Man-machine verification method, device, equipment and medium

Abstract

The application discloses a man-machine verification method and device, equipment and medium. The man-machine verification method comprises the following steps: acquiring first motion data of a mobile terminal collected by a motion sensor; generating a motion graph according to the first motion data; querying a database according to the motion graph to determine the matching degree between the motion graph and historical data in the database; generating a target trajectory according to the motion graph in the case that the matching degree is greater than a first preset threshold and less than a second preset threshold; sending the target trajectory to the mobile terminal, so that an operation object of the mobile terminal feeds back according to the target trajectory; acquiring motion feedback information of the mobile terminal collected by the motion sensor in real time; and determining whether the operation object is a natural person according to the target trajectory and the motion feedback information, to obtain a man-machine verification result. According to the embodiment of the application, the accuracy of man-machine verification can be improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of network security, and particularly relates to a man-machine verification method and device, equipment and a medium. BACKGROUND

[0002] In many scenarios, real natural persons need to participate in interaction, but due to the development of artificial intelligence, a large number of new means appear to simulate the behavior of natural person users, for example, AI (artificial intelligence) robots simulate human behavior, resulting in security risks. Therefore, it is necessary to reduce risks by identifying the operation object of the mobile terminal to realize man-machine verification.

[0003] In the related art, man-machine verification is usually realized based on a pre-trained identification model, however, since an attacker can generate relevant data by an AI algorithm to simulate the behavior of a person, and input the data into the identification model to realize the purpose of identifying AI as a person, that is, for the man-machine verification scheme in the related art, there is a risk that AI behavior is disguised as human behavior to pass the man-machine verification and crack the man-machine verification system. SUMMARY

[0004] The embodiments of the application provide a man-machine verification method, device, equipment and medium, which improve the accuracy of man-machine verification.

[0005] In a first aspect, the embodiments of the application provide a man-machine verification method, applied to a mobile terminal, wherein the mobile terminal is provided with a motion sensor;

[0006] The method comprises:

[0007] acquiring first motion data of the mobile terminal collected by the motion sensor;

[0008] generating a motion graph according to the first motion data;

[0009] querying in a database according to the motion graph to determine the matching degree of the motion graph and historical data in the database;

[0010] in a case where the matching degree is greater than a first preset threshold and less than a second preset threshold, generating a target trajectory according to the motion graph;

[0011] sending the target trajectory to the mobile terminal, so that the operation object of the mobile terminal feeds back according to the target trajectory;

[0012] acquiring motion feedback information of the mobile terminal collected by the motion sensor in real time;

[0013] determining whether the operation object is a natural person according to the target trajectory and the motion feedback information, to obtain a man-machine verification result.

[0014] In a second aspect, the embodiments of the present application provide a man-machine verification device, applied to a mobile terminal, wherein the mobile terminal is provided with a motion sensor.

[0015] The device comprises:

[0016] a first obtaining module, configured to obtain first motion data of the mobile terminal collected by the motion sensor;

[0017] a first generating module, configured to generate a motion graph according to the first motion data;

[0018] a querying module, configured to query a database according to the motion graph to determine a matching degree between the motion graph and historical data in the database;

[0019] a second generating module, configured to generate a target trajectory according to the motion graph when the matching degree is greater than a first preset threshold and less than a second preset threshold;

[0020] a sending module, configured to send the target trajectory to the mobile terminal, so that an operation object of the mobile terminal feeds back according to the target trajectory;

[0021] a second obtaining module, configured to obtain motion feedback information of the mobile terminal collected by the motion sensor in real time;

[0022] a determining module, configured to determine whether the operation object is a natural person according to the target trajectory and the motion feedback information, to obtain a man-machine verification result.

[0023] In a third aspect, the embodiments of the present application provide an electronic device, comprising a processor and a memory storing computer program instructions; the processor implements the steps of the man-machine verification method in any one of the embodiments of the first aspect when executing the computer program instructions.

[0024] In a fourth aspect, the embodiments of the present application provide a computer readable storage medium, which stores computer program instructions; the computer program instructions are executed by a processor to implement the steps of the man-machine verification method in any one of the embodiments of the first aspect.

[0025] The human-computer verification method, device, equipment and medium provided by the embodiments of the present application can obtain information collected by a motion sensor carried by a mobile terminal, generate a corresponding motion graph based on the obtained information, and perform matching degree judgment with historical data in a database, so as to realize preliminary judgment based on a non-perception mode, prevent interruption of the original operation of an operation object, and improve the experience of the operation object. In the case that the matching degree is within a preset range, a target trajectory for feedback of the operation object is generated based on the matching degree, and whether the operation object is a natural person is verified through motion feedback information of the user. Since the motion feedback information is motion feedback information of the target trajectory, and the AI technology maps spatial motion information into fusion data of the motion sensor, the calculation resources consumed are large, so an attacker cannot complete large-scale human-computer identification attacks by generating data through AI. Therefore, it is difficult for the AI technology to realize real-time motion feedback of the motion trajectory. Therefore, whether the operation object is a natural person or a non-natural person can be accurately identified through the motion feedback information, and the accuracy of human-computer verification is improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required to be used in the embodiments of the present application will be briefly introduced. Those skilled in the art can obtain other drawings according to these drawings without creating any creative labor.

