A bionic robot facial movement debugging method and related equipment

Abstract

The present application provides a method and related equipment for debugging facial movements of a bionic robot, and relates to the field of intelligent robot technology. The present application can automatically adjust the motor movement amplitude of a bionic robot to be debugged by using a manually fine-tuned reference bionic robot as a reference object, ensuring that the reference bionic robot and the adjusted bionic robot to be debugged can directly display the same or similar facial expression effects when performing the same facial movements, thereby effectively reducing the manual adjustment workload and the deviation in the robot's facial movement performance during the robot mass production process, and improving the robot's adjustment efficiency and the consistency of the robot's facial movement performance, thereby meeting the timeliness requirements and facial movement performance consistency requirements during robot mass production.

Description

Technical Field

[0001] The present application relates to the field of intelligent robot technology, and in particular to a bionic robot facial movement debugging method and related equipment. Background Art

[0002] As the aging population becomes more and more mature and sophisticated, people's requirements for robots have also evolved from simple repetitive mechanical movements to the development of bionic robots with high intelligence and the ability to interact with other intelligent entities. People hope that bionic robots can express real situations by controlling facial expressions and communicate naturally with humans.

[0003] Currently, the facial movements of bionic robots rely on the coordination of the robot's bionic facial skin and the numerous tiny motors built into the robot's head. These tiny motors typically need to be driven together to pull the robot's bionic facial skin to perform a specific facial movement. It's worth noting that during the mass production of bionic robots, due to assembly process limitations, different finished products of the same bionic robot have certain differences in head motor installation and / or facial skin installation. As a result, when the same motor drive solution is run on different finished products to perform specific facial movements, the amplitude of the facial movements for that specific facial movement varies significantly, making it impossible to present the same or similar facial expressions to the user.

[0004] Therefore, to ensure that numerous mass-produced robots can achieve the same or similar facial movement amplitudes when performing any facial gesture, it is usually necessary to manually adjust the movement amplitude parameters of each motor involved in different facial movements for each robot before shipment. This ensures that the facial movement amplitudes of each robot, when manually inspected, are consistent. However, this manual inspection and adjustment solution is labor-intensive and inefficient, failing to meet the time constraints of mass production. Furthermore, manual inspection of robot expressions is too subjective, making it difficult to meet the consistency requirements for facial movement performance during mass production. Summary of the Invention

[0005] In view of this, the purpose of the present application is to provide a bionic robot facial movement debugging method, a robot debugging device and a readable storage medium, which can use a manually fine-tuned reference bionic robot as a reference object to automatically adjust the motor movement amplitude of the bionic robot to be debugged, so as to ensure that the reference bionic robot and the adjusted bionic robot to be debugged can directly show the same or similar facial expression effects when performing the same facial movements, thereby effectively reducing the manual adjustment workload and the deviation of the robot's facial movement performance during the mass production of the robot, and improving the robot adjustment efficiency and the consistency of the robot's facial movement performance, so as to meet the timeliness requirements and facial movement performance consistency requirements during the mass production of the robot.

[0006] In order to achieve the above objectives, the technical solutions adopted in the embodiments of the present application are as follows:

[0007] In a first aspect, the present application provides a method for debugging facial movements of a bionic robot, the method comprising:

[0008] Acquire reference facial expression feature data of a reference bionic robot performing a target facial action according to an expected action instruction, and acquire actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction;

[0009] Calculating the difference in facial expression between the bionic robot to be debugged and the reference bionic robot that matches the target facial action based on the reference facial expression feature data and the actual facial expression feature data;

[0010] Detecting whether the facial expression difference satisfies a difference constraint condition;

[0011] If it is detected that the facial expression difference does not satisfy the difference constraint condition, adjusting the maximum motor movement amplitude of the target motor involved in the target facial action at the target facial action in the bionic robot to be debugged, and jumping to the step of obtaining actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction to continue execution;

[0012] If it is detected that the facial expression difference satisfies the difference constraint condition, the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement is used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial movement.

[0013] In an optional embodiment, the expected action instruction includes multiple expected action state weights of the target facial action; and the step of calculating the difference in facial expression matching the target facial action between the bionic robot to be debugged and the reference bionic robot based on the reference facial expression feature data and the actual facial expression feature data comprises:

[0014] Calculating, based on the reference facial expression feature data, a plurality of first facial expression quantization values ​​of the reference bionic robot that match the target facial action, wherein each first facial expression quantization value corresponds to a desired action state weight;

[0015] Calculating, based on the actual facial expression feature data, a plurality of second facial expression quantization values ​​of the bionic robot to be debugged that match the target facial action, wherein each second facial expression quantization value corresponds to a desired action state weight;

[0016] For each expected action state weight, calculating an actual quantized difference between the second facial expression quantized value and the first facial expression quantized value corresponding to the expected action state weight;

[0017] A weighted sum operation is performed on the actual quantized differences of the weights of the multiple expected action states to obtain a difference degree of facial expression that matches the target facial action.

[0018] In an optional embodiment, the step of calculating a plurality of target facial expression quantization values ​​that match the target facial action for any set of target facial expression feature data among the reference facial expression feature data and the actual facial expression feature data comprises:

[0019] Extracting target dimension data whose corresponding dimension matches the target facial action from the target facial expression feature data;

[0020] For each expected action state weight, the target dimension data is screened according to a preset data distribution interval corresponding to the expected action state weight to obtain candidate dimension data that matches the expected action state weight;

[0021] Among the candidate dimensional data that matches the expected action state weight, at least one group of continuous steady-state data is screened out, and the at least one group of screened continuous steady-state data is averaged to obtain a target facial expression quantization value corresponding to the expected action state weight; wherein the target facial expression quantization value is a first facial expression quantization value that matches the reference facial expression feature data, or the target facial expression quantization value is a second facial expression quantization value that matches the actual facial expression feature data; the total number of data frames involved in each group of continuous steady-state data is greater than a preset frame number threshold, and the absolute value of the change between two adjacent frames of data in each group of continuous steady-state data is less than the same change threshold.

