Method and device for adjusting character appearance in game and electronic equipment

Abstract

The invention discloses a role appearance adjusting method and device, electronic equipment and a computer readable storage medium, the method and device are applied to a game scene, the game scene at least comprises a virtual role controlled through a game account, and the method comprises the steps that game experience data corresponding to the virtual role are collected, the game experience data comprises at least one of task experience data, environment experience data, combat experience data, social experience data and achievement experience data; determining an appearance adjustment parameter corresponding to the virtual character according to the game experience data; and according to the appearance adjustment parameter, adjusting the appearance asset of the virtual character, so that the appearance expression effect of the virtual character is matched with the game experience of the virtual character. According to the scheme provided by the invention, mapping from the game experience to the role appearance can be realized, so that the role appearance can be dynamically adjusted along with the game experience of the role, and the specificity of the role appearance is greatly improved.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method, device, electronic device and computer-readable storage medium for adjusting the appearance of a character in a game. Background Art

[0002] In virtual games, the appearance of characters is not only the visual embodiment of players, but also an important element to enhance the immersion and sense of involvement in the game. At present, the character appearance design in most games adopts a relatively static method, such as adjusting the character appearance through preset appearance options or players manually changing equipment. The current method cannot truly reflect the growth of the character, and it is difficult to establish a sense of involvement in the character's growth, thereby reducing the player's gaming experience. Summary of the invention

[0003] The present application provides a method, device, electronic device and computer-readable storage medium for adjusting the appearance of a character, which can realize the mapping of game experience to character appearance, so that the character appearance can be dynamically adjusted according to the character's game experience, greatly improving the specificity of the character appearance. The specific scheme is as follows:

[0004] In a first aspect, an embodiment of the present application provides a method for adjusting the appearance of a character, which is applied to a game scene, wherein the game scene includes at least a virtual character controlled by a game account, and the method includes:

[0005] Collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data;

[0006] Determining appearance adjustment parameters corresponding to the virtual character according to the game experience data;

[0007] The appearance assets of the virtual character are adjusted according to the appearance adjustment parameters so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0008] In a second aspect, an embodiment of the present application provides a device for adjusting the appearance of a character, which is applied to a game scene, wherein the game scene includes at least a virtual character controlled by a game account, and the device includes:

[0009] A collection unit, used for collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data;

[0010] A determination unit, configured to determine appearance adjustment parameters corresponding to the virtual character according to the game experience data;

[0011] The adjustment unit is used to adjust the appearance assets of the virtual character according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0012] In a third aspect, the present application further provides an electronic device, including:

[0013] Processor; and

[0014] The memory is used to store a data processing program. After the electronic device is powered on and the program is run by the processor, the method described in the first aspect is executed.

[0015] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium storing a data processing program, which is executed by a processor to perform the method described in the first aspect.

[0016] Compared with the prior art, this application has the following advantages:

[0017] The method for adjusting the appearance of a character provided in an embodiment of the present application is applied to a game scene, and the game scene includes at least a virtual character controlled by a game account, including: collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data; then, according to the game experience data, determining the appearance adjustment parameters corresponding to the virtual character, since the game experience data is data corresponding to the experience of the virtual character in the game, so that the appearance adjustment parameters determined based on the game experience data are appearance adjustment parameters that match the game experience of the virtual character; finally, according to the appearance adjustment parameters, adjusting the appearance assets of the virtual character, since the appearance adjustment parameters match the game experience of the virtual character, the appearance performance effect of the adjusted virtual character matches the game experience of the virtual character. It can be seen that the method for adjusting the appearance of a character provided in an embodiment of the present application can realize the mapping of game experience to character appearance, so that the character appearance can be dynamically adjusted following the game experience of the character, which greatly improves the specificity of the character appearance. In addition, from the player's perspective, based on the appearance of each virtual character in the game, you can intuitively feel the experience and growth of each virtual character in the game, thereby creating a more personalized and immersive gaming experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a system diagram for adjusting the appearance of a character for implementing a method for adjusting the appearance of a character provided in an embodiment of the present application;

[0019] Figure 2is a schematic diagram of the structure of a terminal in a system for adjusting the appearance of a character provided in an embodiment of the present application;

[0020] Figure 3 is a flow chart of a method for adjusting the appearance of a character provided in an embodiment of the present application;

[0021] Figure 4 This is an interface schematic diagram of an example of appearance change of a virtual character doing a task in a flame area in the method for adjusting the appearance of a character provided in an embodiment of the present application;

[0022] Figure 5 This is an interface schematic diagram of an example of the appearance change of a virtual character that has experienced a battle in the method for adjusting the appearance of a character provided in an embodiment of the present application;

[0023] Figure 6 is a structural block diagram of an example of a device for adjusting the appearance of a character provided in an embodiment of the present application;

[0024] Figure 7 It is a structural block diagram of an example of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0026] It should be noted that the terms "first", "second", "third", etc. in the claims, description and drawings of the present application are used to distinguish similar objects and are not used to describe a specific order or sequence. The data used in this way are interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including", "having" and their variants are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0027] It should be understood that in the embodiments of the present application, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" is merely a way to describe the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and subsequent associated objects are in an "or" relationship. "Including A, B and / or C" means including any one, any two, or any three of A, B, and C.

[0028] It should be understood that in the embodiments of the present application, "B corresponding to A", "B corresponding to A", "A corresponds to B", or "B corresponds to A" means that B is associated with A, and B can be determined according to A. Determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0029] Before describing the implementation methods of the present application in detail, the prior art is first further described.

[0030] With the continuous development of computer technology, virtual games have received widespread attention. In virtual games, players usually control virtual characters to explore, fight and develop in the virtual world. During the game, the change of character appearance has become an extremely critical link. The design of the character's appearance should not only be beautiful, but also reflect the character's growth process and personality characteristics, so as to strengthen the player's sense of identity with the character and enhance the game immersion.

[0031] In the related art, in order to achieve the change of the appearance of the character in the game, preset character appearance options are usually provided for players to choose, and players are allowed to perform a certain degree of personalization by changing equipment, customizing the character appearance, etc. In addition, some games will unlock new appearance options based on the tasks completed or the achievements achieved by the players. However, the above method has at least the following limitations:

[0032] First, the character appearance provided by most games is static, that is, the appearance will not automatically change according to the player's experience accumulated in the game. Players need to manually change equipment or adjust the character's appearance to change the character's appearance. This static appearance design makes it difficult to reflect the character's growth process and reduces the player's sense of immersion;

[0033] Second, because the appearance options provided by the game are relatively limited, different player characters are prone to similar appearances. Even if there are differences, these differences are mostly due to the subjective choices of the players, rather than passive changes during the game. This leads to a lack of uniqueness and recognition in the appearance of the characters, which cannot fully meet the players' pursuit of personalization.

[0034] Third, the design of the character's appearance is not closely connected with the environment, tasks and other factors in the game world, and lacks the necessary interactivity. This results in the character's appearance not being affected by the environment or tasks it completes, reducing the player's sense of interaction and immersion with the game world.

[0035] In summary, existing role-playing games have serious homogeneity in character appearance and lack of interactivity with the game world, which limits the role of appearance as a visual avatar and cannot effectively enhance the player's sense of presence and subjectivity. Therefore, there is an urgent need for a method that can dynamically change the character's appearance as the player's game experience, so as to create a more personalized and immersive gaming experience.

[0036] Based on the above reasons, in order to realize the mapping of game experience to character appearance, the character appearance can be dynamically adjusted according to the character's game experience, which greatly enhances the specificity of the character's appearance and creates a more personalized and immersive gaming experience. The first embodiment of the present application provides a method for adjusting the appearance of a character. The method is applied to an electronic device, which can be a desktop computer, a laptop computer, a mobile phone, a tablet computer, an electronic watch, etc., or other electronic devices that can adjust the appearance of the character, which is not specifically limited in the embodiments of the present application.

[0037] In an optional embodiment, when the method for adjusting the appearance of a character is run on a terminal device, the terminal device may include a display screen and a processor, the display screen being used to present a game screen and receive instructions generated by a player acting on the game screen. The game screen may include a portion of a virtual game scene, which is a virtual world in which a virtual character performs activities. The processor is used to store game applications, run the game, generate game screens, respond to instructions, and control the display of the game screen on the display screen. When the player operates the game screen through the display screen, the game screen can control the local content of the terminal device by responding to the received operation instructions. The terminal device may provide a graphical user interface to the player in a variety of ways, for example, it may be rendered and displayed on the display screen of the terminal device, or the graphical user interface may be presented through holographic projection.

