Optimization method of interactive living body detection scheme and related equipment

Abstract

The invention discloses an optimization method of an interactive living body detection scheme and related equipment, and relates to the technical field of biological recognition, and the method comprises the steps: obtaining a to-be-optimized living body detection scheme, and a detection task set and a detection model set of the to-be-optimized living body detection scheme; based on the detection model set, performing task error rate detection on the detection task set to obtain a detection error rate of the to-be-optimized living body detection scheme; performing time index calculation on the detection task set and a preset task time index to obtain a detection time cost of the to-be-optimized living body detection scheme; performing difficulty index calculation on the detection task set and a preset task difficulty index to obtain an action difficulty value of the to-be-optimized living body detection scheme; and based on the detection error rate, the detection time cost and the action difficulty, through a preset multi-objective optimization algorithm, selecting an optimal living body detection scheme from the living body detection schemes to be optimized. According to the method and the device, the consumed time and the action execution difficulty are reduced on the basis of ensuring the face recognition accuracy.

Description

Technical Field

[0001] The present application relates to the field of biometric identification technology, and in particular to an optimization method and related equipment for an interactive liveness detection solution. Background Art

[0002] At present, with the widespread application of face recognition technology, how to prevent malicious attackers from deceiving face recognition systems by means of photos, videos or 3D models has become an urgent problem to be solved. Based on this, interactive liveness detection technology was proposed, which aims to verify the authenticity of the face through the interaction between the user and the system. Interactive liveness detection technology plays an important role in improving the security of face recognition and responding to various attack methods.

[0003] In the interactive liveness detection of related technologies, the system requires the user to perform specific actions or expressions, such as blinking, opening the mouth, shaking the head, etc., in order to determine whether the user is a real person, so as to effectively distinguish between real humans and fake faces or videos generated by deception. However, in order to improve the accuracy of face recognition, this method often brings high time and performance costs. For example, the system requires the user to perform complex actions multiple times or for a long time, which not only consumes the user's time, but also reduces the user experience due to the complexity of the operation. Therefore, how to reduce the user's time consumption and the difficulty of user action execution on the basis of face recognition accuracy, and thus improve the user experience, is a technical problem that urgently needs to be solved.

[0004] The above contents are only used to assist in understanding the technical solution of the present application and do not constitute an admission that the above contents are prior art. Summary of the invention

[0005] The main purpose of this application is to provide an optimization method for an interactive liveness detection solution, aiming to solve the technical problem of how to reduce user time consumption and the difficulty of user action execution on the basis of face recognition security, thereby improving user experience.

[0006] To achieve the above objectives, the present application proposes an optimization method for an interactive liveness detection scheme, and the optimization method for the interactive liveness detection scheme includes:

[0007] Obtaining a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized;

[0008] Based on the detection model set, performing task error rate detection on the detection task set to obtain the detection error rate of the living body detection scheme to be optimized;

[0009] The detection task set and the preset task time index are used to calculate the time index, so as to obtain the detection time cost of the live body detection scheme to be optimized; the detection task set and the preset task difficulty index are used to calculate the difficulty index, so as to obtain the action difficulty value of the live body detection scheme to be optimized;

[0010] Based on the detection error rate, the detection time cost and the action difficulty, an optimal liveness detection scheme is selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm, and the optimal liveness detection scheme is used as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization targets.

[0011] In one embodiment, the step of selecting an optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost, and the action difficulty through a preset multi-objective optimization algorithm, and using the optimal liveness detection scheme as the optimized liveness detection scheme includes:

[0012] Based on the detection error rate, the detection time cost and the action difficulty, sorting the non-dominated solutions of the living body detection schemes to be optimized to obtain sorted living body detection schemes to be optimized;

[0013] A preset number of liveness detection schemes are selected from the sorted liveness detection schemes to be optimized in descending order as optimal liveness detection schemes.

[0014] In one embodiment, the step of selecting a preset number of liveness detection schemes from the sorted liveness detection schemes to be optimized in descending order as the optimal liveness detection scheme includes:

[0015] Calculating the congestion of the sorted liveness detection schemes to be optimized to obtain congestion information;

[0016] Based on the congestion information, a preset number of liveness detection schemes are selected from the sorted liveness detection schemes to be optimized in descending order as optimal liveness detection schemes.

[0017] In one embodiment, the step of selecting an optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost, and the action difficulty through a preset multi-objective optimization algorithm, and using the optimal liveness detection scheme as the optimized liveness detection scheme includes:

[0018] Based on the detection time cost, the action difficulty and the detection performance, a preset number of optimal liveness detection schemes are selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm;

[0019] Calculating the coverage volume of the optimal living body detection solution in the objective function space to obtain hypervolume data;

[0020] Determining whether the super volume data is greater than a preset super volume threshold;

[0021] If the super volume data is less than or equal to the super volume threshold, the optimal liveness detection scheme is used as the liveness detection scheme to be optimized in the next round of iteration, and the process returns to the step of obtaining the liveness detection scheme to be optimized, the detection task set of the liveness detection scheme to be optimized, and the detection model set until the super volume data is greater than the super volume threshold, then the selection of the optimal liveness detection scheme is stopped, and the preset number of optimal liveness detection schemes in this round are used as the optimized liveness detection scheme.

[0022] In one embodiment, after the step of calculating the difficulty index of the detection task set and the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized, the method includes:

[0023] Based on the detection error rate, the detection time cost and the action difficulty, screening the liveness detection scheme to be optimized to obtain the liveness detection scheme to be exchanged;

[0024] Randomly select intersection points from the liveness detection schemes to be exchanged, and exchange the intersection points between the liveness detection schemes to be exchanged to obtain exchanged liveness detection schemes;

[0025] The liveness detection scheme to be optimized is combined with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme.

