the priority of the surface accessory adjustment mode call is as follows: a1. The artificial intelligence toy has a static induction surface or a static coating, or needs to show the fur effect after static induction, such as surprise and terror, and then the static control adjustment module L6 is called to control the form of the surface accessory of the artificial intelligence toy based on the surface accessory static control adjustment mode; a2. If the artificial intelligence toy has a magnetic control induction surface or a magnetic attraction coating, or needs to show the fur effect after magnetic induction, such as fur state recovery, low crouch softness, standing up, and lying down, then the magnetic control adjustment module L7 is called to control the form of the surface accessory of the artificial intelligence toy based on the surface accessory magnetic control adjustment mode; a3. If the artificial intelligence toy has an electric field induction surface or a conductive coating, or needs to show the fur effect after electric control induction, such as vertical, messy, lying down, curling, and fluffy, then the electric control adjustment module L8 is called to control the form of the surface accessory of the artificial intelligence toy based on the surface accessory electric control adjustment mode; a4. If the artificial intelligence toy has a comb structure and a circuit, or needs to show a comb effect, such as different shapes, states, patterns, or actions, then the comb control adjustment module L9 is called to control the form of the surface accessory of the artificial intelligence toy based on the surface accessory comb control adjustment mode; a5. If the artificial intelligence toy does not have direct interaction with the user, the artificial intelligence toy can adjust the fur state according to the environmental state or the emotional state of the artificial intelligence toy.

[0044] a6. If the form of the surface accessory of the artificial intelligence toy needs to be adjusted to a variety of changes and states, or a series of combinations of changes, then the L6-L9 adjustment modules can be adjusted and called one by one, or two by two, or more combinations can be used together until the emotional needs of the user interaction and the pet's own state change are met.

[0045] In some embodiments, the corresponding artificial intelligence toy specific surface accessory adjustment mode is matched according to the control reference data matching, comprising: determining the adjustment fine parameter according to the control reference data, the adjustment fine parameter comprising one of the following: form adjustment strength, form adjustment area and form adjustment accuracy; and matching the corresponding artificial intelligence toy specific surface accessory adjustment mode based on the adjustment fine parameter. In this embodiment, corresponding to different adjustment modes, in addition to adjusting different forms of output, the artificial intelligence toy specific surface accessory adjustment mode can also be more finely controlled according to the judgment and needs of the adjustment strength, form adjustment area and form adjustment accuracy.

[0046] Exemplary, Figure 4 An adjustment fine parameter and accessory adjustment mode matching relationship diagram shown in an exemplary embodiment of the present application is shown. As Figure 4 shown, different form adjustment strength, form adjustment area and form adjustment accuracy can correspond to different accessory adjustment modes. It can be understood that, Figure 4 The adjustment fine parameter and accessory adjustment mode matching relationship shown is only exemplary and is used to explain the present application and does not limit the present application.

[0047] In some embodiments, the artificial intelligence toy specific surface accessory adjustment mode further comprises: an intelligent environment adaptive adjustment mode, which can realize dynamic surface accessory form feedback and change according to user or artificial intelligence toy peripheral environment detection, without user interaction instruction, and the surface accessory change is completely autonomous. Using the environment state data collected in step S101, the pet fur can adjust the adjustment mode of the surface accessory according to the above environment state data, and call different adjustment modules to realize adaptive form presentation. Exemplary, when the room temperature is high, the fur can change to swell, simulating a very hot state, when the room temperature is low, the fur can change to shrink, simulating a very cold state. If the room temperature is suitable, the air detection is suitable, and the fur can adaptively show a floating state. Through this adaptive adjustment control, the user can indirectly understand the change of the artificial intelligence toy and the surrounding environment. The priority of pet fur form adjustment based on pet emotion and user emotion is as follows: first based on user emotion adjustment, second based on pet emotion adjustment, and lastly environment information adaptive adjustment when both have no detection information.

[0048] The collection of environment state data and the calling of intelligent environment adaptive adjustment mode can be an independent adjustment process, which does not depend on the interaction between the user and the artificial intelligence toy. When the user does not interact with the toy, the artificial intelligence toy can adaptively change its fur state according to the current environmental change.

[0049] In some embodiments, the static control adjustment module L6, the magnetic control adjustment module L7, the electric control adjustment module L8 and the comb control adjustment module L9 are built-in in the artificial intelligence toy. The static control adjustment module L6 is used for adjusting the state of the fluff based on the electrostatic principle, the magnetic control adjustment module L7 is used for adjusting the state of the fluff based on the magnetic control principle, the electric control adjustment module L8 is used for adjusting the state of the fluff based on the electric control principle, and the comb control adjustment module L9 is used for adjusting the state of the fluff based on the comb structure and the micro motor controller.

[0050] In some embodiments, the electronic body appendage adjustment mode is used to control the body appendage shape of the artificial intelligence toy, including at least one of the following modes based on the adjustment module.

[0051] (1) Based on the static control adjustment mode of the body appendage, the body appendage shape of the artificial intelligence toy is controlled by at least one of the following modes: The arrangement direction of the body appendage is changed by the change of the electrostatic field; the local shape and movement direction of the body appendage are changed by the electrostatic polarization effect; the distance and local and overall shape of the body appendage are controlled by the movable electrostatic controller; the electric field direction, intensity and area distribution of the body appendage are edited by the electrostatic motion pattern unit, and the directional arrangement, angle change and pattern formation of the body appendage are controlled. (2) Based on the magnetic control adjustment mode of the body appendage, the body appendage shape of the artificial intelligence toy is controlled by the following modes: The magnetic field parameters are adjusted to control the body appendage to arrange along the magnetic field direction according to the preset angle, pattern and sequence. (3) Based on the electric control adjustment mode of the body appendage, the body appendage shape of the artificial intelligence toy is controlled by at least one of the following modes: The corresponding electric field is formed by the micro electrode to control the body appendage to arrange along the electric field line direction according to the preset angle, pattern and sequence; the local shape and movement direction of the body appendage are changed by the influence of dielectric force; the thickness and length of the body appendage are changed by the micro electric controller. (4) Based on the comb control adjustment mode of the body appendage, the body appendage shape of the artificial intelligence toy is controlled by the following modes: The combing control parameters are configured, and the combing components at the root of the body appendage are moved by the motor based on the combing control parameters to form different combing shapes.

[0052] The following will be described in detail in combination with specific application examples and Figures 5 to 10 The above four modes will be described and explained in detail For mode (1): In this embodiment (I), the body surface accessory form of the artificial intelligence toy is controlled based on the electrostatic control adjustment mode by calling the static control adjustment module L6. Exemplarily, Figure 5 A structural schematic diagram of the static control adjustment module L6 is shown according to an exemplary embodiment of the present application. As shown in the figure, Figure 5 The static control adjustment module L6 includes: a static induction control unit L61, a static polarization control unit L62, a static distance control unit L63, a static motion pattern unit L64, and a static deformation recovery unit L65. Exemplarily, Figure 6 A schematic diagram of controlling the body surface accessory form of the artificial intelligence toy based on the electrostatic control adjustment mode by calling the static control adjustment module L6 is shown according to an exemplary embodiment of the present application.