[0038] In an optional embodiment, when the method for adjusting the appearance of the character is run on the server, the method can be implemented and executed based on the cloud gaming system. The cloud gaming system refers to a gaming method based on cloud computing. The cloud gaming system includes a server and a client device. The operating body of the game application and the game screen presentation body are separated, and the storage and operation of the method for adjusting the appearance of the character are completed on the server. The game screen presentation is completed on the client, and the client is mainly used for receiving and sending game data and presenting the game screen. For example, the client can be a display device with a data transmission function close to the player side, such as a mobile terminal, a TV, a computer, a handheld computer, a personal digital assistant, a head-mounted display device, etc., but the terminal device for processing game data is a server in the cloud. When playing the game, the player instructs the client to send instructions to the server, and the server controls the running of the game according to the instructions, encodes and compresses the game screen and other data, and returns it to the client through the network. Finally, the client decodes and outputs the game screen.

[0039] It should be noted that in the embodiment of the present application, the execution subject of the method for adjusting the appearance of the character can be a terminal device or a server, wherein the terminal device can be a local terminal device or a client device in the aforementioned cloud game. The embodiment of the present application does not limit the type of the execution subject.

[0040] For example, in combination with the above introduction, Figure 1 A system 100 for adjusting character appearance for implementing a method for adjusting character appearance provided in an embodiment of the present application is shown, including at least one terminal 101 , a server 102 and a network 103 .

[0041] In the above-mentioned system 100 for adjusting the appearance of a character, the terminal 101 is used to install and run a game application. In some cases, the game application may not be installed in advance in the terminal 101, and the player may directly access the game through a client such as a browser. The player logs in to the game application using a registered game account, and can manipulate the virtual character corresponding to the game account to participate in the game. When the player logs in to the game application, the terminal 101 sends a login request to the server 102, and the server 102 verifies the game account used by the player. At the same time, the server 102 determines the game mechanism corresponding to the game account according to the login request, and returns a login success notification to the terminal 101 after the verification is passed. In the process of the player participating in the game through the game application, the terminal 101 and the server 102 interact with each other, and the terminal 101 sends various information to the server 102. The server 102 determines the display data of the terminal 101 according to the stored game mechanism and the received information, and sends the display data to the terminal 101, so that the display data sent by the server 102 is displayed to the player through the terminal 101.

[0042] In a possible application scenario, different terminals 101 may be served by different servers 102 , and the servers 102 corresponding to different terminals 101 may be the same server.

[0043] In addition, when the game system 100 includes multiple terminals, multiple servers, and multiple networks, different terminals can be connected to each other through different networks and different servers.

[0044] Among them, the terminal 101 can have one or more multi-touch screens for sensing and obtaining the input of the user through touch or sliding operations performed at multiple points on one or more touch display screens. The terminal 101 can also be connected to a keyboard and / or a mouse and / or a game controller, so that the user can perform interface operations through the keyboard and / or mouse and / or a game controller.

[0045] The network 103 may be a wireless network or a wired network, such as a wireless network such as a wireless local area network (WLAN), a local area network (LAN), a cellular network, a 2G network, a 3G network, a 4G network, a 5G network, etc. In addition, different terminals may also use their own Bluetooth network or hotspot network to connect to other terminals or to a server, etc. In addition, the system 100 may include multiple databases, which are coupled to different servers, and game-related information may be continuously stored in the database when different players play a multi-player game online.

[0046] It should be noted that Figure 1 The game system diagram shown is only an example. The game system 100 described in the embodiment of the present application is to more clearly illustrate the technical solution of the embodiment of the present application, and does not constitute a limitation on the technical solution provided in the embodiment of the present application. A person of ordinary skill in the art can know that with the evolution of the game system and the emergence of new business scenarios, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.

[0047] like Figure 2As shown, it is a schematic diagram of the structure of a terminal in a system for adjusting the appearance of a character provided in an embodiment of the present application. The terminal 101 may include an experience collection plug-in 1011, a pre-trained language model 1012, and an appearance generation engine 1013. Among them, the experience collection plug-in 1011 is used to collect and label the experience accumulated by the virtual character in the game world, obtain labeled game experience data, and store the labeled game experience data in the experience database of the game server. The pre-trained language model 1012 is used to pull the labeled game experience data from the experience database, and generate corresponding appearance adjustment parameters according to the labeled game experience data, and send the appearance adjustment parameters to the appearance generation engine 1013. The appearance generation engine 1013 is used to dynamically adjust the appearance of the virtual character according to the appearance adjustment parameters.

[0048] The technical solution of the present application is described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0049] The following, combined Figure 3 to Figure 5 The present invention introduces a method for adjusting the appearance of a character provided in an embodiment of the present application.

[0050] like Figure 3 As shown, it is a flow chart of the method for adjusting the appearance of a character provided by the present application. The method is applied in a game scene, and the game scene at least includes a virtual character controlled by a game account, including the following steps S101 to S103.

[0051] Step S101: collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data.

[0052] It can be understood that virtual games refer to applications developed according to the needs of game applications, and the types of games may include but are not limited to at least one of the following: two-dimensional (Two Dimension) game applications, three-dimensional (Three Dimension) game applications, virtual reality (Virtual Reality, VR) game applications, augmented reality (Augmented Reality, AR) game applications, and mixed reality (Mixed Reality, MR) game applications.

[0053] In the embodiment of the present application, the virtual game may be a role-playing game, a sandbox game, an action-adventure game, a shooting game, etc. In the role-playing game, the player plays a specific role and has a rich gaming experience in the virtual world; the sandbox game allows the player to freely explore and create in an open world; the action-adventure game may generally include game elements such as combat, puzzle solving, and exploration; and the shooting game may allow the player to defeat the enemy virtual character by shooting.

[0054] The above game scenes may include one or more virtual characters controlled by a game account, which refers to a personal account registered and used by a player on a game platform. The game account can save the game progress, ensure data security, and allow players to continue playing the game on different devices with the current game progress corresponding to the game account. In addition, the account also supports social interaction in multiplayer games, such as playing games with friends.

[0055] In the embodiment of the present application, the game experience data corresponding to the virtual character can be collected in real time, or the game experience data corresponding to the virtual character can be collected according to a preset period. The game experience data refers to the data generated by various game activities and game events participated in by the virtual character in the virtual game, which can be used to influence the appearance of the character. The game experience data may include at least one of the following: task experience data, environmental experience data, combat experience data, social experience data, and achievement experience data.

[0056] Among them, the task experience data refers to the data corresponding to the tasks triggered, performed and completed by the virtual character in the task system. It should be noted that the task experience can affect the overall image style of the virtual character. For example, virtual character A has completed more exploration tasks, and its clothing style tends to be light and practical. Virtual character B has completed more social tasks, and its clothing style tends to be gorgeous and fashionable.

[0057] Environmental experience data refers to the data corresponding to the exploration trajectory and residence time of the virtual character in each game environment. Each game environment may include but is not limited to a virtual forest, a virtual ocean, a virtual desert, a virtual swamp, and a virtual snow mountain. It should be noted that environmental experience can affect the appearance of the virtual character. For example, when virtual character A explores in a swamp, his clothes and skin will be stained with mud, and when virtual character B moves on a snow mountain, frost will condense on his equipment.

[0058] The combat experience data refers to the data corresponding to the combats that the virtual character participated in. Each game environment may include but is not limited to a virtual forest, a virtual ocean, a virtual desert, a virtual swamp, and a virtual snow mountain. It should be noted that combat experience can reflect the shaping of the appearance of the virtual character by the combat style and intensity. For example, if virtual character A frequently participates in high-intensity combats, his armor will be damaged and his skin will have scars. Virtual character B rarely participates in combat and his clothes are relatively neat.

[0059] Social experience data refers to the interaction data between the virtual character and other player characters or NPCs (non-player characters). It should be noted that social experience data can be used to determine appearance accessory identification related to social interaction. For example, if virtual character A has good interaction with NPCs, it can wear designated accessories related to the NPC (such as accessories given by the NPC). If virtual character B often fights with other virtual characters, its clothing can have exclusive combat style identification.