[0022] In an optional embodiment, the step of detecting whether the facial expression difference satisfies a difference constraint condition includes:

[0023] Comparing the absolute value of the facial expression difference with a preset difference threshold;

[0024] When the absolute value of the facial expression difference is greater than or equal to the preset difference threshold, determining that the facial expression difference does not satisfy the difference constraint condition;

[0025] When the absolute value of the facial expression difference is less than the preset difference threshold, it is determined that the facial expression difference satisfies the difference constraint condition.

[0026] In an optional embodiment, the method further comprises:

[0027] Obtaining multiple expected action state weights for the target facial action, and determining actual action frame numbers of each of the multiple expected action state weights within a corresponding driving time period based on a unit action driving time and an action driving frame rate of the bionic robot;

[0028] An expression reset time period of a preset number of action frames is reserved before and after the driving time period of each expected action state weight among the multiple expected action state weights to obtain the expected action instruction; wherein the actual action state weight corresponding to the expression reset time period is 0.

[0029] In an optional embodiment, the step of adjusting the maximum motor movement amplitude of the target motor involved in the target facial movement in the bionic robot to be debugged at the target facial movement includes:

[0030] Selecting one target motor from all target motors of the bionic robot to be debugged that are involved in the target facial action as a motor to be adjusted;

[0031] Detecting whether the facial expression difference is positive or negative;

[0032] When it is detected that the facial expression difference is a positive number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is reduced;

[0033] When it is detected that the facial expression difference is a negative number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is increased.

[0034] In an optional embodiment, the step of selecting a target motor as the motor to be adjusted from all target motors of the bionic robot to be debugged that are involved in the target facial action includes:

[0035] In each polling process of all the target motors, each target motor is selected as the motor to be adjusted in descending order of the expression influence of all the target motors for a preset number of times.

[0036] In an optional embodiment, the method further comprises:

[0037] During the process of adjusting the motor movement amplitude of the bionic robot to be debugged related to the target facial movement, detecting in real time whether the total number of motor adjustments of the bionic robot to be debugged for the target facial movement exceeds a preset number threshold;

[0038] When it is detected that the total number of motor adjustments exceeds a preset number threshold, the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement is directly used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial movement.

[0039] In a second aspect, the present application provides a robot debugging device, comprising a processor and a memory, wherein the memory stores a computer program that can be executed by the processor, and the processor can execute the computer program to implement the bionic robot facial movement debugging method described in any one of the aforementioned embodiments.

[0040] In a third aspect, the present application provides a readable storage medium having a computer program stored thereon. When the computer program is executed by a computer device, the bionic robot facial action debugging method described in any one of the aforementioned embodiments is implemented.

[0041] In this case, the beneficial effects of the embodiments of the present application may include the following:

[0042] The present application obtains reference facial expression feature data of a reference bionic robot performing a target facial action according to an expected action instruction, and obtains actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction, and then calculates the facial expression difference between the bionic robot to be debugged and the reference bionic robot that matches the target facial action based on the reference facial expression feature data and the actual facial expression feature data, and then detects whether the facial expression difference satisfies a difference constraint condition, and if the difference constraint condition is not satisfied, adjusts the maximum motor movement amplitude used by the target motor involved in the target facial action in the bionic robot to be debugged when performing the target facial action, and then jumps to the step of obtaining the actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction and continues to execute until the facial expression difference meets the difference constraint condition. When the difference in facial expressions meets the difference constraint condition, the maximum motor movement amplitude actually used by all target motors involved in the target facial action in the bionic robot to be debugged for the target facial action is directly used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial action, so as to automatically adjust the motor movement amplitude of the bionic robot to be debugged with the manually fine-tuned reference bionic robot as a reference, to ensure that the reference bionic robot and the adjusted bionic robot to be debugged can directly show the same or similar facial expression effects when performing the same facial action, so as to effectively reduce the manual adjustment workload and the deviation of the robot's facial action performance during the mass production of the robot, and improve the robot adjustment efficiency and the consistency of the robot's facial action performance, so as to meet the timeliness requirements and facial action performance consistency requirements during the mass production of the robot.

[0043] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 A schematic diagram of the composition of the robot debugging device provided in an embodiment of the present application;

[0046] Figure 2 This is a flow chart of a method for debugging facial movements of a bionic robot provided in an embodiment of the present application;

[0047] Figure 3 for Figure 2A schematic flow chart of the sub-steps included in step S230;

[0048] Figure 4 for Figure 2 A schematic flow chart of the sub-steps included in step S250;

[0049] Figure 5 The second flowchart of the bionic robot facial movement debugging method provided in the embodiment of the present application;

[0050] Figure 6 This is the third flow chart of the bionic robot facial movement debugging method provided in the embodiment of the present application.

[0051] Icons: 10-Robot debugging equipment; 11-Memory; 12-Processor; 13-Communication unit. DETAILED DESCRIPTION

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0053] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for protection, but merely represents selected embodiments of the present application. All other embodiments obtained by persons of ordinary skill in the art based on the embodiments in the present application without creative work are within the scope of protection of the present application.

[0054] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, it does not require further definition or explanation in subsequent drawings.

[0055] In the description of this application, it should be understood that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, or are the orientation or position relationship in which the product of the application is usually placed when in use, or are the orientation or position relationship commonly understood by those skilled in the art. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.

[0056] It should also be noted that, in the description of this application, unless otherwise expressly specified or limited, the terms "disposed," "installed," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0057] In addition, in the description of the present application, it is understood that relational terms such as the terms "first" and "second" are merely used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also include elements inherent to such process, method, article or equipment. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or equipment comprising the elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0058] The following describes some embodiments of the present application in detail with reference to the accompanying drawings. In the absence of conflict, the following embodiments and features in the embodiments may be combined with each other.