[0053] Exemplarily, the static induction control unit L61 generates electric fields of positive and negative static electricity, and controls the standing direction of the pet fur through the action of the static electric field. As shown in the figure, Figure 6 The artificial intelligence toy fur can be arranged according to the direction of the electric field under the action of static electricity, such as vertical standing, forward-leaning standing, backward-leaning standing, compliance, lying flat on the surface, or fur arranged in different angles, patterns, or patterns, or floating.

[0054] Exemplarily, the static polarization control unit L62 utilizes the polarization effect of the fur in the static electric field. When the fur contacts the negative electrode, a conductive current will be generated in the fur due to the higher conductivity of the electrode than the fur, and the fur will generate a negative static charge, so that the fur has a large deformation degree in the electric field, and different patterns are presented according to the opening and closing of different positions of the static control surface. As shown in the figure, Figure 6 The artificial intelligence toy fur undergoes local deformation and moving direction deformation under the action of static polarization, such as local deformation, opening and closing deformation, directional inclination, directional bending, concave-convex deformation, and gathering deformation.

[0055] Exemplarily, a movable static distance control unit L63 is arranged in the artificial intelligence toy. When the static distance control unit L63 and the fur sensing surface of the artificial intelligence toy are relatively close or far away, the system capacitance on the sensing surface will change, the sensing charge will be rearranged, thereby generating induced static electricity, and further controlling the fur state to change. As shown in the figure, Figure 6As shown, the artificial intelligence toy's fluff is under the action of the electrostatic proximity control unit L63, and the deformation core of the fluff is driven by the distance change between the controller and the sensing surface. Different distances correspond to different charge densities and electric field strengths, realizing adjustable deformation different from natural static electricity, such as directional deformation, directional fluffiness, directional contraction, directional attachment, directional separation, local bulging, local concave, dynamic deformation. The distance between the electrostatic proximity control unit L63 and the fluff sensing surface is dynamically adjusted. When the electrostatic proximity control unit L63 approaches the fluff sensing surface, the electrostatic density on the sensing surface increases. When the electrostatic force acting on the fluff exceeds its own gravity and friction, the fluff changes from the flat state to the fluffy and upright state. When the electrostatic proximity control unit L63 moves away from the sensing surface, the electrostatic density on the sensing surface decreases. When the electrostatic force is less than the gravity and friction of the fluff, the fluff returns to the initial flat state. The fluff sensing surface of the artificial intelligence toy is divided into regions, such as the head, ears, neck, torso, limbs, and tail, each corresponding to an independent electrostatic proximity control unit L63. Local proximity or local distance control is achieved through a single electrostatic proximity control unit L63. By adjusting the local electrostatic charge density, the fluff can produce local bulging or local concave deformation. For example, the electrostatic motion pattern unit L64 can generate electric field direction, strength, and regional distribution editing for the fluff of the artificial intelligence toy, realizing directional arrangement, angle change, pattern formation, and other morphological changes. In electrostatic flocking processing, in addition to being charged by positive and negative contact, the fluff can also be charged by electric field polarization, ensuring that the fluff moves towards the positive or negative plate direction, and prompting the fluff to change in different electric fields, such as Figure 6 As shown, the fluff can change in different directions, angles, patterns, and flexibility, forming line pattern, filling pattern, and dynamic change pattern. For example, through the electrostatic motion pattern unit L64 on the back of the artificial intelligence toy, the fluff is arranged in a heart-shaped outline line or simulates pet hair stripes through a curved electrode electric field.

[0056] For example, when the fluff of the artificial intelligence toy needs to be restored, the electrostatic deformation recovery unit L65 can be called to ground the sensing surface and conduct the static electricity on the fluff to the ground. As shown, Figure 6 As shown, the artificial intelligence toy's fluff is under the action of the electrostatic proximity control unit L63, and the deformation core of the fluff is driven by the distance change between the controller and the sensing surface. Different distances correspond to different charge densities and electric field strengths, realizing adjustable deformation different from natural static electricity, such as directional deformation, directional fluffiness, directional contraction, directional attachment, directional separation, local bulging, local concave, dynamic deformation. The distance between the electrostatic proximity control unit L63 and the fluff sensing surface is dynamically adjusted. When the electrostatic proximity control unit L63 approaches the fluff sensing surface, the electrostatic density on the sensing surface increases. When the electrostatic force acting on the fluff exceeds its own gravity and friction, the fluff changes from the flat state to the fluffy and upright state. When the electrostatic proximity control unit L63 moves away from the sensing surface, the electrostatic density on the sensing surface decreases. When the electrostatic force is less than the gravity and friction of the fluff, the fluff returns to the initial flat state. The fluff sensing surface of the artificial intelligence toy is divided into regions, such as the head, ears, neck, torso, limbs, and tail, each corresponding to an independent electrostatic proximity control unit L63. Local proximity or local distance control is achieved through a single electrostatic proximity control unit L63. By adjusting the local electrostatic charge density, the fluff can produce local bulging or local concave deformation.

[0057] For mode (two): In this embodiment (two), the body surface accessory shape of the artificial intelligence toy is controlled based on the magnetic control adjustment mode by calling the magnetic control adjustment module L7. Exemplarily, Figure 7 The structural diagram of the magnetic control adjustment module L7 shown in an exemplary embodiment of the present application is shown. As shown in the figure, Figure 7 The magnetic control adjustment module L7 includes a magnetic control fluff state adjustment unit L71, a magnetic control fluff coating unit L72, a magnetic control state recovery unit L73, and a programmable dynamic magnetic control unit L74.

[0058] Exemplarily, the surface of the artificial intelligence toy is provided with a magnetic control induction surface, and the magnetic control fluff state adjustment unit L71 is arranged inside the toy. The fluff contains magnetic particles and magnetic induction composite materials, so that the fluff can respond to the external magnetic field, and at the same time, it can also maintain a soft touch. After the fluff is magnetized, the magnetic control fluff state adjustment unit L71 can control the magnetic moment of the fluff to arrange it in a specific angle, pattern, and order along the direction of the magnetic field, and produce state changes such as standing or lying down. In some embodiments, the strength, direction, frequency, and other parameters of the magnetic field generated by the magnetic control fluff state adjustment unit L71 can be adjusted to flexibly control the swing or floating mode of the fluff, and form a dynamic change effect.

[0059] Exemplarily, a magnetic response coating is added to the fluff of the artificial intelligence toy, which contains magnetic materials or particles. The magnetic control fluff coating unit L72 is arranged inside the toy to generate different magnetic field strength and direction signals. Under the action of the external magnetic field, the magnetic coating will arrange itself along the direction of the magnetic field, thereby driving the fluff to change its state along the direction of the magnetic field, such as standing up or lying down.

