Method, device, medium and equipment for detecting robots in games

Abstract

The embodiment of the present application provides a method, apparatus, storage medium and terminal device for detecting robots in a game, the method comprising: obtaining trajectory data of the virtual object moving in the game scene; obtaining corresponding semantic sequence data and topological structure data according to the trajectory data; obtaining a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological structure data, and a preset neural network; obtaining a judgment result of whether the virtual object is a robot according to the target trajectory information representation. The embodiment of the present application extracts semantic sequence data and topological structure data that can effectively reflect the characteristics of the virtual object from the trajectory data of the virtual object, and fuses the two to obtain a target trajectory information representation that integrates information of different dimensions, and judges whether the client account corresponding to the virtual object has violated the rules based on the target trajectory information representation.

Description

Technical Field

[0001] The present application relates to the field of electronic communication technology, and more particularly to the field of technology for detecting robots in games, and more particularly to a method, device, medium and equipment for detecting robots in games. Background Art

[0002] At present, various intelligent services are usually deployed in games, such as game robot detection, social recommendation, etc. In recent years, how to detect game robots in online games has gradually become the focus of game technology developers. The existence of game robots seriously undermines the fairness of the game and seriously reduces the game fun of ordinary players. In addition, the existence of game robots destroys the rules of item trading in the game, which devalues ​​the virtual property of players and increases the difficulty of operation of game companies. Therefore, detecting game robots is a win-win situation. Game developers can guarantee profits and players can also enjoy it.

[0003] Therefore, there is an urgent need for an efficient method to detect robots in games. Summary of the invention

[0004] The embodiments of the present application provide a method, apparatus, medium and equipment for detecting robots in a game, by extracting semantic sequence data and topological structure data that can effectively reflect the characteristics of the virtual object from the trajectory data of the virtual object, and fusing the two to obtain a target trajectory information representation that integrates information of different dimensions, and based on the target trajectory information representation, judging whether the client account corresponding to the virtual object has any violation.

[0005] In one aspect, an embodiment of the present application provides a method for detecting a robot in a game, wherein a graphical user interface is provided through a screen of a terminal device, and content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object. The method includes:

[0006] Acquire trajectory data of the virtual object moving in the game scene;

[0007] Acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data;

[0008] Based on the semantic sequence data and the topological graph structure data, and a preset neural network, a target trajectory information representation corresponding to the virtual object is obtained;

[0009] A judgment result of whether the virtual object is a robot is obtained according to the target trajectory information representation.

[0010] In the method for detecting a robot in a game described in an embodiment of the present application, the step of obtaining corresponding semantic sequence data according to the trajectory data includes:

[0011] Dividing the scene image into a plurality of cells;

[0012] The cell into which any trajectory point falls is determined according to the trajectory data, and the semantic sequence data is obtained according to the cell into which the trajectory point falls.

[0013] In the method for detecting a robot in a game described in an embodiment of the present application, the method further includes:

[0014] Assigning corresponding numbers to the multiple cells respectively;

[0015] The obtaining of semantic sequence data according to the cell into which the trajectory point falls includes:

[0016] The semantic sequence data is obtained according to the number of the cell into which the trajectory point falls.

[0017] In the method for detecting a robot in a game described in an embodiment of the present application, obtaining a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological graph structure data, and a preset neural network, includes:

[0018] Obtaining a first vector representation of a data node in the topological graph structure data and degree information of the topological graph structure data;

[0019] Obtaining a graph structure representation of the virtual object according to the first vector representation and the degree information, and a preset graph neural network;

[0020] Obtaining a second vector representation of the node data in the semantic sequence data;

[0021] Obtaining a behavior intention representation of the virtual object according to the second vector representation and a preset neural network;

[0022] The semantic sequence data, the graph structure representation and the intention representation are input into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object.

[0023] In the method for detecting a robot in a game described in an embodiment of the present application, the method further includes:

[0024] Obtaining degree information of the topological graph structure data;

[0025] The obtaining, according to the first vector representation and a preset graph neural network, a graph structure representation of the virtual object includes:

[0026] According to the first vector representation and the degree information, and a preset graph neural network, a graph structure representation of the virtual object is obtained.

[0027] In the method for detecting a robot in a game described in an embodiment of the present application, obtaining the behavior intention representation of the virtual object according to the second vector representation and a preset neural network includes:

[0028] According to the second vector representation and the preset GRU neural network, the behavior intention representation of the virtual object is obtained.

[0029] In the method for detecting a robot in a game described in an embodiment of the present application, the semantic sequence data, the graph structure representation, and the intention representation are input into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object, including:

[0030] The semantic sequence data, the graph structure representation and the intention representation are input into a preset fusion unit, and the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation to obtain the target trajectory information representation corresponding to the virtual object.

[0031] In the method for detecting a robot in a game described in an embodiment of the present application, the trajectory data is data obtained after filtering the initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

[0032] In the method for detecting a robot in a game described in an embodiment of the present application, the step of obtaining a judgment result of whether the virtual object is a robot according to the target trajectory information representation includes:

[0033] Processing the target trajectory information representation by a normalized exponential function to obtain a probability that the virtual object is a robot;

[0034] A judgment result of whether the virtual object is a robot is obtained according to the probability.