[0027] Figure 1 is a flow diagram of a human-computer verification method provided by the embodiments of the present application;

[0028] Figure 2 is a structural diagram of a human-computer verification device provided by the embodiments of the present application;

[0029] Figure 3 is a structural diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION

[0030] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application more clear and apparent, the present application will be further described in detail below with reference to the drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, but not to limit the present application. The present application can be implemented without some of these specific details by those skilled in the art. The following description of the embodiments is only intended to provide a better understanding of the present application by showing examples of the present application.

[0031] It is to be noted that the relative terms such as first and second and the like are used herein solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0032] It should be noted that the acquisition, storage, use and processing of data in the embodiments of the present application comply with the relevant provisions of national laws and regulations.

[0033] Gait cycle, used to represent the process of one side of the foot landing to the foot landing again, is usually expressed in time seconds (s). The gait cycle of an adult is about 1-1.32 s. Each gait cycle in walking contains a series of typical posture transitions, which are usually divided into a series of time periods, called gait phase. A gait cycle can be divided into a support phase and a swing phase, and can be further divided into eight phases, which are usually expressed in percentage of gait cycle (GC%) or seconds (s).

[0034] Weight decay, a regularization technique, is used to solve overfitting. Overfitting is a statistical term used to describe the phenomenon of fitting too closely or accurately to a specific data set, so as to be unable to well fit other data or predict future observations. Therefore, weight decay is used to represent that there are multiple parameters, and the weights of all parameters are the same at the beginning. When a parameter is too active and has too much influence on the overall judgment, even the influence degree of the parameter is decisive, overfitting will occur. Therefore, when a parameter has too much influence, its importance, i.e. weight, should be reduced, so as to avoid meaningless existence of other parameters.

[0035] Atomic action, also known as atomic visual action, originates from the computer term atomic operation. A plurality of operations constitute an operation. If the operation cannot be executed atomically, either all steps are executed or none is executed. It is impossible to execute only a subset of all steps.

[0036] In many scenarios, real natural persons need to participate in interactions, but due to the development of artificial intelligence, a large number of new means have appeared to simulate the behavior of natural person users, such as AI (artificial intelligence) robots simulating human behavior, resulting in security risks. Therefore, it is necessary to reduce risks by identifying the operation objects of the mobile terminal to realize human-computer verification.

[0037] In related technologies, human-computer verification is usually implemented based on a pre-trained identification model, however, since an attacker can generate relevant data to simulate human behavior through an AI algorithm, and input the data into the identification model to achieve the purpose of identifying AI as a human, that is, for the human-computer verification scheme in related technologies, there is a risk that AI behavior can be disguised as human behavior to pass the human-computer verification and crack the human-computer verification system.

[0038] That is, based on a pre-trained model, the way of determining the identity of the verifier by matching during verification requires a large amount of data samples, and the purpose is to identify the identity by inputting data. If the attacker understands the intention of identification, he can generate relevant sensor data to simulate human behavior through an AI algorithm, and bypass the workflow of the sensor to directly input the data disguised as sensor-generated data into the identification model. Although the AI-generated data cannot correspond to the pre-recorded posture-sensor data in the background database, and thus cannot achieve the purpose of disguising as a specific identity, it can achieve the purpose of identifying AI as a human, that is, AI behavior can be disguised as human behavior during the process of human-computer verification, thereby passing the human-computer verification.

[0039] Alternatively, in related technologies, a human-computer verification is performed by using an interactive mode of graphic-text recognition, so that the user performs a relevant operation after identifying the true intention of the graphic-text, which is poor in user experience; or a relatively fixed verification mode or verification process is used, for example, by listening to whether the user pulls a progress bar, clicks a verification point, slides a screen, drags a control on the page, inputs a verification code, and the like to determine whether the user is a human or a machine, which is easy to be cracked by AI and easily bypassed.

[0040] To solve the problems in related technologies, the embodiments of the present application provide a human-computer verification method, device, equipment and medium.

[0041] The human-computer verification method provided by the embodiments of the present application will be described in detail below in combination with the drawings, specific embodiments and application scenarios.