[0059] Please refer to Figure 1 , Figure 1: This is a schematic diagram of the composition of the robot debugging device 10 provided in an embodiment of the present application. In this embodiment of the present application, the robot debugging device 10 can be respectively connected to at least one bionic robot to be debugged and a manually fine-tuned reference bionic robot, and is used to drive the at least one bionic robot to be debugged and the reference bionic robot to perform any facial action with the same action state, and based on the facial action performance status of the reference bionic robot as a reference, the head motor of the bionic robot to be debugged related to the facial action is automatically adjusted according to the deviation of the facial action performance between the at least one bionic robot to be debugged and the reference bionic robot, so as to ensure that the reference bionic robot and the finally adjusted bionic robot to be debugged can show the same or similar facial expression effect for the same facial action, thereby effectively reducing the manual adjustment workload and the deviation of the robot's facial action performance during the robot mass production process, and improving the robot adjustment efficiency and the consistency of the robot's facial action performance, so as to meet the timeliness requirements and facial action performance consistency requirements of the robot mass production. The robot debugging device 10 can be a computer device independent of the bionic robot, and the computer device can be, but is not limited to, a personal computer, a cloud server, a laptop computer, a tablet computer, etc.

[0060] In an embodiment of the present application, the reference bionic robot and the at least one bionic robot to be debugged belong to the same bionic robot; the executable facial actions of the reference bionic robot and all bionic robots to be debugged can be the 52 single-dimensional facial actions involved in the BlendShape (BS) expression animation, and each single-dimensional facial action can be characterized by an action state weight with a numerical range of [0,1] to characterize the degree of exaggeration of the action. When the action state weight of a single-dimensional facial action is "0", the actual facial action amplitude of the single-dimensional facial action is 0, and the corresponding bionic robot returns to a normal state of not executing the single-dimensional facial action. When the action state weight of a single-dimensional facial action is "1", the actual facial action amplitude of the single-dimensional facial action reaches the extreme action effect. When any bionic robot performs a one-dimensional facial action according to a certain action state weight, all head motors in the bionic robot involved in the one-dimensional facial action need to move according to a theoretical action amplitude that matches the action state weight and the one-dimensional facial action; wherein, the theoretical action amplitude of each head motor performing the one-dimensional facial action is obtained by multiplying the action state weight by the maximum action amplitude of the corresponding head motor to achieve the one-dimensional facial action.

[0061] In this embodiment, for any one-dimensional facial movement, the number of head motors involved in the one-dimensional facial movement is at least one, and the composition of the head motors involved in different one-dimensional facial movements may partially overlap or may not overlap at all; when the same head motor is shared by multiple one-dimensional facial movements, the maximum movement amplitudes used by the head motor when implementing the multiple one-dimensional facial movements are usually different from each other, and the maximum movement amplitude corresponding to the head motor at any one-dimensional facial movement is less than or equal to the upper limit value of the effective movement amplitude range of the head motor (for example, one-dimensional facial movement A and one-dimensional facial movement B both involve head motor c, and the effective movement amplitude range of head motor c is [0, 110], where the maximum movement amplitude used by head motor c when implementing one-dimensional facial movement A is 100, and the maximum movement amplitude used by head motor c when implementing one-dimensional facial movement B is 20).

[0062] When a head motor is involved in executing multiple single-dimensional facial movements simultaneously, it is necessary to first calculate the theoretical movement amplitudes of each single-dimensional facial movement executed by the head motor individually. Then, motor movements are superimposed based on the motor movement directions and theoretical movement amplitudes corresponding to the multiple single-dimensional facial movements, and the theoretical superimposed movement amplitudes required for the head motor to simultaneously achieve the multiple single-dimensional facial movements are calculated. Then, by comparing the theoretical superimposed movement amplitude with the effective movement amplitude range of the head motor, the effective superimposed movement amplitude required for the head motor to simultaneously achieve the multiple single-dimensional facial movements is determined. When the corresponding theoretical superimposed movement amplitude is within the effective movement amplitude range of the head motor, the effective superimposed movement amplitude of the head motor is the corresponding theoretical superimposed movement amplitude; when the corresponding theoretical superimposed movement amplitude exceeds the effective movement amplitude range of the head motor, the effective superimposed movement amplitude of the head motor is the amplitude range boundary value of the corresponding effective movement amplitude range that is close to the theoretical superimposed movement amplitude.

[0063] In an embodiment of the present application, the robot debugging device 10 may include a memory 11, a processor 12, and a communication unit 13. The memory 11, the processor 12, and the communication unit 13 are electrically connected to each other, directly or indirectly, to enable data transmission or interaction. For example, the memory 11, the processor 12, and the communication unit 13 may be electrically connected to each other via one or more communication buses or signal lines.

[0064] In the embodiment of the present application, the memory 11 is used to store a computer program, and the processor 12 can execute the computer program accordingly after receiving an execution instruction.

[0065] In an embodiment of the present application, the communication unit 13 is used to establish a communication connection between the robot debugging device 10 and other electronic devices via a network, and to send and receive data via the network, wherein the network includes a wired communication network and a wireless communication network. For example, the robot debugging device 10 can be connected to a facial motion capture device via the communication unit 13 to obtain the facial action performance (i.e., facial expression feature data) of any bionic robot performing a certain facial action through the facial motion capture device, wherein the facial action performance can be expressed using BlendShape expression data.

[0066] In an embodiment of the present application, the robot debugging device 10 may pre-store a specific computer program related to the bionic robot facial movement debugging function in the memory 11, and by driving the processor 12 to execute the specific computer program stored in the memory 11, the manually fine-tuned reference bionic robot is used as a reference object to automatically adjust the motor movement amplitude of the bionic robot to be debugged, thereby ensuring that the reference bionic robot and the adjusted bionic robot to be debugged can directly show the same or similar facial expression effects when performing the same facial movement, thereby effectively reducing the manual adjustment workload and the deviation of the robot's facial movement performance during the mass production of the robot, and improving the robot adjustment efficiency and the consistency of the robot's facial movement performance, so as to meet the timeliness requirements and facial movement performance consistency requirements during the mass production of the robot.

[0067] It is understandable that Figure 1 The block diagram shown is only a schematic diagram of a composition of the robot debugging device 10. The robot debugging device 10 may also include Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown. Figure 1 Each component shown in the figure can be implemented by hardware, software or a combination thereof.