[0035] Accordingly, another aspect of the embodiment of the present application further provides a device for detecting a robot in a game, wherein a graphical user interface is provided through a screen of a terminal device, and content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object, and the device includes:

[0036] A trajectory data acquisition module, used to acquire trajectory data of the virtual object moving in the game scene;

[0037] An intermediate data acquisition module, used to acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data;

[0038] A target data acquisition module, used to obtain a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological graph structure data, and a preset neural network;

[0039] The judgment result acquisition module is used to obtain a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0040] Correspondingly, on the other hand, an embodiment of the present application further provides a computer-readable storage medium, which stores a plurality of instructions, and the instructions are suitable for a processor to load to execute the method for detecting a robot in a game as described above.

[0041] Correspondingly, on the other hand, an embodiment of the present application further provides a terminal device, including a processor and a memory, wherein the memory stores a plurality of instructions, and the processor loads the instructions to execute the method for detecting a robot in a game as described above.

[0042] The embodiments of the present application provide a method, device, medium and equipment for detecting robots in games, wherein the method provides a graphical user interface through the screen of a terminal device, wherein the content displayed by the graphical user interface includes a scene image of a game scene, wherein the game scene includes a virtual object, and the method includes: obtaining trajectory data of the virtual object moving in the game scene; obtaining corresponding semantic sequence data and topological structure data according to the trajectory data; obtaining a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological structure data, and a preset neural network; obtaining a judgment result of whether the virtual object is a robot according to the target trajectory information representation. The embodiments of the present application extract semantic sequence data and topological structure data that can effectively reflect the characteristics of the virtual object from the trajectory data of the virtual object, and fuse the two to obtain a target trajectory information representation that integrates information of different dimensions, and judges whether the client account corresponding to the virtual object has any violations based on the target trajectory information representation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application, and those skilled in the art can obtain other drawings based on these drawings without creative work.

[0044] Figure 1 A flowchart of a method for detecting robots in a game provided in an embodiment of the present application.

[0045] Figure 2A schematic flowchart of the process of constructing a target trajectory information representation in a method for detecting a robot in a game provided in an embodiment of the present application.

[0046] Figure 3 A schematic diagram of the structure of an apparatus for detecting robots in a game provided in an embodiment of the present application.

[0047] Figure 4 Another schematic diagram of the structure of the device for detecting robots in games provided in an embodiment of the present application.

[0048] Figure 5 A schematic diagram of the structure of a terminal device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0050] The present application provides a method for detecting a robot in a game, and the method for detecting a robot in a game can be applied to a terminal device, which can be a smart phone, a tablet computer, or other device.

[0051] At present, various intelligent services are usually deployed in games, such as game robot detection, social recommendation, etc. In recent years, how to detect game robots in online games has gradually become the focus of game technology developers. The existence of game robots seriously undermines the fairness of the game and seriously reduces the game fun of ordinary players. In addition, the existence of game robots destroys the rules of item trading in the game, which devalues ​​the virtual property of players and increases the difficulty of operation of game companies. Therefore, detecting game robots is a win-win situation. Game developers can guarantee profits and players can also enjoy it.

[0052] However, the detection of gaming robots is often limited by the traditional steps of machine learning, and usually requires tedious feature engineering (such as manually constructing user portraits based on expert experience). The feature engineering in different games is also different, that is, the features contained in different games are not the same. This leads to the lack of generalization of current detection schemes, low recognition accuracy, and time-consuming and labor-intensive.

[0053] In order to solve the above technical problems, the embodiment of the present application provides a method for detecting robots in games. Using the method for detecting robots in games provided by the embodiment of the present application, semantic sequence data and topological structure data that can effectively reflect the characteristics of virtual objects are extracted from the trajectory data of virtual objects, and the two are fused to obtain a target trajectory information representation that integrates information of different dimensions, and based on the target trajectory information representation, it is judged whether the client account corresponding to the virtual object has violated the rules.

[0054] See also Figure 1-Figure 2 , Figure 1 A flowchart of a method for detecting robots in a game provided in an embodiment of the present application. Figure 2 A schematic flowchart of the process of constructing a target trajectory information representation in a method for detecting a robot in a game provided in an embodiment of the present application.

[0055] The method for detecting a robot in a game is applied to a terminal device, and a graphical user interface is provided through a screen of the terminal device. The content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object. The method may include the following steps:

[0056] Step 101, obtaining trajectory data of the virtual object moving in the game scene.

[0057] It should be explained that, taking the game scene as an example, the virtual object refers to the game character in the game scene. In this embodiment, the trajectory data corresponding to the virtual object means that when the game user needs to control the game character to move or release skills in a specific area, it is necessary to move the mouse cursor while clicking the mouse button. At this time, each time the game user clicks the mouse button, the current position coordinates of the mouse cursor will be recorded, such as (x, y). When the game user continuously moves the mouse cursor and presses the mouse button, multiple recorded position coordinates constitute the trajectory data. That is, in this embodiment, the trajectory data is the data obtained after filtering the initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

[0058] In other embodiments, in addition to controlling the game character with a mouse, the game character may also be controlled to move or release skills by using keyboard keys or directly touching the touch screen.

[0059] Step 102: Acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data.