[0042] Figure 1 A flowchart of a human-computer verification method 100 according to an embodiment of the present application is shown. As shown in Figure 1 The human-computer verification method 100 is applied to a mobile terminal, and the mobile terminal is provided with a motion sensor. The human-computer verification method 100 can specifically include the following steps:

[0043] S101, acquire first motion data of the mobile terminal collected by a motion sensor;

[0044] S102, generate a motion graph according to the first motion data;

[0045] S103, query in a database according to the motion graph to determine a matching degree of the motion graph and historical data in the database;

[0046] S104, in a case where the matching degree is greater than a first preset threshold and less than a second preset threshold, generate a target trajectory according to the motion graph;

[0047] S105, send the target trajectory to the mobile terminal, so that an operation object of the mobile terminal feeds back according to the target trajectory;

[0048] S106, acquire motion feedback information of the mobile terminal collected by the motion sensor in real time;

[0049] S107, determine whether the operation object is a natural person according to the target trajectory and the motion feedback information, to obtain a man-machine verification result.

[0050] Therefore, by acquiring information collected by a motion sensor carried by a mobile terminal, and generating a corresponding motion graph from the acquired information to determine a matching degree with historical data in a database, a preliminary judgment based on a non-awareness mode is realized, the original operation of an operation object is prevented from being interrupted, and the experience of the operation object is improved. In a case where the matching degree is within a preset range, a target trajectory for feedback of the operation object is generated based on the matching degree, and whether the operation object is a natural person is verified through motion feedback information of the user. Since the motion feedback information is motion feedback information of the target trajectory, and AI technology maps spatial motion information into fusion data of the motion sensor, a large amount of computing resources are consumed, so an attacker cannot complete a large-scale man-machine identification attack by generating data through AI. Therefore, it is difficult for AI technology to realize real-time motion feedback of the motion trajectory. Therefore, through the motion feedback information, it can be accurately identified whether the operation object is a natural person or a non-natural person, and the accuracy of man-machine verification is improved.

[0051] The specific implementation of each step will be described below.

[0052] It should be noted that the mobile terminal in the embodiments of the present application can include a smart phone, a VR glasses, a handheld terminal device or a wearable device, or other large-scale controllable sensor-carrying instrument devices, and the embodiments do not make specific limitations.

[0053] In addition, the mobile terminal in the embodiment of the present application is provided with a motion sensor, and considering the universal applicability of the human-computer verification method of the embodiment of the present application, the motion sensor can include an accelerometer, a gyroscope, etc., because the terminal device with these sensors is relatively universal at present.

[0054] In some embodiments, in step S101, the motion sensor carried by the mobile terminal collects the spatial motion information thereof in real time. When the mobile terminal is in a target scene, for example, enters a pre-set service page (target page), or meets other pre-set conditions, the collection signal of the motion sensor is acquired and defined as first motion data. Specifically, since the motion sensor can include an accelerometer, a gyroscope, etc., the first motion data can include linear acceleration information and angular velocity change information.

[0055] In some embodiments, before step S102, according to the first motion data, the state information of the operation object of the mobile terminal is determined; in the case that the state information of the operation object is static, the invalid data in the first motion data that does not meet the demand of the target scene is determined according to the demand of the target scene, and these invalid data are removed from the first motion data. The state information of the operation object can include motion and static.

[0056] It should be understood that when the state information of the operation object is motion, the corresponding mobile terminal will generate relatively obvious acceleration signals within a certain time period because the operation object is in a motion state. Therefore, the moving speed of the operation object can be calculated according to the acceleration information collected by the accelerometer, and the first motion data is generated in real time by fusing the acceleration information and the angular velocity change information collected by the gyroscope to judge the attitude change of the terminal device, so as to judge whether the operation object is in a motion (for example, walking) state according to the first motion data.

[0057] Or, when the state information of the operation object is static, but the corresponding mobile terminal can have spatial motion within a certain range, which can include two types of motion: the first type is that the operation object needs to interact according to the demand of the target scene or the requirement of the service, which can be determined according to the state of the operation object in the business scene because it is related to the service; the second type is unrelated to the service and cannot be identified and determined according to the operation object information in the business scene, but it can be understood that since the operation object will not swing or shake the mobile device meaninglessly when interacting (for example, reading and understanding the information conveyed by the terminal device), the meaningless motion of the mobile terminal can be identified.

[0058] For example, in the case that the target scenario is that the operation object reads the content, the mobile terminal should generate a few discontinuous and directionally discrete signals according to the natural swing of the body of the operation object, and should not generate a large number of acceleration signals; or in the case that the target scenario is that the operation object moves (for example, advances, turns around, etc.), the mobile terminal should generate a relatively continuous linear velocity signal and a relatively uniform angular velocity signal.

[0059] In addition, when the operation object is in a walking movement, the first motion data in a certain time range can be generated into a plurality of gait cycles according to the periodic characteristics of the walking movement; when the operation object is in a stationary state, whether the operation object is in an interactive state can be determined in real time according to the spatial movement and posture transformation of the terminal device, and when it is determined that the operation object is not in an interactive state, the data collection should be stopped, and when it is determined that the operation object is in an interactive state, the data collection can be triggered according to the device movement generated by the interaction of the operation object in the business scenario, for example, the VR user determines or flips the page in the scene by nodding or shaking his head, and uploads the collected data for subsequent matching verification.