[0060] Achievement experience data refers to the data corresponding to the achievements achieved by the virtual character. It should be noted that achievement experience can highlight the growth and milestones of the character, and achievement experience data can be used to determine the appearance reward logo related to the achievement, and the appearance reward logo highlights the achievement of the virtual character through visual presentation. For example, if virtual character A achieves achievement 1, it can unlock the specified tattoo reward, and virtual character B achieves achievement 2, it can unlock the specified special effect reward.

[0061] In the embodiment of the present application, the collected game experience data can be uploaded and stored in the experience database of the game server according to the time series. In this way, the security and availability of the data can be guaranteed, so that the data can be analyzed and processed conveniently, and the storage of the data according to the time series is helpful to analyze the changing trend of the game behavior of the virtual character.

[0062] In addition, the collected game experience data can be associated with the virtual character, thereby forming an "experience portrait" of each virtual character in the game world. In this way, each piece of game experience data can clearly point to a specific virtual character, thereby forming a personalized experience feature.

[0063] Step S102: determining appearance adjustment parameters corresponding to the virtual character according to the game experience data.

[0064] Since the game experience data collected in step S101 refers to data generated by various game activities and game events in which the virtual character participates in the virtual game, and the game experience data can be used to influence the appearance of the character, thus, the changes in the appearance of the virtual character after participating in various game activities and game events can be determined based on the game experience data.

[0065] In a specific implementation, the appearance adjustment parameters corresponding to the virtual character can be determined according to the game experience data. The appearance adjustment parameters are used to characterize the adjustment method of the appearance assets of the virtual character.

[0066] For example, the appearance adjustment parameter corresponding to the virtual character a is that the body becomes thinner to 80% of the previous size, then the adjustment method of the appearance assets of the virtual character a can be characterized as scaling the character model to 80% of the original appearance; the appearance feature parameter corresponding to the virtual character b is that the clothes are gorgeous red clothes, then the adjustment method of the appearance assets of the virtual character b can be characterized as removing the original clothing assets and wearing gorgeous red clothing assets.

[0067] In an optional specific implementation, the above step S102 can be implemented through a pre-trained language model (such as Chatgpt), that is, the above step S102 can include: inputting the game experience data into the pre-trained language model, so that the pre-trained language model outputs corresponding appearance adjustment parameters.

[0068] In a specific implementation, end-to-end output can be performed through a pre-trained language model, thereby directly mapping the game experience data to the corresponding appearance adjustment parameters. End-to-end output refers to the entire process of a machine learning model or system, in which the input data is processed to directly produce the final output result without human intervention or additional intermediate steps. In the field of natural language processing, an end-to-end approach usually means directly generating the final output text from the original input text without the need for feature extraction, intermediate conversion or other complex preprocessing steps in advance.

[0069] Step S103: adjusting the appearance assets of the virtual character according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0070] After the appearance adjustment parameters corresponding to the virtual character are determined, the appearance assets of the virtual character can be adjusted accordingly according to the adjustment method of the appearance assets represented by the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the corresponding game experience.

[0071] The method for adjusting the appearance of a character provided in an embodiment of the present application is applied to a game scene, and the game scene includes at least a virtual character controlled by a game account, including: collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data; then, according to the game experience data, determining the appearance adjustment parameters corresponding to the virtual character, since the game experience data is data corresponding to the experience of the virtual character in the game, so that the appearance adjustment parameters determined based on the game experience data are appearance adjustment parameters that match the game experience of the virtual character; finally, according to the appearance adjustment parameters, adjusting the assets of the virtual character, since the appearance adjustment parameters match the game experience of the virtual character, the appearance performance effect of the adjusted virtual character matches the game experience of the virtual character. It can be seen that the method for adjusting the appearance of a character provided in an embodiment of the present application can realize the mapping of game experience to character appearance, so that the character appearance can be dynamically adjusted following the game experience of the character, which greatly improves the specificity of the character appearance. In addition, from the player's perspective, based on the appearance of each virtual character in the game, you can intuitively feel the experience and growth of each virtual character in the game, thereby creating a more personalized and immersive gaming experience.

[0072] In an optional embodiment, the pre-trained language model can be obtained by training the following steps S20 and S21:

[0073] Step S20: Acquire first sample data, where the first sample data includes sample game experience data and sample appearance adjustment data corresponding to the sample game experience data;

[0074] Step S21: training the language model to be trained according to the first sample data until a preset training end condition is reached to obtain a pre-trained language model.

[0075] In an embodiment of the present application, the language model to be trained can be trained using the first sample data to obtain a pre-trained language model that can output appearance adjustment parameters based on the game experience data.

[0076] Optionally, the sample game experience data in the first sample data is obtained from a game log, that is, the sample game experience data may be historical game experience data of a real player, and the sample appearance adjustment data in the first sample data may be sample appearance adjustment parameters constructed based on and matching the sample game experience data.

[0077] After the first sample data is obtained, the model parameters of the language model to be trained can be optimized through the first sample data, so as to train the language model to be trained to have the ability to map the game experience data into appearance adjustment parameters.

[0078] The preset training end conditions may include but are not limited to: the accuracy of the predicted appearance adjustment data output by the language model to be trained reaches a threshold, the number of training times reaches a preset number of times, and the loss value of the loss function corresponding to the language model to be trained is within a preset loss value.

[0079] In an optional specific implementation, the above step S21 can be implemented by the following steps:

[0080] Inputting the first sample data into the language model to be trained, so that the language model to be trained outputs predicted appearance adjustment data for the sample game experience data;

[0081] The model parameters of the language model to be trained are adjusted so that the difference between the predicted appearance adjustment data and the sample appearance adjustment data is within a first preset range, thereby obtaining a pre-trained language model.

[0082] It should be noted that the training goal of the above training process is to make the predicted appearance adjustment data output by the language model to be trained as close as possible to the sample appearance adjustment data, that is, to minimize the difference between the predicted appearance adjustment data and the sample appearance adjustment data.

[0083] In an embodiment of the present application, the first sample data may be input into the language model to be trained, so that the language model to be trained may output corresponding predicted appearance adjustment data for the sample game experience data in the input first sample data.

[0084] Afterwards, a difference between the predicted appearance adjustment data and the corresponding sample appearance adjustment data may be determined, and a loss value of a loss function corresponding to the language model to be trained may be determined based on the difference.

[0085] Finally, the model parameters of the language model to be trained can be adjusted by an adjustment strategy that makes the loss value less than a preset loss value, thereby obtaining a pre-trained language model. It should be noted that the above “so that the difference between the predicted appearance adjustment data and the sample appearance adjustment data is within a first preset range” can be understood as an adjustment strategy that makes the loss value less than a preset loss value.

[0086] In another optional specific implementation, the above step S21 may include the following steps S201 to S205:

[0087] Step S201: Obtain the first sample game corpus, where the first sample game corpus includes at least one of gameplay corpus, game mechanism corpus, game strategy corpus, and game analysis corpus;

[0088] Step S202: Perform word masking and sentence order shuffling on the first sample game corpus to obtain the processed first sample game corpus;

[0089] Step S203: Input the processed first sample game corpus into the language model to be trained, so that the language model to be trained outputs predicted masked words for the processed first sample game corpus and predicted sentence orders for the processed first sample game corpus;

[0090] Step S204: Adjust the model parameters of the language model to be trained so that the difference between the predicted masked words and the true masked words is within a second preset range, and the difference between the predicted sentence order and the true sentence order is within a third preset range, to obtain the trained first model;

[0091] Step S205: Train the first model according to the first sample data until a preset training end condition is reached to obtain a pre-trained language model.

[0092] In the embodiments of this application, a self-supervised learning method can be used to pre-train a language model based on the first sample game corpus to obtain a pre-trained language model, so that the pre-trained language model has the ability to master the vocabulary, grammar, and semantics of the game corpus and the ability to understand the internal logic of game text expressions.

[0093] The self-supervised learning method is a machine learning method that lies between supervised learning and unsupervised learning. In self-supervised learning, the algorithm uses the unlabeled data itself to generate useful labels or supervision signals, thereby learning the internal structure of the data. The main goal of this method is to train the model by mining the inherent information in the data so that the model can learn useful representations from the data. In natural language processing, models such as the BERT model and the GPT model can be trained through self-supervised learning.

[0094] The first sample game corpus in Step S201 can be obtained from channels such as game expansion documents, game official websites, and game forum posts. The first sample game corpus can include but is not limited to at least one of gameplay corpus, game mechanism corpus, game strategy corpus, game analysis corpus, game plot interpretation corpus, and player game language. Among them, player game language refers to the slang and jargon among players. For example, "krypton gold" means recharge to purchase items.