[0060] In this embodiment, the semantic sequence data and the topological graph structure data need to be explained in detail.

[0061] For the semantic sequence data, after the trajectory data is converted into the semantic sequence data, the semantic sequence data is subjected to a first vector representation extraction operation through a first vector representation extraction model to obtain a representation of whether the virtual object moves according to a preset movement trajectory.

[0062] It should be explained that the trajectory data in this embodiment is composed of position coordinates. For example, a sample can be [(xt1, yt1), (xt2, yt2), ..., (xtn, ytn)], where xtn represents the horizontal coordinate of the mouse cursor at time tn in the game scene when the game user clicks the mouse button, and ytn represents the vertical coordinate of the mouse cursor at time tn in the game scene when the game user clicks the mouse button.

[0063] Set a suitable trajectory length K (for example, k = 1000), fill the trajectory with 0 for insufficient length, and take the first 1000 position coordinates for the trajectory with a length of more than 1000. Since the position coordinates of the trajectory data appear on the display screen of the display terminal, in order to discretize the continuous values ​​(i.e., the position coordinates), it is necessary to divide the virtual area under the virtual scene into n×m cells, and number each cell according to the arrangement order of the cells in the virtual area, for example, from the upper left corner of the virtual area to 1, 2, 3...n, n+1,..., n×m from left to right. Based on the virtual area, the coordinate data extracted from the moving virtual object in the virtual area are distributed in the corresponding cells in the virtual area, and the number of each coordinate data in the cell where it is located is recorded, for example, 1, 2, 3. Replace the position coordinate point falling in the corresponding cell with the corresponding cell number to obtain the semantic sequence data [v1, v2,......vn]. At this time, the trajectory data originally composed of two-dimensional data is reduced in dimension to obtain semantic sequence data composed of one-dimensional data, thereby simplifying the data processing process and improving the data processing speed.

[0064] In this embodiment, the process of acquiring semantic sequence data can be implemented by the following steps:

[0065] Dividing the scene image into a plurality of cells;

[0066] The cell into which any trajectory point falls is determined according to the trajectory data, and the semantic sequence data is obtained according to the cell into which the trajectory point falls.

[0067] In further explanation, the process of acquiring semantic sequence data also includes the following steps:

[0068] Assigning corresponding numbers to the multiple cells respectively;

[0069] Based on this step, the step of "obtaining semantic sequence data according to the cell where the trajectory point falls" is explained and can be converted into: obtaining semantic sequence data according to the number of the cell where the trajectory point falls.

[0070] For topological graph structure data, it is necessary to understand that although the recursive neural network is specially designed for variable-length sequence data, it has functions such as forgetting and memory through the gated unit, and can mine the semantic information existing in the sequence data. At the same time, it was found in the research process that the game robot has distinct pattern characteristics, such as random displacement around a position, or a pattern of going back and forth between several coordinate points. For such a simple pattern, it is not sufficient to use a recursive neural network to capture its information. Therefore, in this embodiment, by representing the semantic sequence data as topological graph structure data, it can help the model to more easily identify the game robot. For example, a robot that goes back and forth between coordinate points will find that the number of degrees between these points is particularly large under the representation of graph structure data. In simple terms, the semantic information of "moving back and forth between these points" is converted into the size of the "degree between several points", that is, the transformation from semantic information to more precise numerical values, and numerical information is easier for the model to identify and detect.

[0071] Specifically, the virtual area under the virtual scene is divided into n×m cells, and each cell is numbered according to the arrangement order of the cells in the virtual area; the cell into which any trajectory point falls is determined according to the trajectory data, and semantic sequence data is obtained according to the cell into which the trajectory point falls; and the topological map structure data is constructed based on each of the numbers in the semantic sequence data.

[0072] Furthermore, the topological map structure data is constructed based on each of the numbers in the semantic sequence data, including: determining the sub-numbers in the semantic sequence data that are adjacent to each number, and representing the numbers and sub-numbers with a preset pattern; taking one of the numbers as the center number, and pointing the sub-numbers that are adjacent to the center number to the center number through arrow lines to construct the topological map structure data.

[0073] In some embodiments, after pointing the sub-numbers having an adjacent relationship with the central number to the central number through arrow lines, the method further includes:

[0074] The number of times each sub-number of the central number points to the central number is counted respectively, and the corresponding marks are made on each sub-number in the topological structure data, so that the semantic information of "moving back and forth between these points" is converted into the size of the "degree between several points", that is, the transformation from semantic information to more precise numerical values, and numerical information is easier to identify and detect for the model.

[0075] Step 103: Based on the semantic sequence data and the topological graph structure data, and a preset neural network, obtain a target trajectory information representation corresponding to the virtual object.

[0076] After obtaining the semantic sequence data and the topological graph structure data, the target trajectory information representation is obtained by fusing the two features, and the advantages of the features of the two different dimensions can be combined to construct a target trajectory information representation with more complete information expression. In order to reasonably utilize the two types of information, in this embodiment, a gated neural network is used to adaptively learn to select different proportions of information representation for different samples.