[0060] In addition, since meaningless movement of the mobile terminal is not conducive to interaction, the data generated by meaningless movement is invalid data, and these invalid data need to be discarded. That is, the invalid data is removed from the first motion data.

[0061] In this way, when the operation object is in a stationary state, the rationality of the stationary state can be determined according to the scenario in which the operation object is located, for example, the motion scenario should not be in a stationary state, and the operation object will necessarily change the posture of the terminal device when processing related businesses. Therefore, the acquired motion information (i.e., the first motion data) can be removed from the invalid data to achieve denoising.

[0062] Further, in some embodiments, in step S102, a graph model, i.e., a motion graph, is constructed according to the linear acceleration information and the angular velocity change information in the first motion data, which maps the spatial position change of the mobile terminal.

[0063] Specifically, the motion graph includes second motion data corresponding to each of three directions (i.e., XYZ axes of a three-dimensional space), and the straight lines of the three directions are perpendicular to each other. The second motion data can include the speed, acceleration, angular velocity in the corresponding direction, and the change information of these parameters over time.

[0064] In some embodiments, in step S103, whether there is data matching the motion graph in the historical data of the database is queried, so as to determine the matching degree of the motion graph and the historical data in the database.

[0065] For example, when the operation object is walking, a motion atlas is generated by using the gait information generated by the terminal, and the generated motion atlas is input into the gait information database to analyze the motion atlas, so as to query whether similar gait information already exists in the database. The gait information includes gait data in a period of time (i.e., multiple gait cycles).

[0066] Therefore, in some embodiments, in the case where the matching degree is greater than the first preset threshold and less than the second preset threshold, the target trajectory is generated according to the motion atlas. Alternatively, the matching degree can be set to three levels of high, medium and low, 30% matching degree is low, 60% is medium, and 80% is high. In the case where the matching degree is medium, the target trajectory is generated according to the motion atlas.

[0067] In addition, in some embodiments, in the case where the matching degree is less than or equal to the first preset threshold, it is determined that the operation object of the mobile terminal is a natural person to obtain a man-machine verification result, and the corresponding motion atlas and the mobile terminal identifier are associated and stored in the database; in the case where the matching degree is greater than or equal to the second preset threshold, it is determined that the operation object of the mobile terminal is a non-natural person to obtain a man-machine verification result.

[0068] In another embodiment, in the case where the matching degree is greater than or equal to the second preset threshold, the matching data corresponding to the motion atlas and the terminal information corresponding to the matching data based on the pre-established association relationship are further determined; it is judged whether the terminal information is consistent with the identifier of the mobile terminal; when the terminal information is consistent with the identifier of the mobile terminal, the target trajectory can be further generated according to the motion atlas.

[0069] Therefore, the present embodiment adopts the method of large data repetition verification of generated data to perform large-scale man-machine verification, which can prevent single user from repeatedly passing the verification or multiple users from simultaneously passing the verification at a time based on the pre-trained sample set by using AI.

[0070] In this way, the matching degree judgment can realize man-machine interaction based on a non-awareness mode, so as to preliminarily judge whether the operation terminal device is a natural person or a non-natural person (such as AI), so as to realize man-machine verification. It should be noted that when the preliminary verification cannot be judged, a second verification is supplemented in a perceptual manner based on the generated target trajectory.

[0071] In addition, in some embodiments, the same initial score weight is set for the three directions of the motion atlas. In the case where the matching degree is greater than the first preset threshold and less than the second preset threshold, for the generation of the target trajectory, the initial score weight of the corresponding direction of the second motion data can be subjected to weight attenuation processing to obtain an updated score weight; and then the target trajectory is generated according to the updated score weight.

[0072] In implementation, the stronger the motion in a certain direction (e.g., the stronger the acceleration signal), the faster the weight of the direction decays.

[0073] In implementation, the score decay value of the direction corresponding to the second motion data can be determined, and then the initial score weight is attenuated according to the score decay value to obtain an updated score weight.

[0074] In implementation, for each direction, the corresponding atomic action can be determined in the pre-established atomic action library according to the updated score weight; thus, the standard verification action is determined by randomly combining the atomic actions corresponding to the three directions respectively; and then, the target trajectory is generated according to the standard verification action.

[0075] In implementation, the smaller the updated score weight of a certain direction, the more intense the motion in the direction, and the motion in the direction needs to be weakened, and vice versa.

[0076] In addition, in some embodiments, an atomic action library is constructed, including a plurality of atomic actions. The atomic action has atomicity, and the final value obtained according to the change of the score weight in the X, Y, and Z three axes during the action process is divided into three groups, i.e., X group, Y group, and Z group. The three groups can each be provided with 30 atomic actions, numbered 1-30, and the 1-30th actions are numbered from small to large in terms of strength.