[0060] Exemplarily, after the fluff of the artificial intelligence toy is disordered during use or changes its shape due to static electricity control, the magnetic control state recovery unit L73 eliminates the static electricity of the fluff at different positions, directions, and angles, neutralizes or cancels the positive and negative charges on the fluff. For example, if the fluff is currently in a standing state, the magnetic control state recovery unit L73 outputs a signal with opposite polarity, reduces the static electricity of the fluff, and restores the fluff to a specified state, such as becoming low and soft.

[0061] Exemplarily, by configuring parameters through the programmable dynamic magnetic control unit L74, the fluff of the artificial intelligence toy can be made to vibrate, wave, protrude and flatten, directionally curl, shrink, or change in bulkiness.

[0062] In an application example of controlling the dynamic shaking of the fur of an artificial intelligence toy through the programmable dynamic magnetic control unit L74, by periodically changing the parameter configuration of the magnetic field direction or strength of the programmable dynamic magnetic control unit L74, the fur is subjected to a periodically changing magnetic force, and the fur will form a swinging or vibrating morphological change under the action of the force, simulating the state of the pet's hair shaking. This shaking can be a synchronous shaking of the whole fur, or a random shaking of the local fur, to feedback the different emotions and states of the weak crown, to simulate the real pet's inner shaking or happy state change.

[0063] In an application example of controlling the wave undulation of the fur of an artificial intelligence toy through the programmable dynamic magnetic control unit L74, by programming the programmable dynamic magnetic control unit L74 to control the strength and direction of the magnetic field in different areas, the fur is bent and stretched in sequence according to the predetermined target, forming a wave undulation form. For example, starting from one end of the body of the artificial intelligence toy, the fur is bent in sequence to the other end, forming an effect similar to the propagation of waves, which can be used to simulate the dynamic change of the hair of a real pet when it is happy.

[0064] In an application example of controlling the wave undulation of the fur of an artificial intelligence toy through the programmable dynamic magnetic control unit L74, by programming the programmable dynamic magnetic control unit L74 to control the strength and direction of the magnetic field in different areas, the fur is bent and stretched in sequence according to the predetermined target, forming a wave undulation form. For example, starting from one end of the body of the artificial intelligence toy, the fur is bent in sequence to the other end, forming an effect similar to the propagation of waves, which can be used to simulate the dynamic change of the hair of a real pet when it is happy.

[0065] In an application example of controlling the wave undulation of the fur of an artificial intelligence toy through the programmable dynamic magnetic control unit L74, by programming the programmable dynamic magnetic control unit L74 to control the strength and direction of the magnetic field in different areas, the fur is bent and stretched in sequence according to the predetermined target, forming a wave undulation form. For example, starting from one end of the body of the artificial intelligence toy, the fur is bent in sequence to the other end, forming an effect similar to the propagation of waves, which can be used to simulate the dynamic change of the hair of a real pet when it is happy.

[0066] In an application example of controlling the wave undulation of the fur of an artificial intelligence toy through the programmable dynamic magnetic control unit L74, by programming the programmable dynamic magnetic control unit L74 to control the strength and direction of the magnetic field in different areas, the fur is bent and stretched in sequence according to the predetermined target, forming a wave undulation form. For example, starting from one end of the body of the artificial intelligence toy, the fur is bent in sequence to the other end, forming an effect similar to the propagation of waves, which can be used to simulate the dynamic change of the hair of a real pet when it is happy.

[0067] In an application example of controlling the dynamic fur lines of an artificial intelligence toy through a programmable dynamic magnetic control unit L74, a dynamic changing magnetic field is generated through the programmable dynamic magnetic control unit L74, and the fur presents a variety of dynamic pattern changes under the programmable periodic parameter configuration of the magnetic field strength, direction, and action area, such as wave fluctuation pattern, ripple wave pattern, dynamic mottle pattern, and gradient pattern.

[0068] For mode (three): In this embodiment (three), the shape of the body surface appendage of the artificial intelligence toy is controlled based on the electrically controlled adjustment mode of the body surface appendage by calling the electrically controlled adjustment module L8. For example, Figure 8 The structure diagram of the electrically controlled adjustment module L8 shown in an example embodiment of the present application is shown. As Figure 8 shown, the electrically controlled adjustment module L8 includes: a polar electric shape adjustment unit L81, a dielectric shape adjustment unit L82, and an electrically controlled shape adjustment unit L83. For example, Figure 9 The schematic diagram of controlling the shape of the body surface appendage of the artificial intelligence toy based on the electrically controlled adjustment mode of the body surface appendage by calling the static control adjustment module L8 shown in an example embodiment of the present application is shown.

[0069] For example, the fur of the artificial intelligence toy is coated with a conductive coating, and a current induction surface is provided on the surface of the toy. A plurality of groups of arrayed or arrayed microelectrodes are provided on the polar electric shape adjustment unit L81 to form corresponding electric fields. After the fur is charged in the electric field, it is arranged along the electric field line direction to form a specific shape. For example, Figure 9 As shown, the fur can realize accurate control of direction, angle, and state under the action of Coulomb force, such as vertical, disheveled, lying, curled, and fluffy shapes. The electric level for controlling the state of the fur is composed of upright electric level, close electric level, fractal electric level, ring electric level, rectangular electric level, and wave electric level according to the different distribution positions of the fur, forming different changes in the state of the fur. For example, straight fur, close-to-surface fur, dynamic wave fur, or specific pattern shapes such as hearts, letters, numbers, stars, and suns.

[0070] For example, the artificial intelligence toy is coated with a neutral nano coating without charge on the fur. A dielectric form adjusting unit L82 is arranged inside or under the surface of the toy. Under the driving of the dielectric form adjusting unit L82, a non-uniform electric field is generated. The conductive coating on the fur can change under the action of the electric field, and part of the fur is positively charged and part of the fur is negatively charged. Under the influence of the unbalanced dielectric force at both ends, the part with strong electric field is subjected to large force and the part with weak electric field is subjected to small force, so that the state of the fur is deformed to form different shapes, such as Figure 9 For example, sagging, bending, lying down, spreading, gathering, and the like. In some embodiments, the dielectric form adjusting unit L82 is arranged to be movable, and a strong or weak electric field can be generated. When the sensing surface connected to the dielectric form adjusting unit L82 changes position, the fur changes direction with the moving direction of the sensing surface, so as to adjust the direction, angle and bending degree of the fur at different positions.