[0095] After obtaining the first sample game corpus, the first sample game corpus can be preprocessed, specifically including masking the vocabulary and scrambling the sentence order of the first sample game corpus. Specifically, a preset proportion of vocabulary can be randomly masked and replaced with a [mask] mark, and the [mask] mark is used to indicate that the vocabulary at this position is masked. The preset proportion can be, for example, 15%, 20%, etc., which can be specifically set based on actual needs, and the order of sentences can be randomly swapped, so as to obtain the processed first sample game corpus.

[0096] As shown in Table 1, it is an example table of a first sample game corpus and an example of the processed first sample game corpus in the method for adjusting the appearance of a character provided in an embodiment of the present application.

[0097] Table 1.

[0098]

[0099]

[0100] In the above Table 1, when the first sample corpus is "Skill 1 can use a strong electric current to shock the enemy, causing a lot of damage and stunning the enemy. Its cooling time is 10 seconds and consumes 30 mana points", the masking words are "electric current, a lot of damage", and the corpus after the sentence order is disrupted is "Its cooling time is 10 seconds and consumes 30 mana points. Skill 1 uses a strong electric current to shock the enemy, causing a lot of damage and stunning the enemy", then the processed first sample corpus is "Its cooling time is 10 seconds and consumes 30 mana points. Skill 1 uses a strong [mask] to shock the enemy, causing [mask] and stunning the enemy"; when the first sample corpus is "Equipment 2 can provide a strong physical Defense. It has a high durability and can be used multiple times in battle. This equipment can also absorb a certain proportion of spell attacks and convert them into its own health value.", the masked words are "defense, durability", and the corpus after the sentence order is disrupted is "This equipment can also absorb a certain proportion of spell attacks and convert them into its own health value. It has a high durability and can be used multiple times in battle. Equipment 2 can provide strong physical defense", then the first sample corpus after processing is "This equipment can also absorb a certain proportion of spell attacks and convert them into its own health value. It has a high [mask] and can be used multiple times in battle. Equipment 2 can provide strong physical [mask]".

[0101] After obtaining the processed first sample game corpus, the processed first sample game corpus can be input into the language model to be trained, and the text features of the processed first sample game corpus are extracted through the multi-layer Transformer encoder of the language model to be trained. Based on the extracted text features, a first prediction task and a second prediction task are performed respectively. The first prediction task is to predict the masked words in the processed first sample game corpus, and the second prediction task is to predict the order of each sentence in the processed first sample game corpus.

[0102] Afterwards, the cross entropy loss corresponding to the first prediction task and the cross entropy loss corresponding to the second prediction task can be calculated respectively, and the model parameters of the language model to be trained can be adjusted so that the difference between the prediction result of the model to be trained for the first prediction task (that is, the predicted masked vocabulary) and the corresponding real masked vocabulary is within the second preset range, and the difference between the prediction result of the second prediction task (that is, the predicted sentence order) and the corresponding real sentence order is within the third preset range. That is, the optimization goal of the language model to be trained is that the difference between the output predicted masked vocabulary and the real masked vocabulary is within the second preset range, and the difference between the output predicted sentence order and the real sentence order is within the third preset range. In a specific implementation, the first preset range, the second preset range and the third preset range may be the same or different, and can be set according to actual needs. The present application does not impose any restrictions on this.

[0103] In this way, a trained first model can be obtained. Through the above training process, the trained first model has the ability to master the vocabulary, grammar and semantics of the game corpus, and has the ability to understand the internal logic of the game text expression. Afterwards, a pre-trained language model can be generated based on the trained first model.

[0104] In an optional specific implementation, the above step S205 may specifically be:

[0105] Inputting the first sample data into the first model so that the first model outputs predicted appearance adjustment data for the sample game experience data;

[0106] The model parameters of the second model are adjusted so that the difference between the predicted appearance adjustment data and the sample appearance adjustment data is within a first preset range, thereby obtaining a pre-trained language model.

[0107] In this implementation, the model parameters of the first model can be further optimized through the first sample data to train the first model to have the ability to map the game experience data into appearance adjustment parameters.

[0108] The preset training end condition may include but is not limited to: the accuracy of the predicted appearance adjustment data output by the first model reaches a threshold, the number of training times reaches a preset number of times, and the loss value of the loss function corresponding to the first model is within a preset loss value.

[0109] In the embodiment of the present application, by pre-training the language model to be trained in terms of game language understanding, a first model is obtained that has the ability to master the vocabulary, grammar and semantics of the game corpus and the ability to understand the internal logic of the game text expression. In this way, when the first model is trained to obtain a pre-trained language model that has the ability to map game experience data to appearance adjustment data, the training difficulty can be effectively reduced and the prediction ability of the pre-trained language model can be improved. In this way, the pre-trained language model can understand and analyze the game experience data and map the corresponding appearance adjustment parameters.

[0110] In another optional specific implementation, based on the obtained trained first model, the trained first model can be further trained to obtain a pre-trained language model that grasps the settings and rules of a specific game. Therefore, step S205 can also be implemented by the following steps S2051 to S2054:

[0111] Step S2051: constructing second sample data, wherein the second sample data includes a sample game question, a sample game statement for the sample game question, and a sample answer text for the sample game question;

[0112] Step S2052: inputting the second sample data into the first model, so that the first model respectively extracts the question semantic features corresponding to the sample game question and the statement semantic features corresponding to the sample game statement, predicts the answer probability of each statement text in the answer sequence for the sample game question based on the question semantic features and the statement semantic features, and generates a predicted answer text for the sample game question according to the answer probability;

[0113] Step S2053: adjusting the model parameters of the first model so that the difference between the predicted answer text and the sample answer text for the sample game question is within a fourth preset range, thereby obtaining a trained second model;

[0114] Step S2054: training the second model according to the first sample data until a preset training end condition is reached to obtain a pre-trained language model.

[0115] In the embodiment of the present application, firstly, second sample data can be constructed, and the second sample data is used to train the first model to master the settings and rules of a specific game. The second sample data can specifically include a sample game question, a sample game statement for the sample game question, and a sample answer text for the sample game question.

[0116] Optionally, the second sample data can be constructed in the following way:

[0117] Acquire a second sample game corpus and a sample game question and answer corpus, wherein the sample game question and answer corpus includes the sample game question and the sample answer text for the sample game question;

[0118] Extracting game concepts, game entities, and semantic relationships between game entities from the second sample game corpus;

[0119] Generate a structured game knowledge graph triple according to the game concept, the game entity, and the semantic relationship;

[0120] The second sample data is constructed according to the game knowledge graph triples and the sample game question and answer corpus.

[0121] It should be noted that the source of the second sample game corpus can be game manuals, game setting collections and other highly structured game materials. In addition to the game materials, the source of the first sample game corpus can also include game forum posts and other channels where players publish game-related information.

[0122] The sample game question-and-answer corpus can be question-and-answer dialogues related to game settings collected from game forum posts, player communities, and other channels. For example, the sample game question-and-answer corpus can be "Question: Can the Dragon Slaying Sword be used with one hand? Answer: The Dragon Slaying Sword is a two-handed weapon and must be held with both hands. It cannot be used with one hand."

[0123] In an embodiment of the present application, game concepts, game entities, and semantic relationships between game concepts and game entities can be extracted from the second sample game corpus. In a specific implementation, game concepts, game entities, and semantic relationships between game entities can be extracted from the second sample game corpus through a knowledge extraction model.

[0124] Among them, game concepts are abstract summaries of game elements, for example, equipment is divided into weapons and armor, and tasks are divided into main and side quests; game entities are objects with unique names in virtual games, for example, "Dragon Slaying Sword" is a specific weapon and equipment; semantic relationships describe the interaction between game concepts and game entities, for example, "Dragon Slaying Sword" contains the attribute of "high damage", and "intelligence" affects the value of "magic attack power".

[0125] After obtaining game knowledge such as game concepts, game entities, and semantic relationships between game concepts and game entities, the obtained game knowledge can be refined into game knowledge graph triples in the structure of <game entity, semantic relationship, game concept> through knowledge construction tools.