[0077] Specifically, obtaining a first vector representation of a data node in the topological graph structure data and degree information of the topological graph structure data;

[0078] Obtaining a graph structure representation of the virtual object according to the first vector representation and the degree information, and a preset graph neural network;

[0079] Obtaining a second vector representation of the node data in the semantic sequence data; specifically, the semantic sequence data may be input into a recursive neural network model to perform a second vector representation extraction operation to obtain the second vector representation;

[0080] Obtaining a behavior intention representation of the virtual object according to the second vector representation and a preset neural network (such as a GRU neural network);

[0081] The semantic sequence data, the graph structure representation and the intention representation are input into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object.

[0082] It should be noted that in actual game scenarios, the trajectory data generated by game users operating the mouse or clicking the screen usually contains semantic information. For example, game players can control the game character from location A to location B to obtain a key item, or when the game user controls the game character to fight, the game user may use the mouse to control the game character to dodge, attack, etc. The resulting trajectory data contains semantic information, which explains the movement or action mode of the corresponding game character in the trajectory data. The movement or action mode is irregular and is usually determined according to the actual scenario or different game users. However, the trajectory data formed under the control of the game robot is generally regular, such as moving back and forth at the same frequency in a specific area, or making actions at the same interval.

[0083] Therefore, in this embodiment, a behavioral intention representation that can characterize whether the virtual object moves according to the preset movement trajectory can be extracted from the trajectory data, and then it can be determined whether the client account corresponding to the virtual object has any violation. The violation specifically refers to the use of game robots to destroy the item trading rules within the game, which not only devalues ​​the player's virtual property, but also greatly increases the operational difficulty of the game company.

[0084] In a specific embodiment, constructing a first vector representation of the trajectory data based on the trajectory data of the virtual object in the virtual scene screen includes:

[0085] Converting the trajectory data into semantic sequence data, wherein the semantic sequence data is obtained by reducing the dimension of the trajectory data;

[0086] The semantic sequence data is input into a pre-trained first vector representation extraction model to perform a first vector representation extraction operation to obtain the first vector representation.

[0087] The first vector representation extraction model can be obtained by training based on models such as decision tree, gradient boosting tree, SVM, naive Bayes and logistic regression. Taking recursive neural network as an example, the specific training process includes:

[0088] Acquire a training sample of a first vector representation extraction model to be trained, wherein the training sample includes semantic sequence data provided with a label, wherein the label is used to indicate whether a virtual object corresponding to the semantic sequence data moves according to a preset movement trajectory;

[0089] Performing feature extraction on the semantic sequence data of the training sample by using the first vector representation extraction model to be trained to obtain a first vector representation corresponding to the semantic sequence data;

[0090] By using the first vector representation extraction model to be trained, judging based on the first vector representation whether the virtual object corresponding to the semantic sequence data in the training sample moves according to a preset movement trajectory, and obtaining a recognition result of the semantic sequence data;

[0091] Based on the recognition result and the label of the semantic sequence data, the parameters of the first vector representation extraction model to be trained are adjusted to obtain the pre-trained first vector representation extraction model.

[0092] According to the first vector representation and the degree information, and a preset graph neural network, a graph structure representation of the virtual object is obtained.

[0093] In order to improve the accuracy and rationality of detecting whether a game user has violated the rules, in this embodiment, in addition to extracting whether the virtual object moves according to the preset movement trajectory through trajectory data, it is also possible to extract whether the virtual object moves according to the preset behavior pattern through trajectory data. The preset behavior pattern refers to the action showing a certain regularity, such as releasing the same skill every 2 seconds, so as to determine whether it is a game robot.

[0094] Based on the above scheme, by inputting the graph structure representation and the intention representation into a preset fusion unit, the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation, so as to obtain the target trajectory information representation corresponding to the virtual object.

[0095] Specifically, the graph structure representation and the intent representation are input into a pre-trained feature fusion model for multiplication and then nonlinear activation to obtain a fusion coefficient, and a weighted fusion operation is performed on the graph structure representation and the intent representation according to the fusion coefficient to obtain a target trajectory information representation.

[0096] It should be noted that the feature fusion model is a model improved based on the recursive neural network. The difference from the traditional recursive neural network is that the feature fusion model only includes an input gate, an update gate and an output gate. The input gate is used to input the graph structure representation and the intent representation; the update gate is used to convert the graph structure representation and the intent representation into the target trajectory information representation; the output gate is used to output the target trajectory information representation.

[0097] Compared with traditional recursive neural networks, the feature fusion model does not have a reset gate. The function of the reset gate is to filter out some useless information when processing time series data. However, in the fusion scenario, there is no need to filter the previous information. Therefore, the reset gate can be removed to help the model converge faster and improve the accuracy of game violation detection.

[0098] Step 104: obtaining a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0099] In this embodiment, after obtaining the target trajectory information representation, the target trajectory information representation is sent to the linear layer, and then the probability of judging whether the client account corresponding to the virtual object has violated the rules is determined through the softmax (classifier) ​​output, that is, whether it is a game robot.

[0100] The target trajectory information representation is processed by a normalized exponential function to obtain a probability that the virtual object is a robot; and a judgment result of whether the virtual object is a robot is obtained according to the probability.

[0101] All the above optional technical solutions can be arbitrarily combined to form optional embodiments of the present application, which will not be described one by one here.