[0068] In this application, to ensure that the robot debugging device 10 can automatically adjust the motor movement amplitude of the bionic robot to be debugged using a manually fine-tuned reference bionic robot as a reference, thereby meeting the timeliness requirements and facial movement consistency requirements for robot mass production, the embodiment of this application provides a bionic robot facial movement debugging method to achieve the aforementioned objectives. The bionic robot facial movement debugging method provided in this application is described in detail below.

[0069] Please refer to Figure 2 , Figure 2 This is one of the flow charts of the bionic robot facial action debugging method provided in the embodiment of the present application. In the embodiment of the present application, Figure 2The bionic robot facial action debugging method shown may include steps S210 to S260.

[0070] Step S210 , obtaining reference facial expression feature data of a reference bionic robot performing a target facial action according to a desired action instruction.

[0071] In this embodiment, the expected action instruction is used to instruct the corresponding bionic robot to perform a target facial action, wherein the expected action instruction includes a plurality of expected action state weights of the target facial action (for example, three expected action state weights of 0.3, 0.6 and 0.9), and a driving time period corresponding to each of the plurality of expected action state weights, wherein the plurality of expected action state weights are different from each other and are all greater than 0; the target facial action can be any one of the 52 single-dimensional facial actions involved in the BlendShape expression animation. In one implementation of this embodiment, the driving time periods of the respective plurality of expected action state weights in the expected action instruction are arranged in ascending order according to the weight values ​​of the plurality of expected action state weights, so as to ensure that when the same bionic robot performs target facial actions with different expected action state weights, there will be no damage to the robot hardware due to drastic changes in the action amplitude in a short period of time.

[0072] After the robot debugging device 10 determines a matching expected action instruction for the target facial action, it can send the expected action instruction to the manually fine-tuned reference bionic robot, so that the reference bionic robot calls all head motors involved in the target facial action and uses the fine-tuned limit action amplitude that matches the target facial action (i.e., the limit action amplitude used by the corresponding head motor when achieving the target facial action under manual fine-tuning) to execute the expected action instruction, thereby achieving the target facial action execution effect in different expected action states. While the reference bionic robot is executing the expected action instruction, the robot debugging device 10 can capture the facial action performance of the reference bionic robot in real time through a facial action capture device, and obtain reference facial expression feature data of the reference bionic robot executing the expected action instruction.

[0073] It can be understood that in order to avoid any bionic robot being interfered with by the robot hardware position of the realized action state when executing the target facial action of different expected action states, an expression reset time period can be reserved before and after the driving time period of each of the multiple expected action state weights in the expected action instruction. The actual action state weight of the target facial action in the expression reset time period is 0, so as to ensure that the corresponding bionic robot is in a normal state before and after executing the target facial action of any expected action state, thereby improving the data reliability of the corresponding facial expression feature data.

[0074] Step S220 , obtaining actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction.

[0075] In this embodiment, after determining a matching desired action instruction for the target facial action, the robot debugging device 10 can send the desired action instruction to any mass-produced bionic robot to be debugged, causing the bionic robot to drive all head motors (i.e., target motors) involved in the target facial action to execute the desired action instruction using the maximum motor action amplitude configured for the target facial action as the current limit action amplitude, thereby achieving the target facial action execution effect in different desired action states. During the process of the bionic robot to be debugged executing the desired action instruction, the robot debugging device 10 can use a facial motion capture device to capture the facial action performance of the bionic robot to be debugged in real time, and obtain the actual facial expression feature data of the bionic robot to be debugged executing the desired action instruction. The facial motion capture device used by each bionic robot is consistent with the relative position relationship between the corresponding bionic robot's face; for all target motors involved in the target facial action in any bionic robot to be debugged, the limit action amplitude (i.e., the maximum motor action amplitude) actually used by each target motor when first executing the desired action instruction is consistent with the fine-tuning limit action amplitude of the target motor to achieve the target facial action on the reference bionic robot.

[0076] Step S230 , calculating the difference in facial expression that matches the target facial action between the bionic robot to be debugged and the reference bionic robot based on the reference facial expression feature data and the actual facial expression feature data.

[0077] In an embodiment of the present application, the robot debugging device 10 can calculate the facial expression difference of the bionic robot to be debugged relative to the reference bionic robot at the target facial action execution level based on the actual facial expression feature data of the bionic robot to be debugged, by taking the reference facial expression feature data of the reference bionic robot as a reference, so as to characterize the facial action performance deviation between the bionic robot to be debugged and the reference bionic robot regarding the target facial action through the facial expression difference.

[0078] Alternatively, see Figure 3 , Figure 3 yes Figure 2 In the embodiment of the present application, step S230 may include sub-steps S231 to S234 to accurately measure the difference in facial expression between the target facial action execution level of the bionic robot to be debugged and the reference bionic robot.

[0079] Sub-step S231, calculating a plurality of first facial expression quantization values ​​of the reference bionic robot that match the target facial action based on the reference facial expression feature data, wherein each first facial expression quantization value corresponds to a desired action state weight.

[0080] In this embodiment, when the facial expression feature data collected by the facial motion capture device belongs to 52-dimensional BlendShape expression data, each dimension of expression data corresponds to a single-dimensional facial action, the sub-step S231 may include:

[0081] Extracting target dimension data whose corresponding dimension matches the target facial action from the reference facial expression feature data;

[0082] For each expected action state weight, the target dimension data is screened according to a preset data distribution interval corresponding to the expected action state weight to obtain candidate dimension data that matches the expected action state weight;

[0083] Among the candidate dimensional data that matches the expected action state weight, at least one group of continuous steady-state data is screened out, and a mean operation is performed on the at least one group of continuous steady-state data screened out to obtain a first facial expression quantization value corresponding to the expected action state weight.