[0071] For example, the electrically controlled deformation of the fur of the artificial intelligence toy is formed in two ways. In some embodiments, a heat change coating material is applied to the fur, and different sensing areas are arranged on the surface of the toy. The sensing areas are connected to an electrically controlled form adjusting unit L83, and different micro currents are generated by the electrically controlled form adjusting unit L83. The micro currents can cause slight heat changes, and the coating on the fur changes in form under the heat changes. In other embodiments, a variable shrinkage rate coating material is applied to the fur, and an electrically controlled form adjusting unit L83 is arranged inside the toy. Different electric fields are formed by the electrically controlled form adjusting unit L83. Due to the special material, the fur generates charge rearrangement under different electric fields, forming corresponding shrinkage deformation and expansion deformation, so as to form different changes in thickness, length and extension, as shown in Figure 9 For example, when the artificial intelligence toy captures that the user is happy, the fur becomes thick and expands to form a fluffy effect, and when the artificial intelligence toy captures that the user is sad, the fur becomes thin and shrinks to form a soft couch effect.

[0072] For mode (four): In this embodiment (four), the comb control adjustment module L9 is called to control the form of the body surface appendage of the artificial intelligence toy based on the comb control adjustment mode of the body surface appendage. For example, Figure 10 As shown in Figure 10 The comb control adjustment module L9 includes a programmable fur combing control unit L91, a micro motor control unit L92 and a combing component L93.

[0073] For example, the artificial intelligence toy is provided with a plush comb structure, and a programmable plush comb control unit L91 and a micro motor control unit L92 are arranged in the toy. Under the action of the programmable plush comb control unit L91, the micro motor control unit L92 drives the combing part L93 at the root of the plush to move, achieving the effect of step-by-step combing, so that the plush is soft and laid down. Optionally, the combing part L93 at the root of the plush can be multiple groups, and the direction can be edited, so that the plush can be combed in different directions according to the control reference data corresponding to the emotional label. The programmable plush comb control unit L91 can adjust the motion direction, speed, force, and grouping mode of the combing part L93 through programming to form different shapes or patterns, and meet the personalized emotional needs of the interactive user.

[0074] In some embodiments, the motor of the combing part L93 can be reversed and the speed can be adjusted by controlling the switch state to form different combing modes, such as smooth hair mode, laid down mode, global regular mode, local regular mode, front combing mode, rear combing mode, middle parting mode, back head mode, wave mode, specific styling mode, zoned styling mode, specific pattern texture mode, cartoon mode, and personification mode. For example, when the artificial intelligence toy is in a happy and relaxed mood, the micro motor control unit L92 controls the combing part L93 to enter the global smooth hair mode. When the artificial intelligence toy is in a nervous and timid mood, the micro motor control unit L92 controls the combing part L93 to enter the local (such as the head) contraction and laid down state. For example, the specific styling mode is that the plush on the top of the head of the artificial intelligence toy is combed into an upright mode, the forehead plush is combed into a 3 / 7 parting mode, and the tail plush is combed into an explosive hair mode, to simulate the shape of the real pet with a spirited spirit.

[0075] In this embodiment, the body surface accessory shape of the artificial intelligence toy is controlled based on the adjustment mode of the body surface accessory of the artificial intelligence toy by calling the above adjustment module. The shape of the body surface accessory of the artificial intelligence toy can be changed in multiple ways according to the changing emotional label of the user and the artificial intelligence toy during the interaction, to meet the emotional needs of the user interaction and the pet state change.

[0076] In some embodiments, before the above steps S201-S203, the method provided by the embodiment of the present application further includes a process of constructing an emotional and body surface accessory shape learning model. Figure 11 A flowchart of the process of constructing an emotional and body surface accessory shape learning model according to an example embodiment of the present application is shown.

[0077] As Figure 11 shown, the process of constructing an emotional and body surface accessory shape learning model can mainly include the following steps (S301-S303): S301, acquire historical scenario data, the historical scenario data includes at least one of the following: user interaction state data in the past preset time, user environment state data in the past preset time, artificial intelligence toy interaction state data in the past preset time and artificial intelligence toy environment state data in the past preset time; In the embodiment, historical scenario data in the past period of time, for example, a week, a month, is acquired as training data of the large model. In the embodiment, the meanings and acquisition manners of the user and artificial intelligence toy interaction state data and environment state data in the historical scenario data are similar to those of the target scenario data in step S101, and the related description in step S101 can be referred to, and details are not described herein.

[0078] S302, data cleaning and feature extraction are performed on the historical scenario data, and a multi-modal feature vector of the historical scenario data is obtained. In some embodiments, the historical scenario data is cleaned by using a preprocessing algorithm, for example, filtering, amplification and normalization processing, to ensure the quality and consistency of the training data and reduce errors and fluctuations. Key feature parameter points are extracted from the cleaned historical scenario data. In some embodiments, one or more parameters in the historical scenario data, i.e., action state information, voice state information, facial expression information, heartbeat state information and self-defined state, and environment state data, are spliced into a multi-modal feature vector. For example, the columns of the multi-modal feature vector include: action, voice, expression and temperature, sunshine, and each column corresponds to the data of the column.

[0079] S303, the multi-modal feature vector of the historical scenario data and the reference weight and emotion label corresponding to each parameter in the historical scenario data are used as training data to learn and train the emotion and body appendage form learning model.

[0080] In the embodiment, each parameter has a different weight value, and the weight of each parameter is continuously optimized to extract effective feature parameters. In some embodiments, the method provided in the embodiment of the application further includes: adjusting the reference weight corresponding to each parameter in the historical scenario data by: adjusting the test body appendage form of the artificial intelligence toy according to the reference weight corresponding to each parameter in the historical scenario data, the emotion label and the body appendage form; and adjusting the reference weight corresponding to each parameter in the historical scenario data by collecting the feedback result of the user to the test body appendage form.

[0081] Exemplarily, the standard action, voice, expression, heart rate amplitude and period, and environmental parameters are reinforced to obtain the baseline value of the historical emotional data, and the baseline value is the initial input data of the emotion and body appendage shape learning model. When the learning model detects multiple parameter inputs, the weight value corresponding to each parameter is calculated according to the detection rate and proportion of each parameter, and the corresponding emotion label is output. The morphology of the body appendage is controlled and adjusted based on the control reference data corresponding to the emotion label, and then the feedback of the user is collected through the closed loop to determine whether the current morphology of the body appendage is appropriate. If not, the weight proportion of each parameter is adjusted, and the adjusted emotion label is output again. The morphology of the body appendage is controlled and adjusted based on the control reference data corresponding to the emotion label, and secondary feedback is performed until the feedback requirement is met. If the adjustment is effective, the label is increased by 1, and if the adjustment is ineffective, the label is decreased by 1. Through the feedback information, if the feedback is correct, the weight of the key feature parameters of the model is increased by 1 in the future, and if the feedback is incorrect, the weight of the key feature parameters of the model is decreased by 1 in the future, according to the effect and confidence of the reinforcement learning model.

[0082] The above steps S301-S303 are the operation flow of one training cycle. In each training cycle, the training is completed according to the above flow, and the operation of the above training cycle is repeated until the preset number of training cycles (iteration times) is reached or the multi-objective loss function is less than the preset iteration convergence value, and the training of the emotion and body appendage shape learning model is ended.