[0102] In specific implementation, the present application is not limited by the execution order of the various steps described. If no conflict occurs, some steps can be performed in other orders or simultaneously.

[0103] As can be seen from the above, the method for detecting a robot in a game provided by an embodiment of the present application provides a graphical user interface through the screen of a terminal device, and the content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object. The method includes: obtaining trajectory data of the virtual object moving in the game scene; obtaining corresponding semantic sequence data and topological structure data based on the trajectory data; obtaining a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological structure data, and a preset neural network; and obtaining a judgment result of whether the virtual object is a robot based on the target trajectory information representation.

[0104] like Figure 2 As shown, first, the trajectory graph data (i.e., trajectory data) is converted into data of a topological graph structure and data representing the player's intention (i.e., semantic sequence data), and the vector representation of the data node is obtained through the Node Embedding layer respectively. At the same time, in order to better utilize the degree information on the graph structure data, we encode the degree information to obtain the degree encoding, and then add it to the first vector representation obtained by Node Embedding, and then propagate the information through the Graph Attention network (i.e., graph attention neural network), and finally obtain the graph representation information (i.e., graph structure representation) of the game player. Similarly, we send the second vector representation obtained by Node Embedding into the GRU neural network to obtain the representation of the user's behavioral intention (i.e., graph structure representation). After obtaining the representation of the user's behavioral intention and the graph information representation, in order to reasonably utilize these two types of information, the graph structure representation and the intention representation are input into the preset fusion unit to obtain the target trajectory information representation corresponding to the virtual object. The embodiment of the present application extracts semantic sequence data and topological graph structure data that can effectively reflect the characteristics of the virtual object from the trajectory data of the virtual object, and fuses the two to obtain the target trajectory information representation of comprehensive information of different dimensions, and judges whether the client account corresponding to the virtual object has any violations based on the target trajectory information representation.

[0105] The embodiment of the present application also provides a device for detecting a robot in a game, and the device for detecting a robot in a game can be integrated in a terminal device, which can be a smart phone, a tablet computer, or other device.

[0106] See also Figure 3 , Figure 3 A schematic diagram of the structure of a device for detecting a robot in a game provided in an embodiment of the present application. The device 30 for detecting a robot in a game may include:

[0107] A trajectory data acquisition module 31 is used to acquire trajectory data of the virtual object moving in the game scene;

[0108] An intermediate data acquisition module 32, used to acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data;

[0109] A target data acquisition module 33, used to obtain a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological graph structure data, and a preset neural network;

[0110] The judgment result acquisition module 34 is used to obtain a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0111] In some embodiments, the intermediate data acquisition module 32 is used to divide the scene image into a plurality of cells; determine the cell into which any trajectory point falls according to the trajectory data, and obtain semantic sequence data according to the cell into which the trajectory point falls.

[0112] In some embodiments, the target data acquisition module 33 is used to obtain a first vector representation of a data node in the topological graph structure data and the degree information of the topological graph structure data; obtain a graph structure representation of the virtual object based on the first vector representation and the degree information, and a preset graph neural network; obtain a second vector representation of the node data in the semantic sequence data; obtain a behavioral intention representation of the virtual object based on the second vector representation and a preset neural network; input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object.

[0113] In some embodiments, the target data acquisition module 33 is used to input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit, and the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation to obtain the target trajectory information representation corresponding to the virtual object.

[0114] In some embodiments, the trajectory data is data obtained after filtering initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

[0115] In some embodiments, the judgment result acquisition module 34 is used to process the target trajectory information representation through a normalized exponential function to obtain the probability that the virtual object is a robot; and obtain the judgment result of whether the virtual object is a robot according to the probability.

[0116] In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined to be implemented as the same entity or several entities.

[0117] As can be seen from the above, the device 30 for detecting robots in the game provided in the embodiment of the present application obtains the trajectory data of the virtual object moving in the game scene through the trajectory data acquisition module 31; the intermediate data acquisition module 32 obtains the corresponding semantic sequence data and topological structure data according to the trajectory data; the target data acquisition module 33 obtains the target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological structure data, as well as a preset neural network; the judgment result acquisition module 34 obtains the judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0118] See also Figure 4 , Figure 4 Another structural diagram of the device for detecting a robot in a game provided in an embodiment of the present application, the device 30 for detecting a robot in a game includes a memory 120, one or more processors 180, and one or more applications, wherein the one or more applications are stored in the memory 120 and configured to be executed by the processor 180; the processor 180 may include a trajectory data acquisition module 31, an intermediate data acquisition module 32, a target data acquisition module 33, and a judgment result acquisition module. For example, the structure and connection relationship of the above components may be as follows:

[0119] The memory 120 can be used to store applications and data. The applications stored in the memory 120 include executable codes. The applications can be composed of various functional modules. The processor 180 executes various functional applications and data processing by running the applications stored in the memory 120. In addition, the memory 120 may include a high-speed random access memory and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 120 may also include a memory controller to provide the processor 180 with access to the memory 120.

[0120] The processor 180 is the control center of the device, and uses various interfaces and lines to connect various parts of the entire terminal, and executes various functions of the device and processes data by running or executing applications stored in the memory 120, and calling data stored in the memory 120, so as to monitor the device as a whole. Optionally, the processor 180 may include one or more processing cores; preferably, the processor 180 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs.