[0077] Taking the Z group No. 1 atomic action as an example, the Z group atomic action indicates that the Z axis direction motion needs to be strengthened, Z1 indicates the weakest degree of strengthening, which can be understood as the motion signal intensity on the three axes being equivalent, but the Z axis motion signal is slightly stronger than the X and Y axes. Taking a handheld terminal device as an example in action design, the action is to hold up the phone from the head position as the starting position, stretch the single arm to the abdominal height, and the direction is the arm on the same side and the face direction at an angle of 45° as the end point. A Bezier curve is generated according to the starting point and the end point, and the handheld terminal device performs spatial motion according to the generated Bezier curve to complete the process from the starting point to the end point as one Z1 atomic action. Z30 indicates the strongest degree of strengthening, and Z30 is understood as the Z axis signal being much larger than the X axis signal and the Y axis signal in the three-axis motion. In action design, taking a head-mounted terminal (VR glasses) as an example, the head position after wearing the device is the starting point, the action is to squat vertically, and the head position after squatting is the end point. A straight line is generated from the starting point to the end point, and the process from the starting point to the end point is completed as one Z30 atomic action.

[0078] In implementation, the number of atomic actions can be set according to experimental experience and commonly used actions, i.e., actions without too much complexity and large distinction, which are not limited in the embodiment.

[0079] In specific implementation, for the determination of the atomic action, the corresponding atomic action can be selected according to the ratio of the maximum weight value to the initial weight value after completing the non-awareness check. For example, the initial weight of the X-axis is 33, the maximum weight is 11, and the selected atomic action of the X-axis is 11\33*30=10, that is, X10. Further, after selecting the corresponding atomic actions of the X, Y and Z axes respectively, the three atomic actions are combined in a random order to generate the standard check action of the current check, so as to generate the target trajectory according to the standard check action.

[0080] Therefore, in step S105, the target trajectory can be sent to the mobile terminal, so that the operation object of the mobile terminal feeds back according to the target trajectory; and in step S106, the motion feedback information of the mobile terminal collected by the motion sensor is acquired in real time.

[0081] In some embodiments, before step S107, an updated target trajectory can be generated according to the motion feedback information, and the step of sending the target trajectory to the mobile terminal is returned until the endpoint of the motion feedback information coincides with the endpoint of the target trajectory.

[0082] In specific implementation, the key check points can be randomly selected within a preset distance range of the target trajectory according to the motion feedback information; the key check points replace any one of the passing points in the target trajectory to obtain an updated target trajectory.

[0083] That is, these key check points are randomly distributed near the path of the standard check action, the key check points will replace the points in the standard trajectory, and a new Bezier curve is generated in real time, so that the check action and the standard check action have obvious differences in the local.

[0084] Moreover, these key points are not pre-generated, but dynamically generated. The basis for dynamic generation is to determine the motion that needs to be strengthened according to the corresponding multiple scoring indicators (conformity to the target trajectory, reaction rate, stay interval, speed, etc.) of the motion feedback information acquired in real time. For example, if the completion trajectory displayed by the motion feedback information has high conformity to the target trajectory, more and more key check points will be generated to make the subsequent action more and more different from the trajectory of the preset standard check action.

[0085] As can be seen, the atomic action group selected based on the result of the non-awareness check, and the random arrangement of the atomic action group will generate different standard check actions, and the performance when completing the standard action will generate the check action by real-time standard action deformation. Therefore, the check action that each operation object participating in the check needs to complete is completely different, that is, the completely dynamic check action.

[0086] Finally, in some embodiments, in step S107, the motion feedback information is scored according to the target trajectory to obtain a score result; in response to the score result being greater than or equal to a third preset threshold, it is determined that the operation object is a natural person to obtain a man-machine verification result; in response to the score result being less than the third preset threshold, it is determined that the operation object is a non-natural person to obtain the man-machine verification result.

[0087] That is, when the motion feedback information passes through a key checkpoint, the score information of the completion degree of the motion feedback information relative to the target trajectory is scored according to multiple indexes, and finally a score result is obtained, so that the man-machine judgment is made according to the score result, and finally the conclusion of the man-machine verification is made.

[0088] In this way, the combination of the non-perception verification process and the perception verification process is adopted, the non-perception process can complete the preliminary man-machine verification without the operation object being aware to prevent interrupting the original operation of the operation object, improve the interaction experience, and the non-perception man-machine verification can complete the man-machine verification of most operation objects, can save time cost, and improve the man-machine recognition effect and efficiency; the perception process generates the selected atomic action based on the preliminary verification result of the non-perception, and guides the operation object to move and track for man-machine verification in the way of dynamically generating the path in real time, since the verification action is completely dynamically generated and unpredictable, the AI simulation can be effectively prevented, and the accuracy of the man-machine verification is improved.