[0026] In one embodiment, after the step of combining the liveness detection scheme to be optimized with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme, the method includes:

[0027] arbitrarily selecting a liveness detection scheme from the amplified liveness detection schemes as the liveness detection scheme to be mutated;

[0028] Randomly selecting a target variation point in the code string of the liveness detection scheme to be mutated, wherein the target variation point includes a detection task set and a detection model set;

[0029] Based on a preset replacement library, the detection task set or the detection model set is replaced for the target variation point to obtain a mutated liveness detection solution.

[0030] In addition, to achieve the above-mentioned purpose, the present application also proposes an optimization device for an interactive liveness detection scheme, and the optimization device for the interactive liveness detection scheme includes:

[0031] An acquisition module, used to acquire a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized;

[0032] A detection module, configured to perform task error rate detection on the detection task set based on the detection model set, and obtain the detection error rate of the living body detection scheme to be optimized;

[0033] A calculation module, used to perform time index calculation on the detection task set and the preset task time index to obtain the detection time cost of the live body detection scheme to be optimized; perform difficulty index calculation on the detection task set and the preset task difficulty index to obtain the action difficulty value of the live body detection scheme to be optimized;

[0034] An optimization module is used to select an optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty through a preset multi-objective optimization algorithm, and use the optimal liveness detection scheme as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization objectives.

[0035] In addition, to achieve the above-mentioned objectives, the present application also proposes an optimization device for an interactive liveness detection scheme, the device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the optimization method for the interactive liveness detection scheme as described above.

[0036] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the optimization method of the interactive liveness detection scheme as described above are implemented.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the optimization method of the interactive liveness detection scheme as described above.

[0038] One or more technical solutions proposed in this application have at least the following technical effects:

[0039] Compared with the related art in which the system requires the user to perform complex actions multiple times or for a long time, which not only consumes the user's time but also reduces the user experience due to the complexity of the operation, the present application performs task error rate detection on the detection task set based on the detection model set to obtain the detection error rate of the liveness detection scheme to be optimized, calculates the time index of the detection task set with the preset task time index to obtain the detection time cost of the liveness detection scheme to be optimized, and calculates the difficulty index of the detection task set with the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized, that is, respectively determines the detection error rate, detection time cost and action difficulty value of the liveness detection scheme to be optimized, which can be It can be understood that the detection error rate, detection time cost and action difficulty value of the liveness detection scheme to be optimized respectively reflect the recognition accuracy, user consumption time and user action execution difficulty in the liveness detection process. Finally, based on the detection error rate, the detection time cost and the action difficulty, the optimal liveness detection scheme is selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm. The optimal liveness detection scheme is the optimal liveness detection scheme that takes into account recognition accuracy, user consumption time and user action execution difficulty. Therefore, after applying the optimal liveness detection scheme, it is achieved to reduce user consumption time and user action execution difficulty on the basis of ensuring face recognition accuracy, thereby improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0042] Figure 1 A flowchart diagram of the first embodiment of the optimization method of the interactive liveness detection solution of the present application;

[0043] Figure 2 A flow chart of the second embodiment of the optimization method of the interactive liveness detection solution of the present application;

[0044] Figure 3 A schematic diagram of the module structure of an optimization device for an interactive living body detection solution according to an embodiment of the present application;

[0045] Figure 4Schematic diagram of the device structure of the hardware operating environment involved in the optimization method of the interactive liveness detection solution in the embodiment of the present application.

[0046] The purpose, features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0047] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of the present application and are not used to limit the present application.

[0048] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0049] The main solution of the embodiment of the present application is: obtain the liveness detection scheme to be optimized, the detection task set of the liveness detection scheme to be optimized, and the detection model set; based on the detection model set, perform task error rate detection on the detection task set to obtain the detection error rate of the liveness detection scheme to be optimized; perform time index calculation on the detection task set and the preset task time index to obtain the detection time cost of the liveness detection scheme to be optimized; perform difficulty index calculation on the detection task set and the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized; based on the detection error rate, the detection time cost and the action difficulty, select the optimal liveness detection scheme from the liveness detection scheme to be optimized through a preset multi-objective optimization algorithm, and use the optimal liveness detection scheme as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with detection error rate, detection time cost and action difficulty as optimization targets.

[0050] In this embodiment, for ease of description, the optimization device of the interactive living body detection solution is used as the execution subject for explanation below.

[0051] In the interactive liveness detection of related technologies, the system requires the user to perform specific actions or expressions, such as blinking, opening the mouth, shaking the head, etc., in order to determine whether the user is a real person, so as to effectively distinguish between real humans and fake faces or videos generated by deception. However, in order to improve the accuracy of face recognition, this method often brings high time and performance costs. For example, the system requires the user to perform complex actions multiple times or for a long time, which not only consumes the user's time, but also reduces the user experience due to the complexity of the operation. Therefore, how to reduce the user's time consumption and the difficulty of user action execution on the basis of face recognition accuracy, and thus improve the user experience, is a technical problem that urgently needs to be solved.

[0052] The present application provides a solution to optimize the liveness detection solution, thereby improving the user experience on the basis of considering the safety performance of the liveness detection system.

[0053] It can be seen from the above embodiments that the present application performs task error rate detection on the detection task set based on the detection model set to obtain the detection error rate of the live detection scheme to be optimized, performs time index calculation on the detection task set and the preset task time index to obtain the detection time cost of the live detection scheme to be optimized, and performs difficulty index calculation on the detection task set and the preset task difficulty index to obtain the action difficulty value of the live detection scheme to be optimized, that is, respectively determines the detection error rate, detection time cost and action difficulty value of the live detection scheme to be optimized. It can be understood that the detection error rate and detection time cost of the live detection scheme to be optimized The cost and action difficulty values ​​respectively reflect the recognition accuracy, user consumed time and user action execution difficulty in the liveness detection process. Finally, based on the detection error rate, the detection time cost and the action difficulty, the optimal liveness detection scheme is selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm. The optimal liveness detection scheme is the optimal liveness detection scheme that takes into account recognition accuracy, user consumed time and user action execution difficulty. Therefore, after applying the optimal liveness detection scheme, it is achieved to reduce user consumed time and user action execution difficulty while ensuring face recognition accuracy, thereby improving user experience.