[0121] Specifically in this embodiment, the processor 180 will load the executable code corresponding to the process of one or more applications into the memory 120 according to the following instructions, and the processor 180 will run the application stored in the memory 120 to implement various functions:

[0122] A trajectory data acquisition module 31 is used to acquire trajectory data of the virtual object moving in the game scene;

[0123] An intermediate data acquisition module 32, used to acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data;

[0124] A target data acquisition module 33, used to obtain a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological graph structure data, and a preset neural network;

[0125] The judgment result acquisition module 34 is used to obtain a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0126] In some embodiments, the intermediate data acquisition module 32 is used to divide the scene image into a plurality of cells; determine the cell into which any trajectory point falls according to the trajectory data, and obtain semantic sequence data according to the cell into which the trajectory point falls.

[0127] In some embodiments, the target data acquisition module 33 is used to obtain a first vector representation of a data node in the topological graph structure data and the degree information of the topological graph structure data; obtain a graph structure representation of the virtual object based on the first vector representation and the degree information, and a preset graph neural network; obtain a second vector representation of the node data in the semantic sequence data; obtain a behavioral intention representation of the virtual object based on the second vector representation and a preset neural network; input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object.

[0128] In some embodiments, the target data acquisition module 33 is used to input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit, and the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation to obtain the target trajectory information representation corresponding to the virtual object.

[0129] In some embodiments, the trajectory data is data obtained after filtering initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

[0130] In some embodiments, the judgment result acquisition module 34 is used to process the target trajectory information representation through a normalized exponential function to obtain the probability that the virtual object is a robot; and obtain the judgment result of whether the virtual object is a robot according to the probability.

[0131] The embodiment of the present application also provides a terminal device, which can be a smart phone, a computer, a tablet computer, or other device.

[0132] See also Figure 5 , Figure 5 The schematic diagram of the structure of the terminal device provided in the embodiment of the present application is shown, and the terminal device can be used to implement the method for detecting robots in the game provided in the above embodiment. The terminal device 1200 can be a smart phone or a tablet computer.

[0133] like Figure 5 As shown, the terminal device 1200 may include an RF (Radio Frequency) circuit 110, a memory 120 including one or more (only one is shown in the figure) computer-readable storage media, an input unit 130, a display unit 140, a sensor 150, an audio circuit 160, a transmission module 170, a processor 180 including one or more (only one is shown in the figure) processing cores, and a power supply 190. Those skilled in the art will understand that Figure 5 The structure of the terminal device 1200 shown in the figure does not constitute a limitation on the terminal device 1200, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0134] The RF circuit 110 is used to receive and send electromagnetic waves, realize the mutual conversion between electromagnetic waves and electrical signals, and thus communicate with a communication network or other devices. The RF circuit 110 may include various existing circuit elements for performing these functions, such as an antenna, a radio frequency transceiver, a digital signal processor, an encryption / decryption chip, a user identity module (SIM) card, a memory, etc. The RF circuit 110 can communicate with various networks such as the Internet, an intranet, a wireless network, or communicate with other devices through a wireless network.

[0135] The memory 120 can be used to store software programs and modules, such as the program instructions / modules corresponding to the method for detecting the robot in the game in the above embodiment. The processor 180 executes various functional applications and data processing by running the software programs and modules stored in the memory 120. The vibration reminder mode can be automatically selected according to the current scene of the terminal device to detect the robot in the game, which can not only ensure that the scene such as the meeting is not disturbed, but also ensure that the user can feel the incoming call, thereby improving the intelligence of the terminal device. The memory 120 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 120 may further include a memory remotely arranged relative to the processor 180, and these remote memories may be connected to the terminal device 1200 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0136] The input unit 130 can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control. Specifically, the input unit 130 may include a touch-sensitive surface 131 and other input devices 132. The touch-sensitive surface 131, also known as a touch display screen or a touch pad, can collect user touch operations on or near it (such as operations performed by users using fingers, styluses, or any other suitable objects or accessories on or near the touch-sensitive surface 131), and drive corresponding connection devices according to a pre-set program. Optionally, the touch-sensitive surface 131 may include a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into touch point coordinates, and then sends it to the processor 180, and can receive and execute commands sent by the processor 180. In addition, the touch-sensitive surface 131 can be implemented in various types such as resistive, capacitive, infrared, and surface acoustic wave. In addition to the touch-sensitive surface 131, the input unit 130 may further include other input devices 132. Specifically, other input devices 132 may include but are not limited to one or more of a physical keyboard, function keys (such as volume control keys, switch keys, etc.), a trackball, a mouse, a joystick, etc.

[0137] The display unit 140 may be used to display information input by the user or information provided to the user and various graphical user interfaces of the terminal device 1200, which may be composed of graphics, text, icons, videos and any combination thereof. The display unit 140 may include a display panel 141. Optionally, the display panel 141 may be configured in the form of an LCD (Liquid Crystal Display), an OLED (Organic Light-Emitting Diode), or the like. Further, the touch-sensitive surface 131 may cover the display panel 141. When the touch-sensitive surface 131 detects a touch operation on or near it, it is transmitted to the processor 180 to determine the type of touch event. Subsequently, the processor 180 provides a corresponding visual output on the display panel 141 according to the type of touch event. Although in Figure 5 In the embodiment, the touch-sensitive surface 131 and the display panel 141 are implemented as two independent components to implement input and output functions. However, in some embodiments, the touch-sensitive surface 131 and the display panel 141 can be integrated to implement input and output functions.