[0126] Exemplarily, a triplet of the game knowledge graph can be <Dragon Slaying Sword, is a kind of, weapon>, <Dragon Slaying Sword, part, two-handed>, <Dragon Slaying Sword, damage type, physical damage>, <Veros, positioning, BOSS>, <Veros, belongs to the camp, the Dark Order>, <Intelligence, influence, magic attack power>.

[0127] After generating structured game knowledge graph triples, second sample data can be constructed based on the game knowledge graph triples and the sample game question and answer corpus, so as to train the trained first model.

[0128] In an optional specific implementation, the step of "constructing the second sample data according to the game knowledge graph triples and the sample game question and answer corpus" may include the following steps:

[0129] Converting the game knowledge graph triples into game knowledge statement text;

[0130] The game knowledge statement text and the sample game question and answer corpus are combined to obtain the second sample data.

[0131] In the embodiment of the present application, the game knowledge graph triple can be converted into a game knowledge statement text. Specifically, a natural language template can be pre-set in the present application, and the game knowledge graph triple can be converted into a game knowledge statement text through the natural language template. For example, if the game knowledge graph triple is <Dragon Slaying Sword, Part, Both Hands>, ​​it can be converted into Dragon Slaying Sword is a weapon that needs to be held with both hands.

[0132] Afterwards, the game knowledge statement text and the sample game question and answer corpus can be combined to obtain a complete training corpus.

[0133] Afterwards, the complete training corpus can be converted into a sample data structure of a sample game question, a sample game statement for the sample game question, and a sample answer text for the sample game question, thereby obtaining second sample data. For example, the sample data structure can be in the format of "[sample game question] + [sample game statement 1] + [sample game statement 2] + [sample game statement 3] ... + [sample game statement n] + [sample answer text]".

[0134] It should be noted that the knowledge statements generated through the second sample game corpus are more comprehensive and systematic, and the knowledge statements are supplemented and expanded through the sample game question and answer corpus, making the constructed second sample data more accurate and comprehensive.

[0135] After obtaining the second sample data, the second sample data can be input into the aforementioned first model, and the problem semantic features corresponding to the sample game questions in the second sample data and the statement semantic features corresponding to the sample game statements in the second sample data are respectively obtained through the first model. Afterwards, based on the problem semantic features and the statement semantic features, the reply probability of each statement text in the answer sequence for the sample game question is predicted, and the predicted reply text for the sample game question is generated according to the reply probability. Afterwards, the cross entropy loss corresponding to the predicted reply text and the sample reply text is calculated, and the model parameters of the first model are adjusted so that the difference between the predicted reply text output by the first model and the corresponding sample reply text is within the fourth preset range. It should be noted that the fourth preset range and the aforementioned first preset range, second preset range and third preset range may be the same or different, and can be specifically set according to actual conditions.

[0136] In this way, a trained second model can be obtained. Through this training process, a second model that can master the settings and rules of a specific game can be trained on the basis of the first model that has the ability to master the vocabulary, grammar and semantics of the game corpus and the ability to understand the internal logic of the game text. That is, the trained second model can master both the general game language knowledge and the settings and rules of a specific game. Afterwards, the second model can be trained based on the above-mentioned first sample data to obtain a pre-trained language model.

[0137] In an optional specific implementation, the above step S2054 can be implemented in the following way:

[0138] inputting the first sample data into the second model so that the second model outputs predicted appearance adjustment data for the sample game experience data;

[0139] The model parameters of the second model are adjusted so that the difference between the predicted appearance adjustment data and the sample appearance adjustment data is within a first preset range, thereby obtaining a pre-trained language model.

[0140] In this implementation, the model parameters of the second model can be further optimized using the first sample data to train the second model to have the ability to map the game experience data into appearance adjustment parameters.

[0141] The preset training end condition may include but is not limited to: the accuracy of the predicted appearance adjustment data output by the second model reaches a threshold, the number of training times reaches a preset number of times, and the loss value of the loss function corresponding to the second model is within a preset loss value.

[0142] In the embodiment of the present application, by pre-training the language model to be trained in terms of game language understanding, and training the settings and rules of a specific game, a second model is obtained that has the ability to master the vocabulary, grammar and semantics of the game corpus, the ability to understand the internal logic of the game text expression, and the ability to master the settings and rules of a specific game. In this way, when the second model is trained to obtain a pre-trained language model that has the ability to map game experience data to appearance adjustment data, the training difficulty can be effectively reduced and the prediction ability of the pre-trained language model can be improved. In this way, the pre-trained language model can understand and analyze the game experience data and map the corresponding appearance adjustment parameters.

[0143] In an optional implementation, the above step S102 can be implemented by the following steps:

[0144] Performing labeling processing on the game experience data to obtain labeled game experience data;

[0145] According to a preset text structure template, the labeled game experience data is converted into a game experience description text;

[0146] The game experience description text is input into the pre-trained language model so that the pre-trained language model outputs appearance adjustment parameters.

[0147] In an embodiment of the present application, the game experience data can be labeled, specifically including but not limited to labeling the aforementioned task experience data according to task tags, labeling the aforementioned environment experience data according to environment tags, labeling the aforementioned combat experience data according to combat tags, labeling the aforementioned social experience data according to social tags, and labeling the aforementioned achievement experience data according to achievement tags.

[0148] Among them, task tags may include but are not limited to task type (such as main line or branch line), task duration, completion status, and ending choice; environmental tags may include but are not limited to tags including temperature, humidity, terrain, and ecology; combat tags may include but are not limited to combat type, combat intensity, skills used, damage output or damage tolerance; social tags may include but are not limited to interaction frequency, favorability, cooperation or competition; achievement tags may include but are not limited to achievement category, difficulty level, and achievement achievement time.

[0149] Afterwards, the tagged game experience data may be formatted according to a preset text structure template, thereby converting the tagged game experience data into a game experience description text.

[0150] As shown in Table 2, this is an example table of converting labeled game experience data into game experience description text in the method for adjusting the character appearance provided in an embodiment of the present application.

[0151] Table 2.

[0152]

[0153]

[0154] In the above Table 2, the task label of virtual character a is challenging the Abyss copy in the past 48 hours, the completion status is obtaining 12,000 experience points and legendary equipment, and the will attribute has been improved. The preset text structure template is "[Virtual character] obtained [experience value] points of experience / [item name]*[quantity] / [achievement title] through [experience type] [experience details] in [time period], and had [degree of influence] on [character attribute]". Based on the preset text structure template, the game experience data of virtual character a can be converted into the game experience description text "Virtual character a has obtained 12,000 experience points and legendary equipment Shadow Amulet*1 by continuously challenging the Abyss copy in the past 48 hours, which has greatly improved the will attribute."

[0155] In an optional implementation, the above step of "inputting the game experience description text into the pre-trained language model so that the pre-trained language model outputs the appearance adjustment parameter" can be implemented by the following steps:

[0156] Inputting the game experience description text into the pre-trained language model;

[0157] identifying key elements of the game experience description text by the pre-trained language model, so that the pre-trained language model generates matching appearance adjustment suggestions according to the key elements of the experience;

[0158] The appearance adjustment suggestion is structured to generate structured appearance adjustment parameters.

[0159] It should be noted that labeling the game experience data to obtain the game experience description text helps the pre-trained language model to identify the key elements of the experience in the game experience description text. The key elements may include but are not limited to key events (for example, challenging the abyss copy), related entities (for example, legendary equipment, will attributes), so that the pre-trained language model can combine the general game language knowledge and the settings and rules of the specific game (for example, shadow means darkness) to infer appearance changes (for example, the dark temperament is enhanced), and thus output appearance adjustment suggestions related to various parts of the virtual character in the form of natural language.

[0160] As shown in Table 3, this is an example table of an example of outputting appearance adjustment suggestions based on the game experience description text in the method for adjusting the appearance of a character provided in an embodiment of the present application.

[0161] Table 3.

[0162]

[0163]

[0164] In the above Table 3, the input game experience description text of the pre-trained language model is "In the past 48 hours, the virtual character a has continuously challenged the abyss dungeon, obtained 12,000 experience points and legendary equipment*1, and has greatly improved the will attribute", so that the key elements of the experience "shadow, will improvement" can be identified, thereby connecting the relationship between shadow, will and dark system, and proposing adjustment suggestions for the appearance of each part that are in line with the game tone and the player's personal experience. Specifically, the following appearance adjustment suggestions can be output: "Based on the recent experience of the virtual character a focusing on challenging high-difficulty copies of the shadow system, his character appearance should be more dark: the facial skin tone is deepened, and purple-black marks emerge; the pupil color becomes a deep dark gold, and the eyes become sharper; the armor incorporates shadow elements, and some areas are wrapped in dark phosphorescence; the cloak shape is changed to a body-protecting style, echoing the shadow amulet; the weapon is wrapped with a dark flame effect, and there is a shadow afterimage when it is swung."