[0054] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of realizing the above functions, a terminal system, etc. The following takes the optimization device of the interactive liveness detection solution as an example to illustrate this embodiment and the following embodiments.

[0055] Based on this, the present application embodiment provides an optimization method for an interactive liveness detection solution, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the optimization method of the interactive living body detection solution of the present application.

[0056] In this embodiment, the optimization method of the interactive living body detection scheme includes steps S100 to S400:

[0057] Step S100, obtaining a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized;

[0058] In a specific implementation, the optimization method of the interactive liveness detection scheme of the present application can be applied in multiple scenarios, such as in the financial field. When a financial enterprise receives a financial transaction request such as a user transfer, it needs to authenticate whether the current request is initiated by the user himself. The specific method can be an interactive liveness detection active defense method. The specific process is that the financial enterprise requires the user to perform a series of predefined actions, such as blinking, opening the mouth, shaking the head, etc., and judges whether it is a real face by analyzing the real-time and coherence of these actions. Among them, the more actions the user is required to perform, the higher the complexity, and the more accurate the system's judgment of the real face. However, in order to improve the security of face recognition, this method often brings high time and performance costs. For example, the system requires the user to perform multiple or long complex actions, which will not only consume the user's time, but also cause the user experience to decline due to the complexity of the operation. Therefore, in order to solve the above problems, the present application proposes to improve the user experience on the basis of considering the security performance of the detection system, and specifically optimizes the liveness detection scheme to find the optimal interactive liveness detection scheme, so that it can improve the user experience on the basis of considering the security performance of the liveness detection system.

[0059] It should be noted that the liveness detection scheme to be optimized refers to multiple interactive liveness detections that require optimization, and the interactive liveness detection system performs corresponding liveness detection according to the liveness detection scheme. The liveness detection scheme to be optimized includes but is not limited to a liveness detection task set and a liveness detection model set.

[0060] Among them, the liveness detection task set is a collection of multiple liveness detection tasks. The liveness detection task can be a silent action required to be completed by the person being detected, such as nodding, turning the head left, turning the head right, opening the mouth, blinking, etc.; it can also be a voice required to be uttered by the person being detected, such as coughing, speaking with emotion, humming, and singing, etc.; it can also be a mixed task of action and voice, such as reading a paragraph of text aloud.

[0061] Furthermore, the liveness detection model set refers to a model that can detect the authenticity of tasks in the liveness detection task set. Among them, the detection error rate of each liveness detection model corresponding to each liveness detection task can be obtained through experiments. Referring to the following Table 1, it is specifically an example of the detection error rate of the liveness detection model corresponding to the liveness detection task.

[0062] Table 1

[0063]

[0064]

[0065] It is understandable that each of the liveness detection solutions to be optimized includes liveness detection task sets and liveness detection model sets of different numbers and / or different attributes. Specifically, the liveness detection solution to be optimized X, wherein X has N solutions. The encoding string of each liveness detection active defense solution is x={s1...,s m ,d1...,d nn},s i ∈S,d j ∈D,n≥m. Every s i is a task selected from the liveness detection task set S. Each d j ∈D is a detection model selected from the liveness detection model set D. The number of tasks m and the number of detection models n in X are also random.

[0066] Step S200, based on the detection model set, performing task error rate detection on the detection task set to obtain the detection error rate of the living body detection scheme to be optimized;

[0067] It should be noted that the detection error rate refers to the error probability of the liveness detection system when performing detection work. The detection error rate is specifically determined by the ability to timely and accurately detect fake faces or videos generated by malicious attackers. The higher the detection success rate of fake faces or videos, the lower the detection error rate of the liveness detection system. Usually, the more detection tasks are set and / or the more complex the tasks are, the lower the detection error rate of the liveness detection system. The detection error rate also reflects the performance of the liveness detection system. The lower the detection error rate, the higher the performance of the liveness detection system.

[0068] In a specific implementation, the device performs task error rate detection on the detection task set through the above Table 1 and the detection model set to obtain the detection error rate ε of the liveness detection scheme to be optimized. u Specifically, the device calculates the detection error rate ε u Specifically, the device uses the following formula to optimize each task s in each living body detection scheme x: i ,i=1,...,mCalculate the error rate of the detection model:

[0069]

[0070] The overall detection error rate of the device calculation solution is:

[0071] Step S300, performing time index calculation on the detection task set and the preset task time index to obtain the detection time cost of the liveness detection scheme to be optimized; performing difficulty index calculation on the detection task set and the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized;

[0072] It should be noted that the detection time cost refers to the time spent on performing liveness detection tasks. Generally, the more detection tasks are set and / or the more complex the tasks are, the higher the detection time cost of the liveness detection system is; the action difficulty refers to the action difficulty of the liveness detection tasks that the user needs to perform. Generally, the more detection tasks are set and / or the more complex the tasks are, the higher the action difficulty of the liveness detection system is.

[0073] In a specific implementation, the preset detection task indicator set is a set of indicators used to help calculate the target value in the interactive liveness detection active defense method. The detection task indicator set can be obtained through experiments, surveys or other methods. The detection task indicator set includes two indicators: the time it takes for the person being detected to complete the task and the difficulty of the person being detected to complete the task. Please refer to the following Table 2 for details.

[0074] Table 2

[0075]

[0076] In a specific implementation, the device calculates the detection time cost and action difficulty of each of the liveness detection solutions to be optimized based on the preset detection task indicator set and the detection task set. Specifically, the liveness detection system can calculate the detection time cost and action difficulty of each solution x={s1...,s m ,d1...,d nn},s i ∈S,d j ∈D is tested, that is, the device inputs the liveness detection scheme to be optimized into the preset liveness detection system, and the device obtains the detection result output by the liveness detection system.