[0138] The terminal device 1200 may also include at least one sensor 150, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor may include an ambient light sensor and a proximity sensor, wherein the ambient light sensor may adjust the brightness of the display panel 141 according to the brightness of the ambient light, and the proximity sensor may turn off the display panel 141 and / or the backlight when the terminal device 1200 is moved to the ear. As a type of motion sensor, the gravity acceleration sensor can detect the magnitude of acceleration in each direction (generally three axes), and can detect the magnitude and direction of gravity when stationary, which can be used for applications that identify the posture of the mobile phone (such as horizontal and vertical screen switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc.; as for other sensors such as gyroscopes, barometers, hygrometers, thermometers, infrared sensors, etc. that can also be configured in the terminal device 1200, they will not be repeated here.

[0139] The audio circuit 160, the speaker 161, and the microphone 162 can provide an audio interface between the user and the terminal device 1200. The audio circuit 160 can transmit the received audio data to the speaker 161 after converting the received audio data into an electrical signal, which is converted into a sound signal for output by the speaker 161; on the other hand, the microphone 162 converts the collected sound signal into an electrical signal, which is received by the audio circuit 160 and converted into audio data, and then the audio data is output to the processor 180 for processing, and then sent to another terminal through the RF circuit 110, or the audio data is output to the memory 120 for further processing. The audio circuit 160 may also include an earplug jack to provide communication between an external headset and the terminal device 1200.

[0140] The terminal device 1200 can help users send and receive emails, browse web pages, and access streaming media through the transmission module 170 (such as a Wi-Fi module), which provides users with wireless broadband Internet access. Figure 5 The transmission module 170 is shown, but it is understandable that it is not a necessary component of the terminal device 1200 and can be omitted as needed without changing the essence of the invention.

[0141] The processor 180 is the control center of the terminal device 1200. It uses various interfaces and lines to connect various parts of the entire mobile phone. By running or executing software programs and / or modules stored in the memory 120, and calling data stored in the memory 120, it executes various functions of the terminal device 1200 and processes data, thereby monitoring the mobile phone as a whole. Optionally, the processor 180 may include one or more processing cores; in some embodiments, the processor 180 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 180.

[0142] The terminal device 1200 also includes a power supply 190 for supplying power to various components. In some embodiments, the power supply can be logically connected to the processor 180 through a power management system, so that the power management system can realize functions such as managing discharge and power consumption. The power supply 190 can also include any components such as one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, and power status indicators.

[0143] Although not shown, the terminal device 1200 may also include a camera (such as a front camera, a rear camera), a Bluetooth module, etc., which will not be described in detail here. Specifically in this embodiment, the display unit 140 of the terminal device 1200 is a touch screen display, and the terminal device 1200 also includes a memory 120, and one or more programs, wherein one or more programs are stored in the memory 120, and are configured to be executed by one or more processors 180. One or more programs include instructions for performing the following operations:

[0144] A trajectory data acquisition instruction, used to acquire trajectory data of the virtual object moving in the game scene;

[0145] Intermediate data acquisition instructions, used to acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data;

[0146] A target data acquisition instruction, used to obtain a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological graph structure data, and a preset neural network;

[0147] The judgment result acquisition instruction is used to obtain a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

[0148] In some embodiments, the intermediate data acquisition instruction is used to divide the scene image into a plurality of cells; determine the cell into which any trajectory point falls according to the trajectory data, and obtain semantic sequence data according to the cell into which the trajectory point falls.

[0149] In some embodiments, the target data acquisition instruction is used to obtain a first vector representation of a data node in the topological graph structure data and the degree information of the topological graph structure data; obtain a graph structure representation of the virtual object based on the first vector representation and the degree information, and a preset graph neural network; obtain a second vector representation of the node data in the semantic sequence data; obtain a behavioral intention representation of the virtual object based on the second vector representation and a preset neural network; input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object.

[0150] In some embodiments, the target data acquisition instruction is used to input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit, and the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation to obtain the target trajectory information representation corresponding to the virtual object.

[0151] In some embodiments, the trajectory data is data obtained after filtering initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

[0152] In some embodiments, the judgment result acquisition instruction is used to process the target trajectory information representation through a normalized exponential function to obtain the probability that the virtual object is a robot; and obtain a judgment result of whether the virtual object is a robot based on the probability.

[0153] The embodiment of the present application also provides a terminal device, which may be a smart phone, a tablet computer, or other device.

[0154] As can be seen from the above, an embodiment of the present application provides a terminal device 1200, and the terminal device 1200 performs the following steps: providing a graphical user interface through the screen of the terminal device, the content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object, and the method includes: obtaining trajectory data of the virtual object moving in the game scene; obtaining corresponding semantic sequence data and topological structure data according to the trajectory data; obtaining a target trajectory information representation corresponding to the virtual object based on the semantic sequence data and the topological structure data, and a preset neural network; obtaining a judgment result of whether the virtual object is a robot according to the target trajectory information representation. The embodiment of the present application extracts semantic sequence data and topological structure data that can effectively reflect the characteristics of the virtual object from the trajectory data of the virtual object, and fuses the two to obtain a target trajectory information representation that integrates information of different dimensions, and judges whether the client account corresponding to the virtual object has any violation based on the target trajectory information representation.