[0089] It should be noted that the above describes some embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims can be performed in an order different than the order described in the above embodiments and still achieve the desired result. In addition, the processes depicted in the figures do not necessarily require the particular order shown or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or possible.

[0090] Based on the same technical concept, corresponding to any of the above-mentioned embodiment methods, the present application also provides a man-machine verification device 200. Specifically, the man-machine verification device 200 is applied to a mobile terminal, and the mobile terminal is provided with a motion sensor.

[0091] As shown in FIG. 2, the man-machine verification device 200 can include: Figure 2

[0092] A first acquisition module 201 is configured to acquire first motion data of the mobile terminal collected by the motion sensor;

[0093] A first generation module 202 is configured to generate a motion atlas according to the first motion data;

[0094] ​The query module 203 is used to query the database based on the motion map to determine the degree of matching between the motion map and historical data in the database;

[0095] The second generation module 204 is used to generate a target trajectory based on the motion map when the matching degree is greater than a first preset threshold and less than a second preset threshold.

[0096] Sending module 205 is used to send the target trajectory to the mobile terminal so that the operation object of the mobile terminal can provide feedback based on the target trajectory;

[0097] The second acquisition module 206 is used to acquire motion feedback information of the mobile terminal collected by the motion sensor in real time;

[0098] The determination module 207 is used to determine whether the operation object is a natural person based on the target trajectory and motion feedback information, so as to obtain the human-machine verification result.

[0099] In some embodiments, the motion map includes second motion data corresponding to each of three directions, the lines containing the three directions are perpendicular to each other, and the three directions correspond to the same initial scoring weight.

[0100] In some embodiments, the second generation module 204 is specifically used to perform weight decay processing on the initial score weight of the corresponding direction based on the second motion data when the matching degree is greater than a first preset threshold and less than a second preset threshold, to obtain an updated score weight; and to generate a target trajectory based on the updated score weight.

[0101] In some embodiments, the second generation module 204 is further configured to, when the matching degree is greater than a first preset threshold and less than a second preset threshold, determine a score decay value corresponding to the direction based on the second motion data; and decay the initial score weight based on the score decay value to obtain an updated score weight.

[0102] In some embodiments, the human-machine verification device 200 further includes an update module ( Figure 2 (Not shown in the image), the update module is used to generate an updated target trajectory based on the motion feedback information, and return to the step of sending the target trajectory to the mobile terminal until the endpoint of the motion feedback information coincides with the endpoint of the target trajectory.

[0103] In some embodiments, the updating module is specifically used to randomly select key verification points within a preset distance range of the target trajectory based on the motion feedback information; and replace any one of the waypoints in the target trajectory with the key verification points to obtain an updated target trajectory.

[0104] In some embodiments, the updating module is further configured to, for each direction, determine the corresponding atomic action in a pre-established atomic action library according to the updated scoring weight; randomly combine the atomic actions corresponding to the three directions to determine the standard verification action; and generate the target trajectory according to the standard verification action.

[0105] In some embodiments, the human-machine verification device 200 further includes a removal module ( Figure 2 (Not shown in the image), the removal module is used to determine the state information of the operation object of the mobile terminal based on the first motion data; when the state information of the operation object is stationary, it determines invalid data in the first motion data that does not meet the requirements of the target scene according to the requirements of the target scene, and removes the invalid data from the first motion data.

[0106] In some embodiments, the determining module 207 is specifically used to score the motion feedback information according to the target trajectory to obtain a scoring result; in response to the scoring result being greater than or equal to a third preset threshold, to determine that the operation object is a natural person to obtain a human-machine verification result; in response to the scoring result being less than the third preset threshold, to determine that the operation object is a non-natural person to obtain a human-machine verification result.

[0107] It should be noted that, for ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing this application, the functions of each module can be implemented in one or more software and / or hardware.

[0108] The apparatus described above is used to implement the corresponding human-machine verification method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0109] Based on the same technical concept, corresponding to any of the above embodiments, this application also provides an electronic device.

[0110] Figure 3 A schematic diagram of a more specific electronic device hardware structure provided in this embodiment is shown.

[0111] The electronic device 300 may include a processor 301 and a memory 302 storing computer program instructions.

[0112] Specifically, the processor 301 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0113] The memory 302 can include mass storage for data or instructions. As an example and not by way of limitation, the memory 302 can include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disc (e.g., a compact disc (CD) or a digital versatile disc (DVD)), a solid-state drive (SSD), a USB drive, or a combination of two or more of these. Where appropriate, the memory 302 can include removable or non-removable (or fixed) media, where appropriate. The memory 302 can be internal or external to the integrated gateway disaster recovery appliance. In particular embodiments, the memory 302 is non-volatile, solid-state memory.