[0077] Furthermore, the device calculates the detection consumption time ε of each liveness detection scheme x to be optimized by the task completion time index and test time of the detected person in Table 2 above: e Among them, the detection time ε e is to complete all tasks s in solution x i ,i=1,...,m and all detection models d j ,j=1,...,n time.

[0078] Furthermore, the device obtains the difficulty of the task ε through the difficulty index of the task completed by the detected person in Table 2 c , where the difficulty of the task ε c It is to complete all tasks s in each liveness detection scheme x to be optimized i ,i=1,...,m,the sum of the difficulties.

[0079] In a specific implementation, the device obtains each living body detection scheme to be optimized x={s1...,s m ,d1...,d n}, the corresponding detection error rate ε u , Detection consumption time ε e and the difficulty of the task involved in the detection c .

[0080] In a specific implementation, after the device calculates the difficulty index of the detection task set and the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized, the method includes:

[0081] Based on the detection error rate, the detection time cost and the action difficulty, the liveness detection scheme to be optimized is screened to obtain the liveness detection scheme to be exchanged; an intersection is arbitrarily selected from the liveness detection schemes to be exchanged, and the intersection is exchanged between the liveness detection schemes to be exchanged to obtain the exchanged liveness detection schemes; the liveness detection scheme to be optimized is combined with the exchanged liveness detection scheme to obtain the amplified liveness detection scheme.

[0082] In the specific implementation, in order to improve the optimization quality of the liveness detection scheme, the present application expands the number of samples of the liveness detection scheme to be optimized. Specifically, the device selects the liveness detection scheme to be exchanged from the liveness detection scheme to be optimized based on the detection error rate, the detection time cost and the action difficulty, and performs cross-replacement processing to obtain a new liveness detection scheme.

[0083] In a specific implementation, the device screens the liveness detection scheme to be optimized based on the detection error rate, the detection time cost and the action difficulty, and obtains the liveness detection scheme to be exchanged in the following manner: the device calculates the fitness and non-inferiority ranking of each liveness detection scheme to be optimized based on the detection error rate, the detection time cost and the action difficulty, and preferentially selects two solutions with good fitness and high ranking for cross-replacement processing, thereby obtaining a new liveness detection scheme. Among them, the detection error rate, the detection time cost and the action difficulty each have a corresponding mapping with the fitness, and the value of the non-inferiority ranking is calculated by the NSGA-II algorithm (Nondominated Sorting GeneticAlgorithm II, a multi-objective algorithm based on a genetic algorithm).

[0084] In a specific implementation, after the device combines the liveness detection scheme to be optimized with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme, the method includes:

[0085] A liveness detection scheme is arbitrarily selected from the amplified liveness detection schemes as the liveness detection scheme to be mutated; a target variation point is arbitrarily selected from the encoding string of the liveness detection scheme to be mutated, wherein the target variation point includes a detection task set and a detection model set; based on a preset replacement library, the target variation point is replaced with a detection task set or a detection model set to obtain a mutated liveness detection scheme.

[0086] In a specific implementation, the device randomly selects one or more intersections in a code string of a liveness detection scheme to be exchanged, wherein the intersections determine which segments of the code string will be exchanged. The device then divides corresponding parts from the code strings of the two liveness detection schemes to be exchanged according to the intersections, and exchanges these parts to obtain the exchanged liveness detection scheme.

[0087] In a specific implementation, the device combines the exchanged coding segments to form a new liveness detection scheme, and combines the initial liveness detection scheme to be optimized with the newly added exchanged liveness detection scheme to obtain an amplified liveness detection scheme.

[0088] On the basis of the amplified liveness detection scheme, the present application also proposes to mutate the amplified liveness detection scheme to further amplify the liveness detection scheme. Specifically, first, the device selects a solution from the current amplified liveness detection scheme to mutate, wherein the mutation is usually performed with a lower probability to avoid excessive damage to excellent solutions; secondly, the device randomly selects one or more mutation points in the coding string of the solution, wherein the mutation point determines which actions or detection models in the solution will be replaced; finally, the device replaces the action or detection model of the solution according to the mutation point, wherein the replacement can be random or based on a certain strategy. Finally, a further amplified liveness detection scheme is obtained.

[0089] Furthermore, after the device amplifies the liveness detection scheme through the cross operation, it also performs a mutation operation on the amplified liveness detection scheme, wherein the mutation operation specifically refers to randomly selecting a target mutation point of the liveness detection scheme to be mutated, and replacing the detection task set or detection model set for the target mutation point based on a preset replacement library and probability v. The probability v is usually very small, such as 1%. It should be noted that the present application realizes the transformation of a small number of detection task sets or detection model sets by setting a small probability scheme detection task set or detection model set transformation, thereby preventing the problem of a large number of transformations causing the result to deteriorate, so as to further expand the liveness detection scheme.

[0090] Step S400, based on the detection error rate, the detection time cost and the action difficulty, through a preset multi-objective optimization algorithm, select the optimal liveness detection scheme from the liveness detection schemes to be optimized, and use the optimal liveness detection scheme as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization targets.

[0091] In the specific implementation, the present application proposes a multi-objective optimization algorithm with detection performance, detection time cost and action difficulty as optimization targets to optimize the liveness detection scheme to be optimized, and obtain an optimized liveness detection scheme. The optimized liveness detection scheme takes into account the recognition accuracy, user consumption time and user action execution difficulty, and is the optimal liveness detection scheme. Therefore, after applying the optimal liveness detection scheme, the interactive liveness detection active defense application is realized, which reduces the user consumption time and the difficulty of user action execution on the basis of ensuring the accuracy of face recognition, thereby improving the user experience.

[0092] It should be noted that the optimization goal of the liveness detection solution proposed in this application is to select the optimal combination of liveness detection tasks and liveness detection models from multiple liveness detection tasks and liveness detection models with different attributes, so as to reduce the user's time consumption and the difficulty of user action execution while considering the safety performance of the liveness detection system, thereby improving the user experience.