[0084] In this process, different expected action state weights have different preset data distribution intervals corresponding to the target facial action at the facial expression feature data; the total number of data frames of at least one group of continuous steady-state data corresponding to the same expected action state weight is greater than the preset frame number threshold, and the absolute value of the change between two adjacent frames of data in the at least one group of continuous steady-state data is less than the same change threshold; wherein the change threshold is greater than 0, and since the action driving frame rate of the bionic robot is fixed, the total number of data frames can be used to characterize the steady-state duration of the corresponding continuous steady-state data, and the preset frame number threshold can be used to characterize the preset duration threshold, and the above-mentioned continuous steady-state data screening operation is used to eliminate error data caused by robot hardware action delay; in the continuous steady-state data screening process for any expected action state weight, the change threshold can be first set according to a smaller value to screen the continuous steady-state data, and then when any group of continuous steady-state data is not screened out, the continuous steady-state data can be screened again by appropriately increasing the change threshold until at least one group of continuous steady-state data is screened out.

[0085] Sub-step S232, calculating a plurality of second facial expression quantization values ​​that match the target facial action of the bionic robot to be debugged based on the actual facial expression feature data, wherein each second facial expression quantization value corresponds to a desired action state weight.

[0086] In this embodiment, when the facial expression feature data collected by the facial motion capture device belongs to 52-dimensional BlendShape expression data, each dimension of expression data corresponds to a single-dimensional facial action, the sub-step S232 may include:

[0087] Extracting target dimension data whose corresponding dimension matches the target facial action from the actual facial expression feature data;

[0088] For each expected action state weight, the target dimension data is screened according to a preset data distribution interval corresponding to the expected action state weight to obtain candidate dimension data that matches the expected action state weight;

[0089] Among the candidate dimensional data that matches the expected action state weight, at least one group of continuous steady-state data is screened out, and a mean operation is performed on the at least one group of continuous steady-state data screened out to obtain a second facial expression quantization value corresponding to the expected action state weight.

[0090] In this process, the preset data distribution intervals corresponding to the target facial action at the facial expression feature data of different expected action state weights are different from each other; the total number of data frames of at least one group of continuous steady-state data corresponding to the same expected action state weight is greater than the preset frame number threshold, and the absolute value of the change between two adjacent frames of data in the at least one group of continuous steady-state data is less than the same change threshold.

[0091] Sub-step S233 : For each expected action state weight, calculating the actual quantized difference between the second facial expression quantized value and the first facial expression quantized value corresponding to the expected action state weight.

[0092] In this embodiment, for each expected action state weight, the actual quantization difference matching the expected action state weight can be obtained by subtracting the corresponding first facial expression quantization value from the second facial expression quantization value corresponding to the expected action state weight.

[0093] Sub-step S234 , performing a weighted sum operation on the actual quantized differences of the weights of the multiple desired action states to obtain a difference degree of facial expression that matches the target facial action.

[0094] Therefore, the present application can accurately measure the difference in facial expression of the bionic robot to be debugged relative to the reference bionic robot at the target facial action execution level by executing the above sub-steps S231 to S234.

[0095] Step S240 , detecting whether the facial expression difference satisfies the difference constraint condition.

[0096] In this embodiment, the difference constraint is used to characterize the distribution of facial expression differences between the bionic robot to be debugged and the reference bionic robot when they exhibit the same or similar facial expression effects during the execution of the same facial action. When the robot debugging device 10 determines the difference in facial expression between the bionic robot to be debugged and the reference bionic robot in response to the desired action instruction, it performs parameter matching on the facial expression difference with the difference constraint to determine whether the bionic robot to be debugged currently requires automatic adjustment of the motor movement amplitude.

[0097] Among them, when it is detected that the facial expression difference meets the difference constraint condition, it indicates that the facial expression effects displayed by the bionic robot to be debugged and the reference bionic robot during the execution of the target facial action are already very close, and the bionic robot to be debugged currently does not need to perform automatic adjustment of the motor movement amplitude. At this time, the robot debugging device 10 will correspondingly execute step S260; when it is detected that the facial expression difference does not meet the difference constraint condition, it indicates that the facial expression effects displayed by the bionic robot to be debugged and the reference bionic robot during the execution of the target facial action still have obvious deviations, and the bionic robot to be debugged currently needs to perform automatic adjustment of the motor movement amplitude. At this time, the robot debugging device 10 will correspondingly execute step S250.

[0098] In one implementation of this embodiment, step S240 may include:

[0099] Comparing the absolute value of the facial expression difference with a preset difference threshold;

[0100] When the absolute value of the facial expression difference is greater than or equal to the preset difference threshold, determining that the facial expression difference does not satisfy the difference constraint condition;

[0101] When the absolute value of the facial expression difference is less than the preset difference threshold, it is determined that the facial expression difference satisfies the difference constraint condition.

[0102] The preset difference threshold is a positive number, and its value can be flexibly configured according to the robot debugger's debugging accuracy requirements. The higher the debugging accuracy requirements, the closer the preset difference threshold is to 0.

[0103] Step S250 , adjusting the maximum motor movement amplitude of the target motor involved in the target facial movement in the bionic robot to be debugged at the target facial movement.

[0104] In this embodiment, when it is determined that the bionic robot to be debugged currently needs to automatically adjust the motor movement amplitude, the robot debugging device 10 will select a target motor from all target motors involved in the target facial movement in the bionic robot to be debugged as the motor to be adjusted, and then adjust the maximum motor movement amplitude (i.e., the current limit movement amplitude) currently actually used by the motor to be adjusted for the target facial movement to complete a motor movement amplitude adjustment operation related to the target facial movement of the bionic robot to be debugged, and then drive the adjusted bionic robot to be debugged to re-execute the target facial movement according to the expected movement instruction. At this time, the robot debugging device 10 will jump to the above-mentioned step S220 to continue execution until it is detected that the corresponding facial expression difference meets the difference constraint condition.

[0105] Alternatively, see Figure 4 , Figure 4 yes Figure 2 In the embodiment of the present application, step S250 may include sub-steps S251 to S254 to effectively reduce the deviation in facial expression effects between the bionic robot to be debugged and the reference bionic robot during the execution of the target facial action.

[0106] Sub-step S251 : selecting a target motor from all target motors involved in the target facial movement of the bionic robot to be debugged as a motor to be adjusted.