[0165] In the embodiment of the present application, the appearance adjustment suggestion in the natural language form may be further structured to generate structured appearance adjustment parameters.

[0166] As shown in Table 4, it is an example table of an example of generating appearance adjustment parameters from appearance adjustment suggestions in the method for adjusting the appearance of a character provided in an embodiment of the present application.

[0167] Table 4.

[0168]

[0169] In the above Table 4, the suggestion for appearance adjustment is "Based on the recent experience of virtual character a focusing on challenging high-difficulty shadow dungeons, his character appearance should be more dark-style: facial skin tone deepens, purple-black marks emerge; eye color becomes deep dark gold, and the eyes become sharper; armor incorporates shadow elements, and some areas are wrapped in dark phosphorescence; the cloak shape is changed to a body-protecting style, echoing the shadow amulet; the weapon is wrapped with a dark flame effect, and there is a shadow afterimage when waving", and the generated structured appearance adjustment parameters are "skin tone: -20% brightness, +30% darkness; eye color: #c39a63; armor material: #1b0d27+10% phosphorescence intensity".

[0170] It should be noted that, compared with appearance adjustment suggestions in natural language form, structured appearance adjustment parameters are easier to be read by machines, so that the appearance assets of the virtual character can be efficiently adjusted accordingly based on the appearance adjustment parameters.

[0171] In an optional implementation, the appearance assets adjusted in step S103 include but are not limited to at least one of the following: model assets, texture assets, clothing assets, and special effect assets.

[0172] In an optional implementation, the appearance adjustment parameters include an appearance tag and an adjustment parameter for the appearance tag. The step of “adjusting the appearance asset of the virtual character” in step S103 can be implemented by the following steps S1031 to S1034:

[0173] Step S1031: determining a matching appearance asset according to the appearance tag;

[0174] Step S1032: performing detail adjustment on the appearance asset according to the adjustment parameters to obtain an adjusted appearance asset;

[0175] Step S1033: determining an adjusted character appearance model corresponding to the virtual character according to the adjusted appearance asset;

[0176] Step S1034: Rendering the adjusted character appearance model.

[0177] In the embodiment of the present application, the appearance asset matching the appearance tag in the appearance adjustment parameter can be determined. Specifically, step S1031 may include the following steps:

[0178] Determine a matching character 3D model based on the body shape label;

[0179] Determining a matching three-dimensional clothing model according to a clothing label, wherein the clothing label includes a style label, a style label, and a covering part label on the virtual character;

[0180] According to the skin color label, determine the matching skin texture map;

[0181] Determine a matching clothing texture map according to a clothing pattern label and / or a clothing material label;

[0182] Determine matching special effect assets based on character ability tags.

[0183] Optionally, step S1032 may include the following steps:

[0184] If the appearance asset is a three-dimensional model of a character, deforming the three-dimensional model of the character according to a height adjustment parameter and / or a body shape adjustment parameter;

[0185] If the appearance asset is a three-dimensional facial model, adjusting corresponding parts of the three-dimensional facial model according to the facial adjustment parameters;

[0186] If the appearance asset is a three-dimensional clothing model, adjusting the three-dimensional clothing model according to the body proportion parameter;

[0187] If the appearance asset is a skin texture map, adjusting the color of the skin texture map according to skin color parameters;

[0188] If the appearance asset is a clothing texture map, the color of the clothing texture map is adjusted according to the clothing color parameter.

[0189] With respect to the body shape and appearance adjustment parameters, the embodiment of the present application can specifically change the body shape in the following ways:

[0190] If the body shape label is getting fatter, you can choose a character 3D model that is fatter than the current character 3D model currently used by the virtual character; if the body shape label is getting thinner, you can choose a character 3D model that is thinner than the current character 3D model currently used by the virtual character; if the body shape label is getting taller, you can choose a character 3D model that is taller than the current character 3D model currently used by the virtual character; if the body shape label is getting shorter, you can choose a character 3D model that is shorter than the current character 3D model currently used by the virtual character.

[0191] Afterwards, the selected character 3D model can be deformed and scaled according to the height adjustment parameter and / or the body adjustment parameter in the body appearance adjustment parameter, so that the deformed character 3D model meets the body appearance adjustment parameter. For example, if the body appearance adjustment parameter is to become taller, and the height adjustment parameter is +20%, a taller character 3D model can be selected, and the selected character 3D model can be controlled to deform so that the height of the deformed character 3D model is 20% of the height of the current character 3D model.

[0192] With respect to clothing appearance adjustment parameters, the embodiment of the present application can specifically change clothing in the following manner:

[0193] If the clothing label is to become fashionable and gorgeous, you can choose a clothing 3D model that is more fashionable and gorgeous than the current clothing 3D model currently used by the virtual character; if the clothing label is to become dull and old, you can choose a clothing 3D model that is duller and older than the current clothing 3D model currently used by the virtual character; if the clothing label is to change short sleeves into vests, you can choose clothing 3D models related to vests.

[0194] Afterwards, the selected clothing 3D model can be adjusted according to the body proportion parameter in the clothing appearance adjustment parameter, so that the adjusted clothing 3D model meets the clothing appearance adjustment parameter. For example, if the clothing appearance adjustment parameter is gorgeous and fashionable, and the body proportion is 1:2.5 for the head and upper body ratio, a more gorgeous and fashionable clothing 3D model can be selected, and the selected clothing 3D model can be adjusted according to the length of the head, so that the adjusted clothing 3D model can cover the upper body with a length 2.5 times the length of the head.

[0195] Regarding the skin appearance adjustment parameters, the embodiment of the present application can specifically change the skin in the following ways:

[0196] If the skin color label is darkening, you can choose a skin texture map that is darker than the current skin texture map currently used by the virtual character; if the skin color label is whitening, you can choose a skin texture map that is whiter than the current skin texture map currently used by the virtual character; if the skin color label is yellowing, you can choose a skin texture map that is yellower than the current skin texture map currently used by the virtual character.

[0197] Afterwards, the selected skin texture map can be adjusted according to the skin color parameter in the skin appearance adjustment parameter, so that the adjusted skin texture map conforms to the skin appearance adjustment parameter. For example, if the skin appearance adjustment parameter is whitening and the skin color parameter is +10%, a whiter skin texture map can be selected and the selected skin texture map can be controlled to adjust so that the whiteness increases by 10%.

[0198] It should be noted that clothing pattern labels refer to designs or patterns on clothing, such as stripes, dots, patterns, etc. These patterns can be added to clothing three-dimensional models through separate clothing texture maps to increase visual diversity and personalization. Clothing material labels refer to the material properties of the surface of clothing three-dimensional models, such as the texture of different materials such as silk, cotton, leather, etc. These materials can be represented by high-precision clothing texture maps, and can also be combined with technologies such as normal maps and gloss maps to enhance the sense of reality. In the embodiment of the present application, after determining the clothing texture map, the color of the clothing texture map can be adjusted according to the clothing color parameters. For example, if the clothing texture appearance adjustment parameter is red and white stripes, the clothing texture map with a striped style can be selected, and its stripe color can be changed to red and white.

[0199] In an embodiment of the present application, the character ability tags may include but are not limited to attack power, defense power, magic power, strength, agility, intelligence, endurance and the like. Matching special effect assets can be selected through the character ability tags. For example, if the agility is high, special effects such as flash or particle effects can be added around the virtual character to reflect the faster movement speed of the virtual character.

[0200] Through the above manner, it is possible to determine an appearance asset that matches the appearance adjustment parameter for the virtual character, and to adjust the appearance asset accordingly according to the appearance adjustment parameter, so that the adjusted appearance asset better conforms to the appearance adjustment parameter.

[0201] In an optional implementation, when the adjusted appearance asset may include multiple assets, the above step S1033 may be implemented by the following steps:

[0202] Combining the adjusted appearance assets to obtain an adjusted character appearance model corresponding to the virtual character;

[0203] The asset combination includes at least one of the following: model nesting, bone binding, texture baking, and special effect integration.