[0093] Furthermore, the present application can systematically select an interactive liveness detection active defense solution that can simultaneously optimize multiple objectives through a multi-objective optimization algorithm. The resulting solution can optimally balance multiple important objectives. For example, in this solution, the present application uses a multi-objective optimization algorithm to find a solution that can simultaneously optimize the detection error rate ε u , Detection consumption time ε e and the difficulty of the task ε c Interactive liveness detection active defense solution.

[0094] It should be noted that the multi-objective optimization algorithm is an optimization algorithm used to solve multiple objective functions. The goal of the multi-objective optimization algorithm is to find a balance between multiple potentially conflicting objectives and ultimately obtain an approximate optimal solution set for the multi-objective optimization problem. The multi-objective optimization algorithm includes but is not limited to a genetic algorithm, a particle swarm algorithm, a differential evolution algorithm, and a simulated annealing algorithm.

[0095] Furthermore, the multi-objective optimization algorithm in the present application is preferably a genetic algorithm, wherein the optimization problem of the interactive liveness detection scheme is specifically expressed by the following formula:

[0096]

[0097] st

[0098] ε u (θ)<φ u ,

[0099] ε e (θ)<φ e ,

[0100] ε c (θ)<φ c ,

[0101] Among them, ε u represents the detection error rate, φ u represents the error rate threshold, ε e represents the time cost, φ e represents the time cost threshold, ε c Indicates the difficulty of the action, φ c represents the action difficulty threshold, X * Represents the optimal liveness detection solution.

[0102] In this embodiment, the multi-objective optimization algorithm is an optimization algorithm with detection error rate, detection time cost and action difficulty as optimization targets, that is, the multi-objective optimization algorithm involves the objective function of detection error rate, the objective function of detection time cost and the objective function of action difficulty. Preferably, the multi-objective optimization algorithm adopted is specifically a multi-objective genetic algorithm. Specifically, the device selects the optimal liveness detection scheme from the liveness detection schemes to be optimized by using a preset multi-objective optimization algorithm. First, the first objective solution (calculated based on the objective function of detection error rate), the second objective solution (calculated based on the objective function of detection time cost) and the third objective solution (calculated based on the objective function of action difficulty) of each liveness detection scheme to be optimized are respectively calculated, and then based on the first objective solution, the second objective solution and the third objective solution, the fitness evaluation of each liveness detection scheme is determined, and the optimal liveness detection scheme is selected based on the fitness evaluation of each liveness detection scheme, wherein the liveness detection scheme with the highest fitness evaluation is the optimal liveness detection scheme.

[0103] Furthermore, the device determines the fitness evaluation of each liveness detection scheme based on the first target solution, the second target solution and the third target solution, and selects the optimal liveness detection scheme based on the fitness evaluation of each liveness detection scheme, including: (1) randomly generating different weights for all individuals in the liveness detection scheme to be optimized, and then calculating the fitness evaluation of each liveness detection scheme based on the first target solution, the second target solution, the third target solution and the weight value, and finally performing an iterative liveness detection scheme selection operation until the optimal liveness detection scheme is obtained; (2) dividing the liveness detection scheme to be optimized into three subpopulations of equal size, and selecting the optimal liveness detection scheme based on the fitness evaluation of each liveness detection scheme. The sub-population sets the target weights according to different objective functions, calculates the fitness evaluation of each liveness detection scheme, and finally performs an iterative liveness detection scheme selection operation until the optimal liveness detection scheme is obtained; (3) based on the first objective solution, the second objective solution and the third objective solution, selects a non-dominated liveness detection scheme from the liveness detection schemes to be optimized (a non-dominated liveness detection scheme refers to a liveness detection scheme in which there is no other liveness detection scheme in the population whose first objective solution, second objective solution and third objective solution are completely higher than itself), and then manually selects a non-dominated liveness detection scheme to obtain the optimal liveness detection scheme.

[0104] In a specific implementation, the device selects the optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty through a preset multi-objective optimization algorithm, and uses the optimal liveness detection scheme as the optimized liveness detection scheme, including:

[0105] Based on the detection error rate, the detection time cost and the action difficulty, non-dominated solutions of each of the liveness detection schemes to be optimized are sorted to obtain sorted liveness detection schemes to be optimized; and a preset number of liveness detection schemes are selected from the sorted liveness detection schemes to be optimized in descending order as the optimal liveness detection scheme.

[0106] It should be noted that the non-dominated solution sorting is a method of sorting individuals in a population based on the dominance relationship between individuals. The non-dominated sorting divides individuals into multiple levels. The higher the level, the better the individual. Among them, individuals dominated by other individuals are ranked lower.

[0107] In a specific implementation, an individual under the non-dominated solution sorting refers to a liveness detection scheme, and the non-dominated solution sorting is performed based on each liveness detection scheme, that is, a preset number of optimal liveness detection schemes are selected from the liveness detection schemes to be optimized.

[0108] In the specific implementation, the non-dominated solution sorting method is an effective sorting method for multi-objective optimization, which can improve the accuracy of selecting the optimal liveness detection solution, that is, improve the quality of the optimal liveness detection solution.

[0109] In a specific implementation, the step of selecting a preset number of liveness detection schemes from the sorted liveness detection schemes to be optimized in descending order as the optimal liveness detection scheme includes:

[0110] The liveness detection schemes to be optimized after sorting are calculated for congestion to obtain congestion information; based on the congestion information, a preset number of liveness detection schemes to be optimized after sorting are selected in descending order as the optimal liveness detection scheme.

[0111] In a specific implementation, after sorting the non-dominated solutions of the liveness detection scheme to be optimized, the device also calculates the crowding degree of each level of individuals. The crowding degree measures the density around the individual, that is, the distance between the individual and its neighboring individuals, wherein the higher the crowding degree, the more crowded the area where the individual is located.