[0114] In particular embodiments, the memory can include read-only memory (ROM), random access memory (RAM), magnetic disk storage mediums, optical storage mediums, flash memory devices, electrical, optical, or other physical / tangible memory storage devices. Thus, in general, the memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., a memory device) encoded with software that, when executed (by one or more processors), is operable to

[0115] The processor 301 implements any of the above-mentioned human verification methods by reading and executing the computer program instructions stored in the memory 302.

[0116] In some examples, the electronic device 300 can further include a communication interface 303 and a bus 310. As shown, the processor 301, the memory 302, and the communication interface 303 are connected through the bus 310 and complete communication among each other. Figure 3

[0117] The communication interface 303 is mainly used to realize the communication among the modules, devices, units and / or equipment in the embodiments of the present application.

[0118] ​Bus 310 includes hardware, software, or both, to couple components of the online data traffic billing device to each other in a known manner. Although embodiments of the application are not limited to these bus forms or bus forms generally, one example of a bus could be the Accelerated Graphics Port (AGP) or other graphics bus, Enhanced Industry Standard Architecture (EISA) bus, Front Side Bus (FSB), HyperTransport, Industry Standard Architecture (ISA) bus, InfiniBand Bus, Low Pin Count (LPC) bus, Memory Bus, Micro Channel Architecture (MCA) bus, Peripheral Component Interconnect (PCI) bus, PCI-Express (PCI-X) bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association local (VLB) bus, or another suitable bus or a combination of two or more of these busses. Where suitable, bus 310 can include one or more buses. Although specific busses are described and illustrated in the embodiments of the application, the application contemplates any suitable bus or interconnect.

[0119] By way of example, the electronic device 300 can be a mobile phone, a tablet computer, a notebook computer, a palm computer, an in-vehicle electronic device, an ultra-mobile personal computer (UMPC), a netbook, or a personal digital assistant (PDA), etc.

[0120] Based on the same technical concept, the present application also provides a non-transitory computer-readable storage medium corresponding to any of the above-mentioned embodiment methods. The computer-readable storage medium stores computer program instructions; the computer program instructions are executed by a processor to implement any of the above-mentioned embodiment human-computer verification methods. Examples of the computer-readable storage medium include non-transitory computer-readable storage media, such as portable discs, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), portable compact disc read-only memories (CD-ROM), optical storage devices, magnetic storage devices, etc.

[0121] Based on the same technical concept, the present application also provides a computer program product, which includes computer program instructions. In some embodiments, the computer program instructions can be executed by one or more processors of a computer to cause the computer and / or the processor to perform the human-computer verification method. Corresponding to the execution subject of each step in each embodiment of the human-computer verification method, the processor performing the corresponding step can belong to the corresponding execution subject.

[0122] It is to be understood that the application is not limited to particular configurations and processes described herein and shown in the drawings. The detailed description is not to be taken as limiting the application. In the above embodiments, several specific steps are described and illustrated in order to provide a thorough understanding of the application. However, the application can be practiced with fewer or additional steps, and in a different order. The application is not limited to the described and illustrated embodiments.

[0123] The functions shown in the structural block diagrams described above can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application specific integrated circuits (ASICs), appropriate firmware, plug-ins, functional cards, and the like. When implemented in software, the elements of the application are program or code segments that are used to perform the required tasks. The program or code segments can be stored in a machine-readable medium or transmitted through a data signal carried in a carrier wave over a transmission medium or communication link. A "machine-readable medium" includes any medium that can store or transfer information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, and the like. The code segments can be downloaded via computer networks such as the Internet, intranets, and the like.

[0124] It is also to be understood that the example embodiments described in this application are based on a series of steps or apparatuses to describe some methods or systems. However, the application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiments, or in an order different from the embodiments, or several steps can be performed simultaneously.

[0125] The above-described aspects of the application are described above with reference to flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / acts specified in the flowcharts and / or block diagrams. Such a processor can be, but is not limited to, a general purpose processor, a special purpose processor, a special purpose application specific processor, or a field programmable logic array. It should also be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by dedicated hardware-based systems that perform the specified functions or acts, or combinations of hardware and software.

[0126] The above merely describes a specific implementation of the present application. Those skilled in the art can clearly understand the specific working processes of the system, modules and units described above for the convenience and brevity of description, and the corresponding processes in the foregoing method embodiments can be referred to, which will not be described herein again. It should be understood that the protection scope of the present application is not limited in this way, and any person skilled in the art can easily think of various equivalent modifications or replacements within the technical range disclosed in the present application, and these modifications or replacements should be covered in the protection scope of the present application.

Claims

1. A method of human-in-the-loop verification, the method comprising: The application is applied to a mobile terminal provided with a motion sensor; The method comprises: acquiring first motion data of the mobile terminal collected by the motion sensor; generating a motion graph according to the first motion data; querying in a database according to the motion graph to determine a matching degree of the motion graph and historical data in the database; generating a target trajectory according to the motion graph in a case that the matching degree is greater than a first preset threshold and less than a second preset threshold; sending the target trajectory to the mobile terminal to make an operation object of the mobile terminal feedback according to the target trajectory; acquiring motion feedback information of the mobile terminal collected by the motion sensor in real time; determining whether the operation object is a natural person according to the target trajectory and the motion feedback information to obtain a man-machine verification result.