[0204] In an optional specific implementation, when the asset combination includes model nesting, bone binding, texture baking, and special effect integration, the step of "assembling the adjusted appearance assets to obtain the adjusted character appearance model corresponding to the virtual character" may include the following steps:

[0205] The decoration three-dimensional model and the character three-dimensional model are nested and assembled to generate a first model covered with the decoration;

[0206] Binding the first model to the character skeleton to obtain a second model with the skeleton bound;

[0207] Assigning the texture map to the surface of the second model to obtain a third model with appearance details;

[0208] The special effect assets are integrated into the third model to obtain an adjusted character appearance model corresponding to the virtual character.

[0209] In an embodiment of the present application, models of clothing, accessories, etc. can be nested and assembled with a basic human body model to form a hierarchical character model, which is a first model covered with decorations; then, the first model can be bound to the character skeleton so that the appearance can change with the character's movements, thereby obtaining a second model with bound skeletons; then, skin texture maps, clothing texture maps, and other maps can be baked onto the second model to achieve a fine expression of the appearance, thereby obtaining a third model with appearance details; finally, special effects assets such as light effects and particles can be integrated into the third model to enhance the dynamic expressiveness of the appearance, thereby obtaining an adjusted character appearance model corresponding to the virtual character.

[0210] Afterwards, the adjusted character appearance model may be rendered so that the virtual character presents an appearance performance effect that matches the game experience of the virtual character in the graphical user interface.

[0211] The following combination Figure 4 and Figure 5 The change of the appearance on the interface in the method for adjusting the appearance of a character provided in an embodiment of the present application is introduced:

[0212] like Figure 4 As shown, it is an interface diagram of an example of the appearance change of a virtual character performing a task in a flame zone in the method for adjusting the character appearance provided in an embodiment of the present application, including interface (4-a) and interface (4-b). In interface (4-a), the virtual character 401 has not yet performed a task in the flame zone, and the current appearance model is the initial appearance model. In interface (4-b), the virtual character 401 is performing a task in the flame zone, and the initial appearance model will be adjusted. Specifically, flame special effects as shown in 402, 403 and 404 are added to the initial appearance model.

[0213] like Figure 5 As shown, it is an interface diagram of an example of the appearance change of a virtual character that has experienced a battle in the method for adjusting the appearance of a character provided in an embodiment of the present application, including interface (5-a) and interface (5-b). In interface (5-a), the virtual character 401 has not experienced a battle, and the current appearance model is the initial appearance model. In interface (5-b), the virtual character 401 has experienced a battle, and the initial appearance model will be adjusted. Specifically, a scar as shown in 405 is added to the initial appearance model.

[0214] In an optional implementation, the above step S1034 may be implemented in the following manner:

[0215] Generating a transition effect according to the initial appearance model of the virtual character and the adjusted character appearance model;

[0216] Rendering the transition effect, and after the rendering of the transition effect is completed, rendering the adjusted character appearance model to switch the initial appearance model presented in the graphical user interface to the adjusted character appearance model.

[0217] In an embodiment of the present application, a transition effect can be generated based on the initial appearance model of the virtual character and the adjusted character appearance model. The transition effect can specifically include but is not limited to at least one of a gradient fusion effect, a dynamic interpretation effect, and a situational rendering effect.

[0218] Among them, the gradient fusion special effect refers to the use of gradient transparency, dissolution, floodlight and other effects to achieve a smooth transition between the old and new appearances; the dynamic interpretation special effect refers to the use of morphological animation, special effects animation and other means to express the appearance switching as a dynamic interpretation process; the situational rendering special effect refers to the combination of light and shadow, post-production, lens and other means to render an atmospheric appearance transformation situation.

[0219] In the embodiment of the present application, the initial appearance model of the virtual character can be smoothly transitioned to the adjusted character appearance model through the above-mentioned appearance generation engine, thereby avoiding visual abruptness during direct switching, thereby improving the player's gaming experience.

[0220] In an optional specific implementation, the method provided in the embodiment of the present application can be applied to a game client, which can have an attached Figure 1 The experience collection plugin, pre-trained language model, and appearance generation engine shown in .

[0221] In this way, the above step S101 can be performed through the experience collection plug-in, and the collected game experience data corresponding to the virtual character is uploaded and stored in a storage location associated with the virtual character in the experience database of the game client in a time series manner. After the pre-trained language model pulls the game experience data corresponding to the virtual character from the storage location, the above step S102 is performed through the pre-trained language model, the game experience data is mapped to the appearance adjustment parameters corresponding to the virtual character, and the appearance adjustment parameters are sent to the appearance generation engine. In this way, after receiving the appearance adjustment parameters corresponding to the virtual character, the appearance generation engine performs the above step S103 through the appearance generation engine to adjust the appearance assets of the virtual character accordingly.

[0222] The method for adjusting the appearance of a character provided in the embodiment of the present application has the following advantages: First, this solution can realize the mapping of game experience to character appearance based on the game experience data and natural language understanding technology corresponding to the appearance generation engine, so that the character image can be dynamically adjusted with the game experience, which greatly improves the specificity of the character; Second, this solution utilizes the knowledge understanding and generation capabilities of the language model, and through flexible template guidance, it can convert the game experience data into appearance adjustment parameters that conform to the game world view, thereby enhancing the semantic rationality and narrative relevance of the appearance adjustment; Third, this solution uses labeled game experience data to generate character appearance, breaking through the current technology that only provides a preset limited appearance combination, and can flexibly adjust the detailed performance of various parts of the virtual character, presenting richer and more realistic visual changes, and enhancing the immersiveness of the appearance; Fourth, this solution can realize real-time dynamic appearance generation, so that the appearance of the virtual character can be continuously adjusted as the game experience accumulates, so that players can intuitively feel the growth of the virtual character through the appearance, deepening the sense of game substitution.

[0223] Corresponding to the method for adjusting the appearance of a character provided in the first embodiment of the present application, the second embodiment of the present application further provides a device for adjusting the appearance of a character, which is applied to a game scene, wherein the game scene at least includes a virtual character controlled by a game account, such as Figure 6 As shown, the device 600 for adjusting the appearance of a character includes:

[0224] The collecting unit 601 is used to collect game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data;

[0225] A determination unit 602, configured to determine appearance adjustment parameters corresponding to the virtual character according to the game experience data;

[0226] The adjustment unit 603 is used to adjust the appearance assets of the virtual character according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0227] Optionally, the adjustment unit 603 is specifically used to adjust at least one of the following appearance assets: model assets, texture assets, clothing assets, and special effect assets.

[0228] Optionally, the appearance adjustment parameter includes an appearance tag and an adjustment parameter for the appearance tag; the adjustment unit 603 is specifically configured to:

[0229] Determining a matching appearance asset according to the appearance tag;

[0230] According to the adjustment parameters, the appearance asset is adjusted in detail to obtain an adjusted appearance asset;

[0231] Determining an adjusted character appearance model corresponding to the virtual character according to the adjusted appearance asset;

[0232] Rendering the adjusted character appearance model.

[0233] Optionally, the adjusted appearance assets include multiple ones, and the adjusting unit 603 is specifically used to:

[0234] Combining the adjusted appearance assets to obtain an adjusted character appearance model corresponding to the virtual character;

[0235] The asset combination includes at least one of the following: model nesting, bone binding, texture baking, and special effect integration.

[0236] Optionally, the adjusting unit 603 is specifically used for at least one of the following steps:

[0237] Determine a matching character 3D model based on the body shape label;

[0238] Determining a matching three-dimensional clothing model according to a clothing label, wherein the clothing label includes a style label, a style label, and a covering part label on the virtual character;

[0239] According to the skin color label, determine the matching skin texture map;

[0240] Determine a matching clothing texture map according to a clothing pattern label and / or a clothing material label;

[0241] Determine matching special effect assets based on character ability tags.

[0242] Optionally, the adjusting unit 603 is specifically configured to:

[0243] If the appearance asset is a three-dimensional model of a character, deforming the three-dimensional model of the character according to a height adjustment parameter and / or a body shape adjustment parameter;

[0244] If the appearance asset is a three-dimensional facial model, adjusting corresponding parts of the three-dimensional facial model according to the facial adjustment parameters;

[0245] If the appearance asset is a three-dimensional clothing model, adjusting the three-dimensional clothing model according to the body proportion parameter;

[0246] If the appearance asset is a skin texture map, adjusting the color of the skin texture map according to skin color parameters;

[0247] If the appearance asset is a clothing texture map, the color of the clothing texture map is adjusted according to the clothing color parameter.