[0083] In some embodiments, the method provided by the embodiments of the present application further includes collecting the feedback result of the user on the morphology of the body appendage of the artificial intelligence toy, and adjusting the reference weight corresponding to each parameter in the historical emotional data of the emotion and body appendage shape learning model. In the embodiment, in the process of controlling the morphology of the body appendage of the artificial intelligence toy based on the control reference data, the learning model can also judge whether the current morphology of the body appendage is appropriate according to the feedback of the user on the morphology of the body appendage in the current real-time interaction process, and continue to adjust the weight proportion of each parameter if it is not appropriate. In the embodiment, the weight values of the parameters are adjusted through the closed loop user feedback to make the model converge, the method of marking effective adjustment of the model is used to train the selected learning model, and the weight values of the model parameters can be continuously adjusted and optimized through the feedback in the real-time interaction process, thereby further improving the effectiveness, accuracy and robustness of the learning model.

[0084] In some embodiments, the emotion and body appendage morphology learning model can select, for example, a convolutional neural network (CNN) model, a Transformer model, and a conditional generative adversarial network model (CGAN), and the selection of the deep learning model is not limited in the present application.

[0085] The model training in the present embodiment is exemplarily explained below in combination with the above deep learning model.

[0086] (1) An example of the emotion and body appendage morphology learning model adopting a CNN model The facial expression information of the interactive user is collected by the artificial intelligence toy built-in camera, and the collected facial expression information is taken as the input of the CNN model. The CNN model realizes the mapping of the facial expression information and the emotion label by extracting the facial organ features of the user, such as eyes, mouth, eyebrows, etc., and outputs the corresponding emotion label, such as anger, happiness, sadness, etc., to drive the artificial intelligence toy to change the fur state, as the interactive emotional feedback response.

[0087] Specifically, the CNN model abstracts each organ on the face as a corresponding expression feature point, and defines different boundary regions for each feature point. The face layout features or feature combinations of the interactive user are taken as the feature input elements by the multi-layer convolution algorithm in the CNN model. The input elements can be a single facial expression or a combination of multiple facial expressions. The effective facial expression features are extracted layer by layer by the multi-layer convolution algorithm. The pooling layer in the CNN model further processes the above expression features by a specific pooling algorithm to form an expression feature map. Finally, the above features are classified by the full connection layer, and the expression feature map is one-to-one mapped with the emotion label, such as smile corresponding to happiness, frown corresponding to anger, etc. The emotion label corresponding to the facial expression is finally output.

[0088] (2) An example of the emotion and body appendage morphology learning model adopting a Transformer model The Transformer model is used to extract and identify the state information of the motion, voice, facial expression, heartbeat and custom of the artificial intelligence toy and the interactive user. Through the camera or other posture sensor built in the artificial intelligence toy, the motion state information, voice state information, facial expression information, heartbeat state information and custom state information of the user or the artificial intelligence toy are collected and captured, and are converted into corresponding sequence data information. The motion state information can be any gesture or body action of the user, such as waving, saying goodbye, shaking hands, waving, lifting the head, nodding, shaking the head, pointing, etc. The collected information can be single action feature, language feature and expression feature, or a combination of a series of action features, language features and expression features of the interactive user. Through the collection and processing of long sequence feature data, the global information or intention of the user is captured, i.e. the coherent behavior of the user is learned through the Transformer model, the corresponding emotion label is associated, and the final emotion label is outputted to drive the change of the fluff state of the artificial intelligence toy as the interactive emotional feedback response. For example, the Transformer model training process specifically includes the following steps: S401, collecting the interactive state data of the user and the artificial intelligence toy through the camera or other posture sensor built in the toy, including motion state information, voice state information and facial expression information, as the input of the Transformer model; S402, preprocessing the interactive state data to extract key feature parameter points; S403, the Transformer model encoder encodes the above key feature parameter points into corresponding feature parameter sequences or sequence combinations; S404, the Transformer model decoder generates the corresponding emotion label according to the emotion response, context and context according to the mapping relationship between the user or the behavior library; S405, the Transformer model collects the feedback information of the user, such as satisfaction or dissatisfaction, interactive behavior, nodding, shaking the head, etc. The model key feature parameters are adjusted through the feedback information. If the feedback is correct, the weight of the model key feature parameters will be increased by 1 in the future, and if the feedback is incorrect, the weight of the model key feature parameters will be reduced by 1 in the future, and the effect and confidence of the reinforcement learning model are adjusted.

[0089] (3) An example of the emotion and body appendage shape learning model using the CGAN model The CGAN model includes a generator and a discriminator provided in the model. The generator can synthesize multiple expressions, languages or actions of a user to enhance the diversity and accuracy of the training data. The discriminator can be used to distinguish the difference between the morphological changes of the artificial intelligence toy and the real pet fur changes. Meanwhile, the discriminator can determine whether the current fur adjustment meets the requirements through user feedback. For example, the specific steps of the CGAN model training process are as follows. S501, collecting the interaction state data of the user and the artificial intelligence toy through the camera, microphone, IMU motion sensor or other posture sensor built in the artificial intelligence toy, including action state information, voice state information and facial expression information, as the input of the CGAN model. S502, preprocessing the interaction state data to extract key feature parameter points, filtering and converting the key feature parameter points to obtain standard normalized data. S503, the generator in the CGAN model selects a corresponding generation model according to the interaction form between the pet and the user. If expression parameters are collected, an expression generation variant is selected. If language parameters are collected, a language generation variant is selected. If action parameters are collected, an action generation variant is selected. The parameter collection and generation variant here can be a single parameter, a continuous parameter or a mixed generation variant of expression, language and action. The generator outputs a corresponding emotion label according to the above input parameters.

[0090] S504, the discriminator in the CGAN model judges whether the fur morphology of the artificial intelligence toy meets the target expectation threshold according to the feedback information of the user. If yes, it indicates that the model selection and calculation are accurate, and the parameter label is increased by 1. If not, it indicates that the model selection or parameter configuration is distorted, and the parameter label is reduced by 1. The model parameters need to be retrained until the fur morphology of the artificial intelligence toy meets the user's requirements.

[0091] The body surface appendage morphology control method of the artificial intelligence toy provided by the embodiment of the present application can collect the target state data of the user and the artificial intelligence toy in real time, identify the changing emotions of the user and the artificial intelligence toy in the interaction process based on the deep learning model, and further control the body surface appendage morphology to change the state according to the emotional changes, realize the multiple changes and morphologies of the body surface appendage morphology of the artificial intelligence toy, feedback the inner world of the artificial intelligence toy through the change of the fur, simulate the interaction between the real pet and the user, give the user a response, meet the emotional needs of the user interaction and the pet state changes, and improve the user interaction experience.