[0107] In this embodiment, when the number of target motors involved in the target facial action of the bionic robot to be debugged is 1, a single target motor involved in the target facial action of the bionic robot to be debugged can be directly used as the motor to be adjusted; when the number of target motors involved in the target facial action of the bionic robot to be debugged is multiple, one target motor can be selected from the multiple target motors as the motor to be adjusted according to a specific strategy (for example, cyclically using the multiple target motors in a specific order as the motors to be adjusted) or randomly.

[0108] In one implementation of this embodiment, when a single-dimensional facial movement needs to rely on the cooperation of multiple head motors to be realized, the multiple head motors can be sorted according to factors such as their respective installation postures and motor models to determine their respective impacts on facial expressions of the single-dimensional facial movement, thereby obtaining a descending order of the impact of the multiple head motors on facial expressions of the single-dimensional facial movement, wherein the head motor that is closer to the front in the descending order of the impact has a stronger influence on the facial expression of the single-dimensional facial movement.

[0109] On this basis, to improve the efficiency of adjusting the amplitude of facial movements of the bionic robot to be debugged, the robot debugging device 10 may perform at least one polling operation on all target motors involved in the target facial movement of the bionic robot to be debugged, and in each polling operation, sequentially select each target motor as the motor to be adjusted for a predetermined number of consecutive times according to the descending order of the expression influence of all target motors. For example, if all target motors involved in a target facial movement include motor G1, motor G2, and motor G3, and the descending order of the expression influence of these three target motors is "motor G3 -> motor G1 -> motor G2," and assuming that the predetermined number of selections for these three target motors for the target facial movement is 3, then in a single polling operation, motor G3 must first be selected as the motor to be adjusted for three consecutive times before motor G1 can be selected for three consecutive times, and then motor G2 can be selected for three consecutive times, thereby completing one polling operation.

[0110] Sub-step S252 , detecting whether the facial expression difference is positive or negative.

[0111] In sub-step S253 , when it is detected that the facial expression difference is a positive number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is reduced.

[0112] In this embodiment, when it is detected that the current facial expression difference of the bionic robot to be debugged with respect to the target facial action is a positive number, the facial movement amplitude of the bionic robot to be debugged for the target facial action is larger than that of the reference bionic robot. At this time, it is necessary to reduce the maximum motor movement amplitude (i.e., the current limit movement amplitude) currently actually used by the motor to be adjusted under the target facial action, so that the subsequent facial movement amplitude of the bionic robot to be debugged for the target facial action is as close as possible to that of the reference bionic robot.

[0113] In this embodiment, the magnitude of the single motor movement amplitude adjustment of the motor to be adjusted for the target facial movement can be calculated by multiplying a preset adjustment amplitude coefficient (e.g., 0.05) by the fine-tuning limit movement amplitude of the motor to be adjusted for the target facial movement on the reference bionic robot. Each single motor movement amplitude adjustment operation involved in the same polling operation can be implemented using the same preset adjustment amplitude coefficient. The preset adjustment amplitude coefficients used in different polling operations can be partially identical or different. In one implementation of this embodiment, the preset adjustment amplitude coefficients used in two adjacent polling operations change in a decreasing trend.

[0114] In sub-step S254 , when it is detected that the facial expression difference is a negative number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is increased.

[0115] In this embodiment, when it is detected that the current facial expression difference of the bionic robot to be debugged with respect to the target facial action is a negative number, the facial movement amplitude of the bionic robot to be debugged for the target facial action is smaller than that of the reference bionic robot. At this time, it is necessary to increase the maximum motor movement amplitude (i.e., the current limit movement amplitude) currently actually used by the motor to be adjusted under the target facial action, so that the subsequent facial movement amplitude of the bionic robot to be debugged for the target facial action is as close as possible to that of the reference bionic robot.

[0116] Therefore, the present application can effectively reduce the deviation in facial expression effects displayed by the bionic robot to be debugged and the reference bionic robot during the execution of the target facial action by executing the above sub-steps S251 to S254.

[0117] Step S260 , taking the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement as the expected limit movement amplitude for the corresponding target motor to achieve the target facial movement.

[0118] In this embodiment, when it is detected that the current facial expression difference of the bionic robot to be debugged with respect to the target facial action satisfies the difference constraint, it indicates that the maximum motor movement amplitude (i.e., the current limit movement amplitude) currently actually used by all target motors in the bionic robot to be debugged for the target facial action can ensure that the reference bionic robot and the bionic robot to be debugged exhibit the same or similar facial expression effects when executing the target facial action. At this time, the maximum motor movement amplitude currently actually used by all target motors for the target facial action can be used as the expected limit movement amplitude of the corresponding target motor for the target facial action that can meet the facial action performance consistency requirement, thereby completing the facial action performance debugging operation of the bionic robot to be debugged and the reference bionic robot for the target facial action.

[0119] Therefore, the present application can execute the above steps S210 to S2560, use the manually fine-tuned reference bionic robot as a reference object, and automatically adjust the motor movement amplitude of the bionic robot to be debugged, so as to ensure that the reference bionic robot and the adjusted bionic robot to be debugged can directly show the same or similar facial expression effects when performing the same facial movements, thereby effectively reducing the manual adjustment workload and the deviation of the robot's facial movement performance during the mass production of the robot, and improving the robot adjustment efficiency and the consistency of the robot's facial movement performance, so as to meet the timeliness requirements and facial movement performance consistency requirements during the mass production of the robot.

[0120] Alternatively, see Figure 5 , Figure 5 This is the second flow chart of the bionic robot facial action debugging method provided in the embodiment of the present application. Figure 2 Compared with the bionic robot facial action debugging method shown in the figure, Figure 5 The bionic robot facial movement debugging method shown may further include steps S270 to S280 to effectively avoid endlessly adjusting the motor movement amplitude of any bionic robot to be debugged, thereby ensuring the timeliness of robot mass production.

[0121] Step S270 , during the process of adjusting the motor movement amplitude related to the target facial movement of the bionic robot to be debugged, detecting in real time whether the total number of motor adjustments of the bionic robot to be debugged for the target facial movement exceeds a preset number threshold.