[0248] Optionally, when the asset combination includes model nesting, skeleton binding, texture baking, and special effect integration, the adjustment unit 603 is specifically used to:

[0249] The decoration three-dimensional model and the character three-dimensional model are nested and assembled to generate a first model covered with the decoration;

[0250] Binding the first model to the character skeleton to obtain a second model with the skeleton bound;

[0251] Assigning the texture map to the surface of the second model to obtain a third model with appearance details;

[0252] The special effect assets are integrated into the third model to obtain an adjusted character appearance model corresponding to the virtual character.

[0253] Optionally, the adjusting unit 603 is specifically configured to:

[0254] Generating a transition effect according to the initial appearance model of the virtual character and the adjusted character appearance model;

[0255] Rendering the transition special effect, and after the rendering of the transition special effect is completed, rendering the adjusted character appearance model.

[0256] Optionally, the determining unit 602 is specifically configured to:

[0257] Performing labeling processing on the game experience data to obtain labeled game experience data;

[0258] According to a preset text structure template, the labeled game experience data is converted into a game experience description text;

[0259] The game experience description text is input into the pre-trained language model so that the pre-trained language model outputs appearance adjustment parameters.

[0260] Optionally, the determining unit 602 is specifically configured to:

[0261] Inputting the game experience description text into the pre-trained language model;

[0262] identifying key elements of the game experience description text by the pre-trained language model, so that the pre-trained language model generates matching appearance adjustment suggestions according to the key elements of the experience;

[0263] The appearance adjustment suggestion is structured to generate structured appearance adjustment parameters.

[0264] Corresponding to the method for adjusting the appearance of a character provided in the first embodiment of the present application, the third embodiment of the present application further provides an electronic device for adjusting the appearance of a character. Figure 7 As shown, the electronic device 700 includes: a processor 701; and a memory 702, which is used to store a program of a method for adjusting the appearance of a character. After the device is powered on and the program of the method for adjusting the appearance of a character is run by the processor, the following steps are performed:

[0265] Collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data;

[0266] Determining appearance adjustment parameters corresponding to the virtual character according to the game experience data;

[0267] The appearance assets of the appearance of the virtual character are adjusted according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0268] Corresponding to the method for adjusting the appearance of a character provided in the first embodiment of the present application, the fourth embodiment of the present application provides a computer-readable storage medium storing a program for the method for adjusting the appearance of a character, the program being executed by a processor to perform the following steps:

[0269] Collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data;

[0270] Determining appearance adjustment parameters corresponding to the virtual character according to the game experience data;

[0271] The appearance assets of the appearance of the virtual character are adjusted according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

[0272] It should be noted that for the detailed description of the devices, electronic devices and computer-readable storage media provided in the second, third and fourth embodiments of the present application, reference can be made to the relevant description of the first embodiment of the present application, and no further details will be given here.

[0273] Although the present application is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present application. Any technical personnel in this field may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

[0274] In a typical configuration, a node device in a blockchain includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0275] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0276] 1. Computer-readable media includes permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), random access memory (RAM) of other attributes, read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage media or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined in this article, computer-readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0277] 2. Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems or computer program products. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0278] Although the present application is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present application. Any technical personnel in this field may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

Claims

1. A method for adjusting the appearance of a character, characterized in that: Applied in a game scene, the game scene at least includes a virtual character controlled by a game account, and the method includes: Collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data; Determining appearance adjustment parameters corresponding to the virtual character according to the game experience data; The appearance assets of the virtual character are adjusted according to the appearance adjustment parameters so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

2. The method according to claim 1, characterized in that The appearance assets include at least one of the following: model assets, texture assets, clothing assets, and special effect assets.

3. The method according to claim 1, characterized in that: The appearance adjustment parameters include an appearance tag and an adjustment parameter for the appearance tag; and adjusting the appearance asset of the virtual character includes: Determining a matching appearance asset according to the appearance tag; According to the adjustment parameters, the appearance asset is adjusted in detail to obtain an adjusted appearance asset; Determining an adjusted character appearance model corresponding to the virtual character according to the adjusted appearance asset; Rendering the adjusted character appearance model.

4. The method according to claim 3, characterized in that: The adjusted appearance assets include a plurality of assets, and determining the adjusted character appearance model corresponding to the virtual character according to the adjusted appearance assets includes: Combining the adjusted appearance assets to obtain an adjusted character appearance model corresponding to the virtual character; The asset combination includes at least one of the following: model nesting, bone binding, texture baking, and special effect integration.

5. The method according to claim 3, characterized in that: The determining, according to the appearance tag, a matching appearance asset includes at least one of the following: Determine a matching character 3D model based on the body shape label; Determining a matching three-dimensional clothing model according to a clothing label, wherein the clothing label includes a style label, a style label, and a covering part label on the virtual character; According to the skin color label, determine the matching skin texture map; Determine a matching clothing texture map according to a clothing pattern label and / or a clothing material label; Determine matching special effect assets based on character ability tags.

6. The method according to claim 3, characterized in that The step of adjusting the appearance asset in detail according to the adjustment parameter to obtain the adjusted appearance asset includes: If the appearance asset is a three-dimensional model of a character, deforming the three-dimensional model of the character according to a height adjustment parameter and / or a body shape adjustment parameter; If the appearance asset is a three-dimensional facial model, adjusting corresponding parts of the three-dimensional facial model according to the facial adjustment parameters; If the appearance asset is a three-dimensional clothing model, adjusting the three-dimensional clothing model according to the body proportion parameter; If the appearance asset is a skin texture map, adjusting the color of the skin texture map according to skin color parameters; If the appearance asset is a clothing texture map, the color of the clothing texture map is adjusted according to the clothing color parameter.

7. The method according to claim 4, characterized in that In the case where the asset combination includes model nesting, bone binding, texture baking, and special effect integration, the asset combination of the adjusted appearance assets to obtain the adjusted character appearance model corresponding to the virtual character includes: The decoration three-dimensional model and the character three-dimensional model are nested and assembled to generate a first model covered with the decoration; Binding the first model to the character skeleton to obtain a second model with the skeleton bound; Assigning the texture map to the surface of the second model to obtain a third model with appearance details; The special effect assets are integrated into the third model to obtain an adjusted character appearance model corresponding to the virtual character.

8. The method according to claim 3, characterized in that The rendering of the adjusted character appearance model comprises: Generating a transition effect according to the initial appearance model of the virtual character and the adjusted character appearance model; Rendering the transition special effect, and after the rendering of the transition special effect is completed, rendering the adjusted character appearance model.

9. The method according to claim 1, characterized in that: Determining the appearance adjustment parameters corresponding to the virtual character according to the game experience data includes: Performing labeling processing on the game experience data to obtain labeled game experience data; According to a preset text structure template, the labeled game experience data is converted into a game experience description text; The game experience description text is input into a pre-trained language model so that the pre-trained language model outputs appearance adjustment parameters.

10. The method according to claim 9, characterized in that The step of inputting the game experience description text into the pre-trained language model so that the pre-trained language model outputs the appearance adjustment parameter comprises: Inputting the game experience description text into the pre-trained language model; identifying, by means of the pre-trained language model, key elements of the experience in the game experience description text, so that the pre-trained language model generates matching appearance adjustment suggestions according to the key elements of the experience; The appearance adjustment suggestion is structured to generate structured appearance adjustment parameters.

11. A device for adjusting the appearance of a character, characterized in that: Applied in a game scene, the game scene at least includes a virtual character controlled by a game account, and the device includes: A collection unit, used for collecting game experience data corresponding to the virtual character, wherein the game experience data includes at least one of the following: task experience data, environment experience data, combat experience data, social experience data, and achievement experience data; A determination unit, configured to determine appearance adjustment parameters corresponding to the virtual character according to the game experience data; The adjustment unit is used to adjust the appearance assets of the virtual character according to the appearance adjustment parameters, so that the appearance performance effect of the virtual character matches the game experience of the virtual character.

12. An electronic device, characterized in that: include: processor; as well as The memory is used to store a data processing program. After the electronic device is powered on and the program is run by the processor, the method according to any one of claims 1 to 10 is executed.

13. A computer-readable storage medium, characterized in that: A data processing program is stored, and the program is run by a processor to execute the method according to any one of claims 1 to 10.

Images