[0092] An example embodiment of the present application provides a body surface appendage morphology control device 100 of an artificial intelligence toy. Figure 12A structural block diagram of the body surface accessory shape control device 100 of the artificial intelligence toy provided by an example embodiment of the present application is shown. The body surface accessory shape control device 100 of the artificial intelligence toy can realize the functions of the body surface accessory shape control method as described above Figures 1 to 11 any of the embodiments. Only the structure and functions of the body surface accessory shape control device 100 of the artificial intelligence toy are briefly described below, and other details can be referred to the related description in the body surface accessory shape control method of the artificial intelligence toy. The embodiments of the body surface accessory shape control device 100 of the artificial intelligence toy correspond to the embodiments of the body surface accessory shape control method of the artificial intelligence toy, and each implementation process and implementation manner of the method embodiments can be applied to the body surface accessory shape control device embodiments, and the same technical effects can be achieved.

[0093] As shown in Figure 12 the body surface accessory shape control device 100 of the artificial intelligence toy includes an information acquisition module 101, a parameter matching module 102, and an intelligent control module 103. In this embodiment, the information acquisition module 101 is configured to acquire target state data, which includes at least one of the following: user interaction state data, user environment state data, artificial intelligence toy interaction state data, and artificial intelligence toy environment state data. The parameter matching module 102 is configured to determine corresponding control reference data according to the target state data, wherein the control reference data is used to indicate the body surface accessory shape corresponding to the target state data. The intelligent control module 103 is configured to control the body surface accessory shape of the artificial intelligence toy based on the control reference data.

[0094] In some embodiments, as Figure 13As shown, the body appendage form control device 100 of the artificial intelligence toy provided in the embodiment of the present application further comprises an emotion and body appendage form learning model module 104. In the embodiment, the intelligent control module 103 is further configured to input the target emotional state data into the pre-constructed emotion and body appendage form learning model module 104, wherein the interaction state data comprises at least one of the following: action state information, voice state information, facial expression information, heartbeat state information and self-defined state information; the output result of the emotion and body appendage form learning model is used as an emotion label corresponding to the target emotional state data, and is output to the parameter matching module 102; wherein the emotion label is used to indicate the emotional state of the user or the artificial intelligence toy; and the parameter matching module 102 is further configured to determine the body appendage form corresponding to the emotion label as the corresponding control reference data according to the pre-set emotion and body appendage form mapping relationship. In the embodiment, by collecting the target emotional state data of the user and the artificial intelligence toy in real time, the corresponding emotion label can be matched, so as to recognize the emotional change of the user or the artificial intelligence toy, and then the body appendage form can be controlled to change the state according to the emotional change.

[0095] In some embodiments, the emotion and body appendage form learning model module 104 is further configured to pre-construct the emotion and body appendage form learning model, comprising: obtaining historical emotional state data, wherein the historical emotional state data comprises at least one of the following: interaction state data of the user in the past preset time, environment state data of the user in the past preset time, interaction state data of the artificial intelligence toy in the past preset time, and environment state data of the artificial intelligence toy in the past preset time; performing data cleaning and feature extraction on the historical emotional state data to obtain a multi-modal feature vector of the historical emotional state data; and using the multi-modal feature vector of the historical emotional state data, the reference weight corresponding to each parameter in the historical emotional state data and the emotion label as training data to learn and train the emotion and body appendage form learning model.

[0096] In some embodiments, the emotion and body appendage form learning model module 104 adjusts the reference weight corresponding to each parameter in the historical emotional state data by: adjusting the test body appendage form of the artificial intelligence toy according to the reference weight corresponding to each parameter in the historical emotional state data, the emotion label and the body appendage form; and adjusting the reference weight corresponding to each parameter in the historical emotional state data by collecting the feedback result of the user on the test body appendage form.

[0097] In some embodiments, the emotion and body appendage form learning model module 104 adjusts the reference weight corresponding to each parameter in the historical emotional state data of the emotion and body appendage form learning model by collecting the feedback result of the user on the body appendage form of the artificial intelligence toy by the information acquisition module 101.

[0098] In this embodiment, the weight values of the parameters are adjusted through closed-loop user feedback to make the model converge, and the model is adjusted through the marking method. The appropriate learning model is trained, and the weight values of the model parameters can be continuously adjusted and optimized through feedback in the real-time interaction process, further improving the effectiveness, accuracy and robustness of the learning model.

[0099] For specific embodiments of the learning model selection and related descriptions of the emotion and body appendage shape learning model, please refer to the related descriptions in the body appendage shape control method of the artificial intelligence toy described above, which will not be repeated here.

[0100] In some embodiments, the types of body appendage shapes include at least one of the following: fixed shape, dynamic shape, and custom shape; as Figure 13 As shown in the figure, the body appendage shape control device 100 of the artificial intelligence toy provided in this embodiment further includes a variable fluff reconstruction module 105 for adjusting the body appendage shape of the artificial intelligence toy through the programmable control unit. In some embodiments, the programmable control unit in the variable fluff reconstruction module 105 includes a programmable electrostatic fluff circuit unit, a programmable magnetic control fluff circuit unit, a programmable electric control fluff circuit unit, and a programmable fluff comb control circuit unit. In this embodiment, the body appendage shape of the artificial intelligence toy can be controlled through the above-mentioned programmable control unit to present various body appendage shapes corresponding to various emotion labels, so as to present the changes and various shapes of the body appendage shape.

[0101] In some embodiments, as Figure 13 As shown in the figure, the body appendage shape control device 100 of the artificial intelligence toy provided in this embodiment further includes a fluff control adjustment module 106, and the intelligent control module 103 is further used to match the corresponding artificial intelligence toy body appendage adjustment mode according to the control reference data. The fluff control adjustment module 106 is used to control the body appendage shape of the artificial intelligence toy based on the artificial intelligence toy body appendage adjustment mode. In this embodiment, the artificial intelligence toy body appendage adjustment mode includes at least one of the following: body appendage electrostatic control adjustment mode, body appendage magnetic control adjustment mode, body appendage electric control adjustment mode, and body appendage comb control adjustment mode. The body appendage electrostatic control adjustment mode is used to display the body appendage shape after electrostatic induction through the electrostatic induction surface or electrostatic coating of the artificial intelligence toy. The body appendage magnetic control adjustment mode is used to display the body appendage shape after magnetic induction through the magnetic induction surface or magnetic attraction coating of the artificial intelligence toy. The body appendage electric control adjustment mode is used to display the body appendage shape after electric control induction through the electric field induction surface or conductive coating of the artificial intelligence toy. The body appendage comb control adjustment mode is used to display the combing shape through the combing structure and circuit of the artificial intelligence toy.

[0102] In some embodiments, the intelligent control module 103 matches the corresponding artificial intelligence toy body surface accessory adjustment mode according to the control reference data by: determining the adjustment mode priority according to the control reference data; and matching the corresponding artificial intelligence toy body surface accessory adjustment mode based on the adjustment mode priority. In this embodiment, for example, the user's emotion during the interaction with the artificial intelligence toy is constantly changing, and therefore the form of the body surface accessory of the artificial intelligence toy also needs to correspond to a plurality of changes and states, or a series of combinations of fur changes. According to different emotion labels, the priority of the above-mentioned four adjustment modes and the combination mode, the order, etc. can be determined, and each is used until the emotional needs of the user interaction and the pet's own state change are met.