[0122] In this embodiment, different single-dimensional facial movements correspond to different preset number thresholds; during the process of adjusting the motor movement amplitude of the bionic robot to be debugged for the target facial movement, the robot debugging device 10 will detect in real time whether the total number of motor adjustments of the bionic robot to be debugged for the target facial movement exceeds the preset number threshold corresponding to the target facial movement, to determine whether it is necessary to continue the motor movement amplitude adjustment operation for the target facial movement. Among them, when the total number of motor adjustments for the target facial action does not exceed the preset number threshold corresponding to the target facial action, it indicates that it is necessary to continue to adjust the motor movement amplitude of the target facial action, and the robot debugging device 10 will continue to execute the above steps S220 to S260; when the total number of motor adjustments for the target facial action exceeds the preset number threshold corresponding to the target facial action, it indicates that it is not necessary to continue to adjust the motor movement amplitude of the target facial action, even if the current facial expression difference of the bionic robot to be debugged with respect to the target facial action does not meet the difference constraint condition, the robot debugging device 10 will also suspend the execution of the above steps S220 to S260 and execute step S280 to avoid endlessly adjusting the motor movement amplitude of the bionic robot to be debugged, thereby effectively ensuring the timeliness of robot mass production.

[0123] During this process, the preset number thresholds corresponding to different unidimensional facial movements can be flexibly configured by the robot debugger based on the timeliness requirements of robot mass production. In one implementation of this embodiment, for any unidimensional facial movement, the preset number threshold for that unidimensional facial movement can be obtained by multiplying the number of head motors corresponding to that unidimensional facial movement, the preset number of selections, and the preset number of pollings. For example, if the number of head motors involved in a target facial movement is 3, the preset number of selections for these three head motors under the target facial movement is 3, and the preset number of pollings for these three head motors is 5, then the preset number threshold for the target facial movement is 3*3*5=45.

[0124] Step S280, when it is detected that the total number of motor adjustments exceeds the preset number threshold, the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement is directly used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial movement.

[0125] In this embodiment, when it is detected that the total number of motor adjustments of the bionic robot to be debugged for the target facial action exceeds a preset number threshold corresponding to the target facial action, the maximum motor movement amplitudes (i.e., current limit movement amplitudes) currently actually used by all target motors involved in the target facial action in the bionic robot to be debugged under the target facial action can be used as the expected limit movement amplitudes for achieving the target facial action, thereby completing the facial movement performance debugging operation of the bionic robot to be debugged and the reference bionic robot for the target facial action.

[0126] Therefore, the present application can effectively avoid endless and unnecessary motor motion amplitude adjustment operations in the automatic adjustment of the motor motion amplitude of the bionic robot to be debugged by executing the above steps S210 to S280, thereby ensuring the timeliness of robot mass production.

[0127] Alternatively, see Figure 6 , Figure 6 This is the third flow chart of the bionic robot facial action debugging method provided in the embodiment of the present application. Figure 2 or Figure 5 Compared with the bionic robot facial action debugging method shown in the figure, Figure 6 The bionic robot facial action debugging method shown can also include steps S290 and S2100 to avoid the corresponding bionic robot being disturbed by the robot hardware position of the realized action state in the process of driving any bionic robot to perform target facial actions in different expected action states, thereby improving the data reliability of the corresponding facial expression feature data.

[0128] Step S290: Obtain multiple expected action state weights for the target facial action, and determine the actual action frame numbers of each of the multiple expected action state weights within the corresponding driving time period based on the unit action driving time and action driving frame rate of the bionic robot.

[0129] In this embodiment, the actual number of action frames of each expected action state weight in the corresponding driving time period is obtained by multiplying the unit action driving time and the action driving frame rate. For example, the unit action driving time of the bionic robot is 2 seconds, and the action driving frame rate of the bionic robot is 30FPS. Then the actual number of action frames of each expected action state weight in the corresponding driving time period is 2*30=60, and the effective duration of the driving time period of each expected action state weight is 2 seconds. In one implementation of this embodiment, the driving time periods of the various expected action state weights are arranged in ascending order according to the weight values ​​of the various expected action state weights.

[0130] Step S2100 , reserving an expression reset time period of a preset number of action frames before and after a driving time period of each of a plurality of expected action state weights to obtain an expected action instruction.

[0131] In this embodiment, the expression reset time period has a weight of 0 for the actual action state of the target facial action. This ensures that the corresponding bionic robot is in a normal state before and after performing any desired action state of the target facial action, thereby improving the data reliability of the corresponding facial expression feature data. In one implementation of this embodiment, the preset action frame number is consistent with the action drive frame rate, so that the effective duration of a single expression reset time period is 1 second.

[0132] It is understandable that when the reserved operation of all expression reset time periods is completed for the multiple expected action state weights, the drive signal corresponding to the target facial action is a sawtooth wave signal. At this time, the drive signal can be smoothed and filtered to obtain a target drive signal with a relatively smooth change, approximately a wave-shaped curve, so as to use the target drive signal to represent the expected action instruction, thereby ensuring that when the bionic robot executes target facial actions with different expected action state weights, there will be no robot hardware damage caused by drastic changes in the action amplitude in a short period of time. Among them, the aforementioned smoothing filter processing operation needs to ensure that the actual action state weight and the corresponding expected action state weight are consistent for at least a period of time in each drive time period.

[0133] Therefore, the present application can avoid the corresponding bionic robot being disturbed by the robot hardware position of the realized action state by executing the above steps S290 and S2100, while driving any bionic robot to perform target facial actions in different expected action states, thereby improving the data reliability of the corresponding facial expression feature data.