[0112] In a specific implementation, the present application calculates the crowding degree of individuals at each level to achieve a more detailed sorting of liveness detection schemes at each level, thereby improving the accuracy of selecting the optimal liveness detection scheme, that is, improving the quality of the optimal liveness detection scheme.

[0113] Compared with the related art in which the system requires the user to perform complex actions multiple times or for a long time, which not only consumes the user's time but also reduces the user experience due to the complexity of the operation, the present application performs task error rate detection on the detection task set based on the detection model set to obtain the detection error rate of the liveness detection scheme to be optimized, calculates the time index of the detection task set with the preset task time index to obtain the detection time cost of the liveness detection scheme to be optimized, and calculates the difficulty index of the detection task set with the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized, that is, respectively determines the detection error rate, detection time cost and action difficulty value of the liveness detection scheme to be optimized, which can be It can be understood that the detection error rate, detection time cost and action difficulty value of the liveness detection scheme to be optimized respectively reflect the recognition accuracy, user consumption time and user action execution difficulty in the liveness detection process. Finally, based on the detection error rate, the detection time cost and the action difficulty, the optimal liveness detection scheme is selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm. The optimal liveness detection scheme is the optimal liveness detection scheme that takes into account recognition accuracy, user consumption time and user action execution difficulty. Therefore, after applying the optimal liveness detection scheme, it is achieved to reduce user consumption time and user action execution difficulty on the basis of ensuring face recognition accuracy, thereby improving user experience.

[0114] Based on the above first embodiment, the present application also proposes another embodiment, referring to Figure 2 , the optimization method of the interactive liveness detection scheme includes:

[0115] The device selects the optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty through a preset multi-objective optimization algorithm, and uses the optimal liveness detection scheme as the optimized liveness detection scheme, including:

[0116] Step A100, based on the detection time cost, the action difficulty and the detection performance, a preset number of optimal liveness detection schemes are selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm;

[0117] In the specific implementation, this step has been specifically described in the first embodiment, and reference may be made to the above first embodiment, and will not be repeated here.

[0118] Step A200, calculating the coverage volume of the optimal living body detection solution in the objective function space to obtain hypervolume data;

[0119] It should be noted that hypervolume is an indicator used to evaluate the performance of a solution set in multi-objective optimization problems. It is usually used to calculate the volume covered by the solution set in the objective function space. The calculation method of hypervolume depends on the number and type of objective functions in the problem, but the basic principle is to calculate the volume of all non-dominated sorted solutions.

[0120] In the specific implementation, for the optimization problem of multiple objectives, the hypervolume can be calculated by the following steps: 1. Normalization: Normalize all objective values ​​to the interval [0,1]. This is usually achieved by subtracting the minimum value of each objective and dividing by the difference between the maximum and minimum values; 2. Non-dominated sorting: Sort the solutions according to the non-dominated relationship; 3. Calculate the hypervolume: For each non-dominated sorted solution, calculate the volume it covers in the objective function space.

[0121] Step A300, determining whether the hypervolume data is greater than a preset hypervolume threshold;

[0122] Step A400: If the super-volume data is less than or equal to the super-volume threshold, the optimal liveness detection scheme is used as the liveness detection scheme to be optimized in the next round of iteration, and the step of obtaining the liveness detection scheme to be optimized, the detection task set of the liveness detection scheme to be optimized, and the detection model set is returned to the step of obtaining the liveness detection scheme to be optimized, until the super-volume data is greater than the super-volume threshold, then the selection of the optimal liveness detection scheme is stopped, and the preset number of optimal liveness detection schemes in this round are used as the optimized liveness detection scheme.

[0123] It should be noted that the present application proposes to adopt an iterative selection method to screen the optimal liveness detection scheme, and iteratively select the optimal liveness detection scheme through multiple cycles, so that the selected liveness detection scheme is constantly close to the optimal.

[0124] In a specific implementation, the iterative stop condition in the iterative selection method proposed in the present application is specifically whether the hypervolume data of the optimal liveness detection solution is greater than a preset hypervolume threshold, wherein the hypervolume is an indicator used to evaluate the search results of a multi-objective problem, and specifically, the hypervolume can measure the size of the area dominated by the non-dominated solution set. If the hypervolume data of the optimal liveness detection solution is less than or equal to the hypervolume threshold, it means that the optimal liveness detection solution of this round has not achieved the set goal, then the device uses the optimal liveness detection solution as the liveness detection solution to be optimized for the next round of iteration, and returns to the step of determining the detection task set and the detection model set of each liveness detection solution to be optimized, until the hypervolume data is greater than the hypervolume threshold, then the selection of the optimal liveness detection solution is stopped, and the preset number of optimal liveness detection solutions of this round are used as the optimized liveness detection solution.

[0125] It should be noted that the iteration stop condition can also be to meet a certain iteration cycle, for example, the device stops selecting the optimal liveness detection scheme after reaching 20 iteration cycles. That is, if the algorithm reaches the end requirement, the algorithm ends, otherwise, the optimal liveness detection scheme is used as the liveness detection scheme to be optimized in the next round of iteration, and the step of determining the detection task set and detection model set of each liveness detection scheme to be optimized is completed.

[0126] It should be noted that the above examples are only used to understand the present application and do not constitute a limitation on the optimization method of the interactive liveness detection solution of the present application. More simple transformations based on this technical concept are all within the scope of protection of the present application.

[0127] This application also provides an optimization device for an interactive liveness detection solution, please refer to Figure 3 , the optimization device of the interactive living body detection scheme includes:

[0128] An acquisition module 10 is used to acquire a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized;

[0129] A detection module 20, configured to perform task error rate detection on the detection task set based on the detection model set, and obtain a detection error rate of the liveness detection scheme to be optimized;

[0130] The calculation module 30 is used to calculate the time index of the detection task set and the preset task time index to obtain the detection time cost of the live body detection scheme to be optimized; and to calculate the difficulty index of the detection task set and the preset task difficulty index to obtain the action difficulty value of the live body detection scheme to be optimized;

[0131] The optimization module 40 is used to select the optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty through a preset multi-objective optimization algorithm, and use the optimal liveness detection scheme as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization objectives.