[0103] In some embodiments, the intelligent control module 103 matches the corresponding artificial intelligence toy body surface accessory adjustment mode according to the control reference data by: determining the adjustment fine parameter according to the control reference data, the adjustment fine parameter including one of: form adjustment strength, form adjustment area, and form adjustment accuracy; and matching the corresponding artificial intelligence toy body surface accessory adjustment mode based on the adjustment fine parameter. In this embodiment, corresponding to different adjustment modes, in addition to adjusting the output of different forms, the artificial intelligence toy body surface accessory adjustment mode can also be more finely controlled according to the judgment and needs of the adjustment strength, the form adjustment area, and the form adjustment accuracy.

[0104] For example, corresponding to the above-mentioned body surface accessory adjustment mode, as shown in FIG. 6, the fur control adjustment module 106 includes: a static control adjustment module L6, a magnetic control adjustment module L7, an electric control adjustment module L8, and a comb control adjustment module L9, which are respectively adapted to each body surface accessory adjustment mode. Figure 13

[0105] In some embodiments, based on the static electric control adjustment mode of the body surface accessory, the static control adjustment module L6 controls the form of the body surface accessory of the artificial intelligence toy by at least one of the following ways: changing the arrangement direction of the body surface accessory through the change of the static electric field; changing the local form and moving direction of the body surface accessory through the static polarization effect; controlling the distance and the local and overall form of the body surface accessory through the movable static electric controller; and controlling the directional arrangement, angle change, and pattern formation of the body surface accessory through the static electric motion pattern unit to edit the electric field direction, intensity, and area distribution of the body surface accessory.

[0106] In some embodiments, based on the magnetic control adjustment mode of the body surface accessory, the magnetic control adjustment module L7 controls the body surface accessory to arrange according to a preset angle, pattern, and order along the magnetic field direction by adjusting the magnetic field parameter.

[0107] ​In some embodiments, based on the electronic control adjustment mode of the body surface appendage, the electrically controlled adjustment module L8 controls the shape of the body surface appendage of the artificial intelligence toy by at least one of the following ways: forming a corresponding electric field by microelectrodes to control the body surface appendage to arrange along the direction of the electric field line according to a preset angle, pattern and sequence; changing the local shape and moving direction of the body surface appendage by the influence of dielectric force; changing the thickness and length of the body surface appendage by the microelectrical controller.

[0108] In some embodiments, based on the comb control adjustment mode of the body surface appendage, the comb control adjustment module L9 controls the movement of the combing component at the root of the body surface appendage by configuring combing control parameters based on the combing control parameters.

[0109] In this embodiment, the shape of the body surface appendage of the artificial intelligence toy is controlled based on the body surface appendage adjustment mode of the artificial intelligence toy by calling the above adjustment modules. The shape of the body surface appendage of the artificial intelligence toy can be changed in multiple ways according to the changing emotional labels of the user and the artificial intelligence toy in the interaction process, so as to meet the emotional needs of the user interaction and the pet state change.

[0110] For specific embodiments and related descriptions of each adjustment module L6-L9 of the fluff control adjustment module 106 controlling the shape of the body surface appendage of the artificial intelligence toy based on the electronic body surface appendage adjustment mode, please refer to the related descriptions in the above-mentioned method for controlling the shape of the body surface appendage of the artificial intelligence toy. Here, no further description is given.

[0111] The body surface appendage shape control device of the artificial intelligence toy provided by the embodiments of the present application can collect target state data of the user and the artificial intelligence toy in real time, identify the changing emotions of the user and the artificial intelligence toy in the interaction process based on a deep learning model, and then control the body surface appendage shape to change the state according to the emotional changes, realize the multiple changes and shapes of the body surface appendage shape of the artificial intelligence toy, feedback the inner world of the artificial intelligence toy through the changes of the fluff, simulate the interaction between the real pet and the user, give the user a response, meet the emotional needs of the user interaction and the pet state change, and improve the user interaction experience.

[0112] Figure 14 A structural block diagram of an electronic device 1000 is shown, which is shown in an example embodiment of the present application. The electronic device 1000 can be implemented as the body surface appendage shape control device of the artificial intelligence toy described above, which can be configured in the artificial intelligence toy.

[0113] Generally, the electronic device 1000 includes a processor 1001 and a memory 1002.

[0114] The processor 1001 can include one or more processing cores, such as a 4-core processor, a 10-core processor, and the like. The processor 1001 can be implemented in at least one of a hardware form of a DSP (Digital Signal Processing), an FPGA (Field-Programmable Gate Array), a PLA (Programmable Logic Array). The processor 1001 can also include a main processor and a coprocessor, the main processor being a processor for processing data in an awake state, also known as a CPU (Central Processing Unit), and the coprocessor being a low-power processor for processing data in a standby state. In some embodiments, the processor 1001 can be integrated with a GPU (Graphics Processing Unit) for rendering and drawing content required to be displayed by the display screen. In some embodiments, the processor 1001 can also include an AI (Artificial Intelligence) processor for processing machine learning-related computing operations.

[0115] The memory 1002 can include one or more computer-readable storage media that can be non-transitory. The memory 1002 can also include high-speed random access memory, and nonvolatile memory such as one or more disk storage devices, flash storage devices. In some embodiments, the non-transitory computer-readable storage medium in the memory 1002 is used to store at least one instruction for being executed by the processor 1001 to implement all or part of the steps in the body attachment form control method of the artificial intelligence toy shown in the method embodiment of the present application.

[0116] Those skilled in the art can understand that the structure shown in the above description is not a limitation on the electronic device 1000, and can include more or fewer components than those shown, or combine certain components, or use different component arrangements. Figure 14 The structure shown in the above description does not constitute a limitation on the electronic device 1000, and can include more or fewer components than those shown, or combine certain components, or use different component arrangements.

[0117] In an exemplary embodiment, a readable storage medium is also provided, in which a program or instruction is stored, which, when executed by a processor, implements all or part of the steps of the body-attached object shape control method of the artificial intelligence toy described above. For example, the computer readable storage medium can be a Read-Only Memory (ROM), a Random Access Memory (RAM), a Compact Disc Read-Only Memory (CD-ROM), a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0118] In an exemplary embodiment, a computer program product is also provided, which includes a computer program stored on a non-transitory computer readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform all or part of the steps of the body-attached object shape control method of the artificial intelligence toy described above. Figures 1 to 11 All or part of the steps of the body-attached object shape control method of the artificial intelligence toy shown in any embodiment.