[0134] In addition, the present application also provides a readable storage medium having a computer program stored thereon, which can be used to execute a computer device (e.g., a personal computer, a server, or a network device) acting as the robot debugging device 10 to implement any possible bionic robot facial action debugging method disclosed in the above method embodiment. The readable storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0135] The above are merely various embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. A bionic robot facial action debugging method, characterized in that: The method comprises: Acquiring reference facial expression feature data of a reference bionic robot performing a target facial action according to an expected action instruction, and acquiring actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction, wherein the expected action instruction includes multiple expected action state weights of the target facial action; Calculating the difference in facial expression between the bionic robot to be debugged and the reference bionic robot that matches the target facial action based on the reference facial expression feature data and the actual facial expression feature data; Detecting whether the facial expression difference satisfies a difference constraint condition; If it is detected that the facial expression difference does not satisfy the difference constraint condition, adjusting the maximum motor movement amplitude of the target motor involved in the target facial action at the target facial action in the bionic robot to be debugged, and jumping to the step of obtaining actual facial expression feature data of the bionic robot to be debugged performing the target facial action according to the expected action instruction to continue execution; If it is detected that the facial expression difference satisfies the difference constraint condition, the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement is used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial movement.

2. The method according to claim 1, characterized in that The step of calculating the difference in facial expressions matching the target facial action between the bionic robot to be debugged and the reference bionic robot based on the reference facial expression feature data and the actual facial expression feature data comprises: Calculating, based on the reference facial expression feature data, a plurality of first facial expression quantization values ​​of the reference bionic robot that match the target facial action, wherein each first facial expression quantization value corresponds to a desired action state weight; Calculating, based on the actual facial expression feature data, a plurality of second facial expression quantization values ​​of the bionic robot to be debugged that match the target facial action, wherein each second facial expression quantization value corresponds to a desired action state weight; For each expected action state weight, calculating an actual quantized difference between the second facial expression quantized value and the first facial expression quantized value corresponding to the expected action state weight; A weighted sum operation is performed on the actual quantized differences of the weights of the multiple expected action states to obtain a difference degree of facial expression that matches the target facial action.

3. The method according to claim 2, characterized in that The step of calculating a plurality of first facial expression quantization values ​​of the reference bionic robot that match the target facial action based on the reference facial expression feature data comprises: Extracting target dimension data whose corresponding dimension matches the target facial action from the reference facial expression feature data; For each expected action state weight, the target dimension data is screened according to a preset data distribution interval corresponding to the expected action state weight to obtain candidate dimension data that matches the expected action state weight; At least one set of continuous steady-state data is screened out from candidate dimensional data that matches the expected action state weight, and a mean operation is performed on the screened at least one set of continuous steady-state data to obtain a first facial expression quantization value corresponding to the expected action state weight; wherein the total number of frames of data involved in each set of continuous steady-state data is greater than a preset frame number threshold, and the absolute value of the change between two adjacent frames of data in each set of continuous steady-state data is less than the same change threshold; The step of calculating a plurality of second facial expression quantization values ​​of the bionic robot to be debugged that match the target facial action based on the actual facial expression feature data comprises: Extracting target dimension data whose corresponding dimension matches the target facial action from the actual facial expression feature data; For each expected action state weight, the target dimension data is screened according to a preset data distribution interval corresponding to the expected action state weight to obtain candidate dimension data that matches the expected action state weight; Among the candidate dimensional data that matches the expected action state weight, at least one group of continuous steady-state data is screened out, and a mean operation is performed on the at least one group of screened continuous steady-state data to obtain a second facial expression quantization value corresponding to the expected action state weight; wherein the total number of data frames involved in each group of continuous steady-state data is greater than a preset frame number threshold, and the absolute value of the change between two adjacent frames of data in each group of continuous steady-state data is less than the same change threshold.

4. The method according to claim 1, wherein The step of detecting whether the facial expression difference satisfies the difference constraint condition comprises: Comparing the absolute value of the facial expression difference with a preset difference threshold; When the absolute value of the facial expression difference is greater than or equal to the preset difference threshold, determining that the facial expression difference does not satisfy the difference constraint condition; When the absolute value of the facial expression difference is less than the preset difference threshold, it is determined that the facial expression difference satisfies the difference constraint condition.

5. The method according to claim 1, wherein The method further comprises: Obtaining multiple expected action state weights for the target facial action, and determining actual action frame numbers of each of the multiple expected action state weights within a corresponding driving time period based on a unit action driving time and an action driving frame rate of the bionic robot; An expression reset time period of a preset number of action frames is reserved before and after the driving time period of each expected action state weight among the multiple expected action state weights to obtain the expected action instruction; wherein the actual action state weight corresponding to the expression reset time period is 0.

6. The method according to any one of claims 1 to 5, characterized in that The step of adjusting the maximum motor movement amplitude of the target motor involved in the target facial movement in the bionic robot to be debugged at the target facial movement comprises: Selecting one target motor from all target motors of the bionic robot to be debugged that are involved in the target facial action as a motor to be adjusted; Detecting whether the facial expression difference is positive or negative; When it is detected that the facial expression difference is a positive number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is reduced; When it is detected that the facial expression difference is a negative number, the maximum motor movement amplitude actually used when the motor to be adjusted performs the target facial movement is increased.

7. The method according to claim 6, characterized in that The step of selecting a target motor as a motor to be adjusted from all target motors of the bionic robot to be debugged that are involved in the target facial action comprises: In each polling process of all the target motors, each target motor is selected as the motor to be adjusted in descending order of the expression influence of all the target motors for a preset number of times.

8. The method according to claim 6, characterized in that The method further comprises: During the process of adjusting the motor movement amplitude of the bionic robot to be debugged related to the target facial movement, detecting in real time whether the total number of motor adjustments of the bionic robot to be debugged for the target facial movement exceeds a preset number threshold; When it is detected that the total number of motor adjustments exceeds a preset number threshold, the maximum motor movement amplitude actually used by all target motors in the bionic robot to be debugged for the target facial movement is directly used as the expected limit movement amplitude of the corresponding target motor to achieve the target facial movement.

9. A robot debugging device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program that can be executed by the processor, and the processor can execute the computer program to implement the bionic robot facial action debugging method according to any one of claims 1 to 8.

10. A readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a computer device, the bionic robot facial action debugging method according to any one of claims 1 to 8 is implemented.

Images