2. The method of claim 1, wherein, Before the step of determining whether the operation object is a natural person according to the target trajectory and the motion feedback information to obtain a man-machine verification result, the method further comprises: generating an updated target trajectory according to the motion feedback information, and returning to the step of sending the target trajectory to the mobile terminal until the end point of the motion feedback information coincides with the end point of the target trajectory.

3. The method of claim 1, wherein, The motion graph comprises three second motion data corresponding to three directions respectively, and the straight lines where the three directions are located are perpendicular to each other respectively, and the three directions correspond to the same initial scoring weight; The step of generating a target trajectory according to the motion graph in a case that the matching degree is greater than a first preset threshold and less than a second preset threshold comprises: In the case that the matching degree is greater than a first preset threshold and less than a second preset threshold, performing weight attenuation processing on the initial scoring weight of the direction corresponding to the second motion data to obtain an updated scoring weight; generating a target trajectory according to the updated scoring weight.

4. The method of claim 2, wherein, The step of generating an updated target trajectory according to the motion feedback information comprises: randomly selecting a key verification point within a preset distance range of the target trajectory according to the motion feedback information; replacing the key verification point with any one passing point in the target trajectory to obtain an updated target trajectory.

5. The method of claim 3, wherein, The step of performing weight attenuation processing on the initial scoring weight of the direction corresponding to the second motion data to obtain an updated scoring weight in a case that the matching degree is greater than a first preset threshold and less than a second preset threshold comprises: determining a scoring decay value of the direction corresponding to the second motion data according to the second motion data in the case that the matching degree is greater than a first preset threshold and less than a second preset threshold; attenuating the initial scoring weight according to the scoring decay value to obtain an updated scoring weight.

6. The method of claim 3, wherein, The step of generating a target trajectory according to the updated scoring weight comprises: for each direction, determining a corresponding atomic action in a pre-established atomic action library according to the updated scoring weight; randomly combining the atomic actions corresponding to the three directions respectively to determine a standard verification action; generating a target trajectory according to the standard verification action.

7. The method of claim 1, wherein, Before the step of generating a motion graph according to the first motion data, the method further comprises: According to the first motion data, state information of an operation object of the mobile terminal is determined; In a case where the state information of the operation object is static, invalid data in the first motion data that does not meet a target scene requirement is determined according to the target scene requirement, and the invalid data is removed from the first motion data.

8. The method of claim 1, wherein, After the matching degree is determined according to the motion graph and the historical data in the database, the method further comprises: In a case where the matching degree is greater than or equal to a second preset threshold, it is determined that the operation object of the mobile terminal is a non-natural person, to obtain a man-machine verification result.

9. The method of claim 1, wherein, After the matching degree is determined according to the motion graph and the historical data in the database, the method further comprises: In a case where the matching degree is less than or equal to a first preset threshold, it is determined that the operation object of the mobile terminal is a natural person, to obtain a man-machine verification result.

10. The method of claim 1, wherein, The determining whether the operation object is a natural person according to the target trajectory and the motion feedback information to obtain a man-machine verification result comprises: The motion feedback information is scored according to the target trajectory, to obtain a scoring result; In response to the scoring result being greater than or equal to a third preset threshold, it is determined that the operation object is a natural person, to obtain a man-machine verification result; in response to the scoring result being less than the third preset threshold, it is determined that the operation object is a non-natural person, to obtain a man-machine verification result.

11. A man-machine verification device, characterized by The method is applied to a mobile terminal, and the mobile terminal is provided with a motion sensor. The device comprises: A first obtaining module is configured to obtain first motion data of the mobile terminal collected by the motion sensor; A first generating module is configured to generate a motion graph according to the first motion data; A querying module is configured to query, according to the motion graph, a database to determine a matching degree of the motion graph and historical data in the database; A second generating module is configured to generate a target trajectory according to the motion graph in a case where the matching degree is greater than a first preset threshold and less than a second preset threshold; A sending module is configured to send the target trajectory to the mobile terminal, so that an operation object of the mobile terminal feeds back according to the target trajectory; A second obtaining module is configured to obtain motion feedback information of the mobile terminal collected by the motion sensor in real time; A determining module is configured to determine whether the operation object is a natural person according to the target trajectory and the motion feedback information, to obtain a man-machine verification result.

12. An electronic device, comprising: The device comprises a processor and a memory storing computer program instructions; the processor invokes the computer program instructions to implement the man-machine verification method of any one of claims 1-10.

13. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer program instructions, and the computer program instructions are invoked by the processor to implement the man-machine verification method of any one of claims 1-10.

Images