[0119] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. It is intended that the application be limited only by the scope of the claims, including all variations and adaptations thereof that are within the spirit and scope of the application. The specification and examples are to be considered exemplary only, with the true scope and spirit of the application indicated by the claims.

[0120] It should be understood that the application is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the claims that follow.

Claims

1. A body attachment shape control method of an artificial intelligence toy, characterized by, The method comprises: acquiring target situational data, the target situational data comprising at least one of the following: interaction state data of a user, environment state data of the user, interaction state data of an artificial intelligence toy, and environment state data of the artificial intelligence toy; determining corresponding control reference data according to the target situational data, wherein the control reference data is used to indicate a body appendage form corresponding to the target situational data; controlling a body appendage form of the artificial intelligence toy based on the control reference data.

2. The method of claim 1, wherein, The determining of the corresponding control reference data according to the target situational data comprises: inputting the target situational data into a pre-constructed emotion and body appendage form learning model, wherein the interaction state data comprises at least one of the following: action state information, voice state information, facial expression information, heartbeat state information, and self-defined state information; taking an output result of the emotion and body appendage form learning model as an emotion label corresponding to the target situational data, wherein the emotion label is used to indicate an emotional state of the user or the artificial intelligence toy; determining a body appendage form corresponding to the emotion label as the corresponding control reference data according to a preset emotion and body appendage form mapping relationship.

3. The method of claim 2, wherein, Before the inputting of the target situational data into the pre-constructed emotion and body appendage form learning model, the method further comprises: acquiring historical situational data, the historical situational data comprising at least one of the following: interaction state data of the user in a past preset time, environment state data of the user in the past preset time, interaction state data of the artificial intelligence toy in the past preset time, and environment state data of the artificial intelligence toy in the past preset time; performing data cleaning and feature extraction on the historical situational data to obtain a multi-modal feature vector of the historical situational data; taking the multi-modal feature vector of the historical situational data, reference weights corresponding to each parameter in the historical situational data, and emotion labels as training data to learn and train the emotion and body appendage form learning model.

4. The method of claim 3, wherein, The method further comprises adjusting the reference weights corresponding to each parameter in the historical situational data in the following manner: adjusting a test body appendage form of the artificial intelligence toy according to the reference weights corresponding to each parameter in the historical situational data, emotion labels, and body appendage forms; adjusting the reference weights corresponding to each parameter in the historical situational data by collecting feedback results of the user on the test body appendage form.

5. The method of claim 3, wherein, The method further comprises: collecting feedback results of the user on the body appendage form of the artificial intelligence toy, and adjusting the reference weights corresponding to each parameter in the historical situational data of the emotion and body appendage form learning model.

6. The method of claim 1, wherein: the type of the body appendage form comprises at least one of the following: a fixed form, a dynamic form, and a self-defined form; before the determining of the corresponding control reference data according to the target situational data, the method further comprises: The body surface accessory form of the artificial intelligence toy is obtained by setting an artificial intelligence toy body surface accessory adjustment mode through a programmable control unit.

7. The method of claim 6, wherein, the body surface accessory form of the artificial intelligence toy is controlled based on the control reference data, including: matching a corresponding artificial intelligence toy body surface accessory adjustment mode according to the control reference data, and controlling the body surface accessory form of the artificial intelligence toy based on the artificial intelligence toy body surface accessory adjustment mode; wherein the artificial intelligence toy body surface accessory adjustment mode includes at least one of: a body surface accessory electrostatic control adjustment mode, a body surface accessory magnetic control adjustment mode, a body surface accessory electric control adjustment mode, and a body surface accessory comb control adjustment mode, wherein: the body surface accessory electrostatic control adjustment mode is used to display the body surface accessory form after electrostatic induction through an electrostatic induction surface or an electrostatic coating of the artificial intelligence toy; the body surface accessory magnetic control adjustment mode is used to display the body surface accessory form after magnetic induction through a magnetic induction surface or a magnetic attraction coating of the artificial intelligence toy; the body surface accessory electric control adjustment mode is used to display the body surface accessory form after electric control induction through an electric field induction surface or a conductive coating of the artificial intelligence toy; the body surface accessory comb control adjustment mode is used to display the combing form through the combing structure and circuit of the artificial intelligence toy.

8. The method of claim 7, wherein, the matching of the corresponding artificial intelligence toy body surface accessory adjustment mode according to the control reference data includes: determining an adjustment mode priority according to the control reference data; matching the corresponding artificial intelligence toy body surface accessory adjustment mode based on the adjustment mode priority.

9. The method of claim 7, wherein, the matching of the corresponding artificial intelligence toy body surface accessory adjustment mode according to the control reference data includes: determining an adjustment fine parameter according to the control reference data, the adjustment fine parameter including one of: form adjustment strength, form adjustment area, and form adjustment accuracy; matching the corresponding artificial intelligence toy body surface accessory adjustment mode based on the adjustment fine parameter.

10. The method of claim 7, wherein, the controlling of the body surface accessory form of the artificial intelligence toy based on the electronic body surface accessory adjustment mode includes at least one of: based on the body surface accessory electrostatic control adjustment mode, the body surface accessory form of the artificial intelligence toy is controlled by at least one of: changing the arrangement direction of the body surface accessory through electrostatic field change; changing the local form and moving direction of the body surface accessory through electrostatic polarization effect; controlling the distance and local and overall form of the body surface accessory through a movable electrostatic controller; editing the electric field direction, intensity, and area distribution of the body surface accessory through an electrostatic motion pattern unit, controlling the directional arrangement, angle change, and pattern formation of the body surface accessory; Based on the magnetic control adjustment mode of the body surface appendage, the body surface appendage is arranged along the magnetic field direction according to a preset angle, pattern and sequence by adjusting the magnetic field parameter; Based on the electric control adjustment mode of the body surface appendage, the body surface appendage of the artificial intelligence toy is controlled in the following ways: the body surface appendage is arranged along the electric field line direction according to a preset angle, pattern and sequence by forming a corresponding electric field through a micro electrode; the local shape and moving direction of the body surface appendage are changed by the influence of dielectric force; the thickness and length of the body surface appendage are changed by a micro electric controller. Based on the comb control adjustment mode of the body surface appendage, different combing shapes are formed by configuring combing control parameters and controlling the movement of the combing components at the root of the body surface appendage based on the combing control parameters.

11. An electronic device, comprising: The electronic device includes a processor and a memory, the memory stores programs or instructions that can be run on the processor, and the programs or instructions are executed by the processor to realize the steps of the body surface appendage shape control method of the artificial intelligence toy according to any one of claims 1 to 10.

12. A readable storage medium, characterized by, The readable storage medium stores programs or instructions, and the programs or instructions are executed by the processor to realize the steps of the body surface appendage shape control method of the artificial intelligence toy according to any one of claims 1 to 10.

13. A computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions that, when executed by a computer, cause the computer to perform the steps of the body surface appendage shape control method of the artificial intelligence toy according to any one of claims 1 to 10.

Images