[0155] An embodiment of the present application further provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium. When the computer program runs on a computer, the computer executes the method for detecting a robot in a game as described in any of the above embodiments.

[0156] It should be noted that, for the method of detecting robots in the game described in the present application, ordinary testers in the field can understand that all or part of the process of implementing the method of detecting robots in the game described in the embodiment of the present application can be completed by controlling the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, such as stored in the memory of a terminal device, and executed by at least one processor in the terminal device, and the execution process may include the process of the embodiment of the method of detecting robots in the game. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), etc.

[0157] For the device for detecting robots in the game of the embodiment of the present application, each functional module can be integrated into a processing chip, or each module can exist physically separately, or two or more modules can be integrated into one module. The above-mentioned integrated module can be implemented in the form of hardware or in the form of software functional modules. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium, such as a read-only memory, a disk or an optical disk.

[0158] The above is a detailed introduction to the method, device, computer-readable storage medium and terminal device for detecting robots in games provided by the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for those skilled in the art, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for detecting robots in a game, characterized in that: A graphical user interface is provided through a screen of a terminal device, wherein the content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object, and the method includes: Acquire trajectory data of the virtual object moving in the game scene; Acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data; Obtaining a first vector representation of a data node in the topological graph structure data and degree information of the topological graph structure data; Obtaining a graph structure representation of the virtual object according to the first vector representation and the degree information, and a preset graph neural network; Obtaining a second vector representation of the node data in the semantic sequence data; Obtaining a behavior intention representation of the virtual object according to the second vector representation and a preset neural network; Inputting the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object; A judgment result of whether the virtual object is a robot is obtained according to the target trajectory information representation.

2. The method according to claim 1, characterized in that The acquiring corresponding semantic sequence data according to the trajectory data includes: Dividing the scene image into a plurality of cells; The cell into which any trajectory point falls is determined according to the trajectory data, and the semantic sequence data is obtained according to the cell into which the trajectory point falls.

3. The method according to claim 2, characterized in that The method further comprises: Assigning corresponding numbers to the multiple cells respectively; The obtaining of semantic sequence data according to the cell into which the trajectory point falls includes: The semantic sequence data is obtained according to the number of the cell into which the trajectory point falls.

4. The method according to claim 1, characterized in that The method further comprises: Obtaining degree information of the topological graph structure data; The obtaining, according to the first vector representation and a preset graph neural network, a graph structure representation of the virtual object includes: According to the first vector representation and the degree information, and a preset graph neural network, a graph structure representation of the virtual object is obtained.

5. The method according to claim 1, characterized in that The step of obtaining the behavior intention representation of the virtual object according to the second vector representation and a preset neural network includes: According to the second vector representation and the preset GRU neural network, the behavior intention representation of the virtual object is obtained.

6. The method according to claim 1, characterized in that The step of inputting the semantic sequence data, the graph structure representation, and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object includes: The semantic sequence data, the graph structure representation and the intention representation are input into a preset fusion unit, and the fusion unit performs nonlinear transformation and addition operations on the semantic sequence data, the graph structure representation and the intention representation to obtain the target trajectory information representation corresponding to the virtual object.

7. The method according to claim 1, characterized in that The trajectory data is data obtained after filtering the initial trajectory data of the virtual object moving in the game scene according to the resolution of the screen.

8. The method according to claim 1, characterized in that The determining result of whether the virtual object is a robot according to the target trajectory information representation includes: Processing the target trajectory information representation by a normalized exponential function to obtain a probability that the virtual object is a robot; A judgment result of whether the virtual object is a robot is obtained according to the probability.

9. A device for detecting robots in a game, characterized in that: A graphical user interface is provided through a screen of a terminal device, wherein the content displayed by the graphical user interface includes a scene image of a game scene, and the game scene includes a virtual object. The device includes: A trajectory data acquisition module, used to acquire trajectory data of the virtual object moving in the game scene; An intermediate data acquisition module, used to acquire corresponding semantic sequence data and topological graph structure data according to the trajectory data; A target data acquisition module is used to obtain a first vector representation of a data node in the topological graph structure data and the degree information of the topological graph structure data; obtain a graph structure representation of the virtual object based on the first vector representation and the degree information, and a preset graph neural network; obtain a second vector representation of the node data in the semantic sequence data; obtain a behavioral intention representation of the virtual object based on the second vector representation and a preset neural network; input the semantic sequence data, the graph structure representation and the intention representation into a preset fusion unit to obtain a target trajectory information representation corresponding to the virtual object; The judgment result acquisition module is used to obtain a judgment result of whether the virtual object is a robot according to the target trajectory information representation.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the method for detecting a robot in a game according to any one of claims 1-8.

11. A terminal device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a plurality of instructions, and the processor loads the instructions to execute the method for detecting a robot in a game as described in any one of claims 1 to 8.

Images