[0132] Optionally, the optimization module 40 includes:

[0133] A sorting module, configured to sort the non-dominated solutions of the living body detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty, so as to obtain sorted living body detection schemes to be optimized;

[0134] The selection module is used to select a preset number of liveness detection schemes from the sorted liveness detection schemes to be optimized in descending order as the optimal liveness detection scheme.

[0135] Optionally, the selection module includes:

[0136] A congestion calculation module is used to calculate the congestion of the sorted liveness detection schemes to be optimized to obtain congestion information;

[0137] The congestion degree selection module is used to select a preset number of liveness detection schemes as the optimal liveness detection schemes from the sorted liveness detection schemes to be optimized in descending order based on the congestion degree information.

[0138] Optionally, the optimization module 40 further includes:

[0139] A scheme selection module, configured to select a preset number of optimal liveness detection schemes from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm based on the detection time cost, the action difficulty and the detection performance;

[0140] A hypervolume data calculation module, used to calculate the coverage volume of the optimal living body detection solution in the objective function space to obtain hypervolume data;

[0141] A judging module, used to judge whether the super volume data is greater than a preset super volume threshold;

[0142] An iterative selection module is used to use the optimal liveness detection scheme as the liveness detection scheme to be optimized in the next round of iteration if the super volume data is less than or equal to the super volume threshold, and return to the step of obtaining the liveness detection scheme to be optimized, the detection task set of the liveness detection scheme to be optimized, and the detection model set, until the super volume data is greater than the super volume threshold, then stop selecting the optimal liveness detection scheme, and use the preset number of optimal liveness detection schemes in this round as the optimized liveness detection scheme.

[0143] Optionally, the optimization device of the interactive living body detection solution further includes:

[0144] A screening module, configured to screen the liveness detection scheme to be optimized based on the detection error rate, the detection time cost and the action difficulty, to obtain the liveness detection scheme to be exchanged;

[0145] An exchange module, configured to arbitrarily select an intersection point from the liveness detection schemes to be exchanged, and exchange the intersection point between the liveness detection schemes to be exchanged to obtain an exchanged liveness detection scheme;

[0146] The amplification module is used to combine the liveness detection scheme to be optimized with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme.

[0147] Optionally, the optimization device of the interactive living body detection solution further includes:

[0148] A module for determining a liveness detection scheme to be mutated, used for arbitrarily selecting a liveness detection scheme from the amplified liveness detection schemes as the liveness detection scheme to be mutated;

[0149] A variation point selection module, used to arbitrarily select a target variation point in the code string of the liveness detection scheme to be mutated, wherein the target variation point includes a detection task set and a detection model set;

[0150] The replacement module is used to replace the detection task set or the detection model set of the target variation point based on a preset replacement library to obtain a mutated liveness detection solution.

[0151] The interactive liveness detection scheme optimization device provided by the present application adopts the interactive liveness detection scheme optimization method in the above-mentioned embodiment, and can solve the technical problem of optimizing the interactive liveness detection scheme. Compared with the prior art, the beneficial effects of the interactive liveness detection scheme optimization device provided by the present application are the same as the beneficial effects of the interactive liveness detection scheme optimization method provided by the above-mentioned embodiment, and other technical features in the interactive liveness detection scheme optimization device are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.

[0152] The present application provides an optimization device for an interactive liveness detection solution, and the optimization device for the interactive liveness detection solution includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the optimization method for the interactive liveness detection solution in the above-mentioned embodiment 1.

[0153] Reference below Figure 4 , which shows a schematic diagram of the structure of an optimization device suitable for implementing the interactive liveness detection solution of the embodiment of the present application. The optimization device of the interactive liveness detection solution in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The optimization device of the interactive liveness detection solution shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0154] like Figure 4As shown, the optimization device of the interactive liveness detection scheme may include a processing device 1001 (such as a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 to a random access memory (RAM: Random Access Memory) 1004. Various programs and data required for the operation of the optimization device of the interactive liveness detection scheme are also stored in the RAM 1004. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems may be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 1003 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 1009. The communication device 1009 may allow the optimization device of the interactive liveness detection scheme to communicate wirelessly or wired with other devices to exchange data. Although the figure shows the optimization device of the interactive liveness detection scheme with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems may be implemented or have instead.

[0155] In particular, according to the embodiments disclosed in the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0156] The interactive liveness detection scheme optimization device provided by the present application adopts the interactive liveness detection scheme optimization method in the above-mentioned embodiment, which can solve the technical problem of optimizing the interactive liveness detection scheme. Compared with the prior art, the beneficial effects of the interactive liveness detection scheme optimization device provided by the present application are the same as the beneficial effects of the interactive liveness detection scheme optimization method provided by the above-mentioned embodiment, and other technical features in the interactive liveness detection scheme optimization device are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0157] It should be understood that the various parts disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0158] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0159] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, computer programs) stored thereon, and the computer-readable program instructions are used to execute the optimization method of the interactive living body detection scheme in the above-mentioned embodiment.

[0160] The computer-readable storage medium provided in the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0161] The computer-readable storage medium may be included in the optimization device of the interactive liveness detection solution; or may exist independently without being assembled into the optimization device of the interactive liveness detection solution.

[0162] The computer-readable storage medium carries one or more programs. When the one or more programs are executed by the interactive liveness detection scheme optimization device, the interactive liveness detection scheme optimization device enables the interactive liveness detection scheme optimization.

[0163] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0164] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0165] The modules involved in the embodiments described in this application may be implemented by software or hardware, wherein the name of the module does not constitute a limitation on the unit itself in some cases.

[0166] The readable storage medium provided in the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned interactive liveness detection scheme optimization method, and can solve the technical problem of optimizing the interactive liveness detection scheme. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in the present application are the same as the beneficial effects of the interactive liveness detection scheme optimization method provided in the above-mentioned embodiment, and will not be repeated here.

[0167] The present application also provides a computer program product, including a computer program, which implements the steps of the above-mentioned optimization method of the interactive living body detection solution when executed by a processor.

[0168] The computer program product provided by this application can solve the technical problem of optimizing the interactive liveness detection solution. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as the beneficial effects of the optimization method of the interactive liveness detection solution provided by the above embodiment, which will not be repeated here.

[0169] The above descriptions are only some embodiments of the present application, and are not intended to limit the patent scope of the present application. All equivalent structural changes made using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. An optimization method for an interactive liveness detection scheme, characterized in that: The method for optimizing the interactive liveness detection scheme includes: Obtaining a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized; Based on the detection model set, performing task error rate detection on the detection task set to obtain the detection error rate of the living body detection scheme to be optimized; The detection task set and the preset task time index are used to calculate the time index, so as to obtain the detection time cost of the live body detection scheme to be optimized; the detection task set and the preset task difficulty index are used to calculate the difficulty index, so as to obtain the action difficulty value of the live body detection scheme to be optimized; Based on the detection error rate, the detection time cost and the action difficulty, an optimal liveness detection scheme is selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm, and the optimal liveness detection scheme is used as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization targets.

2. The optimization method of the interactive liveness detection scheme according to claim 1, characterized in that: The step of selecting the optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost, and the action difficulty through a preset multi-objective optimization algorithm, and using the optimal liveness detection scheme as the optimized liveness detection scheme includes: Based on the detection error rate, the detection time cost and the action difficulty, sorting the non-dominated solutions of the living body detection schemes to be optimized to obtain sorted living body detection schemes to be optimized; A preset number of liveness detection schemes are selected from the sorted liveness detection schemes to be optimized in descending order as optimal liveness detection schemes.

3. The optimization method of the interactive liveness detection scheme as claimed in claim 2, characterized in that: The step of selecting a preset number of liveness detection schemes from the sorted liveness detection schemes to be optimized in descending order as the optimal liveness detection scheme comprises: Calculating the congestion of the sorted liveness detection schemes to be optimized to obtain congestion information; Based on the congestion information, a preset number of liveness detection schemes are selected from the sorted liveness detection schemes to be optimized in descending order as optimal liveness detection schemes.

4. The optimization method of the interactive liveness detection scheme according to claim 1, characterized in that: The step of selecting the optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost, and the action difficulty through a preset multi-objective optimization algorithm, and using the optimal liveness detection scheme as the optimized liveness detection scheme includes: Based on the detection time cost, the action difficulty and the detection performance, a preset number of optimal liveness detection schemes are selected from the liveness detection schemes to be optimized through a preset multi-objective optimization algorithm; Calculating the coverage volume of the optimal living body detection solution in the objective function space to obtain hypervolume data; Determining whether the super volume data is greater than a preset super volume threshold; If the super volume data is less than or equal to the super volume threshold, the optimal liveness detection scheme is used as the liveness detection scheme to be optimized in the next round of iteration, and the process returns to the step of obtaining the liveness detection scheme to be optimized, the detection task set of the liveness detection scheme to be optimized, and the detection model set until the super volume data is greater than the super volume threshold, then the selection of the optimal liveness detection scheme is stopped, and the preset number of optimal liveness detection schemes in this round are used as the optimized liveness detection scheme.

5. The optimization method of the interactive liveness detection scheme according to claim 1, characterized in that: After the step of calculating the difficulty index of the detection task set and the preset task difficulty index to obtain the action difficulty value of the liveness detection scheme to be optimized, the method includes: Based on the detection error rate, the detection time cost and the action difficulty, screening the liveness detection scheme to be optimized to obtain the liveness detection scheme to be exchanged; Randomly select intersection points from the liveness detection schemes to be exchanged, and exchange the intersection points between the liveness detection schemes to be exchanged to obtain exchanged liveness detection schemes; The liveness detection scheme to be optimized is combined with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme.

6. The optimization method of the interactive liveness detection scheme as claimed in claim 5, characterized in that: After the step of combining the liveness detection scheme to be optimized with the exchanged liveness detection scheme to obtain an amplified liveness detection scheme, the method includes: arbitrarily selecting a liveness detection scheme from the amplified liveness detection schemes as the liveness detection scheme to be mutated; Randomly selecting a target variation point in the code string of the liveness detection scheme to be mutated, wherein the target variation point includes a detection task set and a detection model set; Based on a preset replacement library, the detection task set or the detection model set is replaced for the target variation point to obtain a mutated liveness detection solution.

7. An optimization device for an interactive liveness detection scheme, characterized in that: The device comprises: An acquisition module, used to acquire a liveness detection scheme to be optimized, a detection task set and a detection model set of the liveness detection scheme to be optimized; A detection module, configured to perform task error rate detection on the detection task set based on the detection model set, and obtain the detection error rate of the living body detection scheme to be optimized; A calculation module, used to perform time index calculation on the detection task set and the preset task time index to obtain the detection time cost of the live body detection scheme to be optimized; perform difficulty index calculation on the detection task set and the preset task difficulty index to obtain the action difficulty value of the live body detection scheme to be optimized; An optimization module is used to select an optimal liveness detection scheme from the liveness detection schemes to be optimized based on the detection error rate, the detection time cost and the action difficulty through a preset multi-objective optimization algorithm, and use the optimal liveness detection scheme as the optimized liveness detection scheme, wherein the multi-objective optimization algorithm is an optimization algorithm with the detection error rate, the detection time cost and the action difficulty as optimization objectives.

8. An optimization device for an interactive liveness detection scheme, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the optimization method of the interactive liveness detection scheme according to any one of claims 1 to 6.

9. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the optimization method of the interactive living body detection solution according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of the optimization method of the interactive living body detection scheme according to any one of claims 1 to 6 are implemented.

Images