Interaction processing method and apparatus for virtual scene, electronic device, computer-readable storage medium, and computer program product

Abstract

Provided in the present application are an interaction processing method and apparatus for a virtual scene, an electronic device, a computer-readable storage medium and a computer program product. The method comprises: displaying a virtual scene, the virtual scene comprising at least one virtual object; in response to a pickup condition being met, controlling the first virtual object to pick up a first virtual prop, the first virtual object being any virtual object in the virtual scene; and in response to a trigger operation for the first virtual prop, controlling the first virtual object to implement an action sequence on the basis of the first virtual prop, the action sequence comprising a prop action adapted to the first virtual prop and an object action adapted to the first virtual object. The present application can improve the diversity and adaptability of actions during interaction between virtual objects and virtual props.

Description

Interactive processing methods, devices, electronic devices, computer-readable storage media, and computer program products for virtual scenes

[0001] Cross-references to related applications

[0002] This application is based on and claims priority to Chinese Patent Application No. 2024117680580, filed on December 3, 2024, the entire contents of which are incorporated herein by reference. Technical Field

[0003] This application relates to computer technology, and more particularly to a method, apparatus, electronic device, computer-readable storage medium, and computer program product for interactive processing of virtual scenes. Background Technology

[0004] Taking game scenarios as an example, the interaction between virtual objects and virtual items can bring players a better gaming experience. However, current technologies have limitations in the interaction methods between virtual objects and virtual items; they are relatively simple and cannot adapt to the increasingly diverse nature of virtual items in virtual scenarios. This fails to meet players' needs for diverse interaction methods, resulting in a poor player experience. Summary of the Invention

[0005] This application provides a method, apparatus, electronic device, computer-readable storage medium, and computer program product for interactive processing of virtual scenes, which can improve the smoothness and diversity of interaction between virtual objects and virtual props.

[0006] The technical solution of this application embodiment is implemented as follows:

[0007] This application provides an interactive processing method for a virtual scene, executed by an electronic device, the method comprising:

[0008] Displaying a virtual scene, wherein the virtual scene includes at least one virtual object;

[0009] In response to the fulfillment of the picking conditions, the first virtual object is controlled to pick up the first virtual prop, wherein the first virtual object is any one of the virtual objects in the virtual scene;

[0010] In response to a trigger operation on the first virtual prop, the first virtual object is controlled to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

[0011] This application provides an interactive processing device for a virtual scene, the device comprising:

[0012] A display module is used to display a virtual scene, wherein the virtual scene includes at least one virtual object;

[0013] A picking module is used to control a first virtual object to pick up a first virtual prop in response to the fulfillment of picking conditions, wherein the first virtual object is any one of the virtual objects in the virtual scene;

[0014] A control module is configured to respond to a trigger operation on the first virtual prop and control the first virtual object to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

[0015] This application provides an electronic device, the electronic device comprising:

[0016] Memory is used to store executable instructions or computer programs.

[0017] The processor, when executing computer-executable instructions or computer programs stored in the memory, implements the interactive processing method for virtual scenes provided in the embodiments of this application.

[0018] This application provides a computer-readable storage medium storing a computer program or computer-executable instructions, which, when executed by a processor, implements the interactive processing method for a virtual scene provided in this application.

[0019] This application provides a computer program product, including a computer program or computer executable instructions. When the computer program or computer executable instructions are executed by a processor, they implement the interactive processing method for virtual scenes provided in this application.

[0020] The embodiments of this application have the following beneficial effects:

[0021] When the pickup conditions are met, the first virtual object in the virtual scene can be controlled to pick up the first virtual item. When the first virtual object picks up the first virtual item, the first virtual object can be triggered to perform an action sequence based on the first virtual item. This allows the action sequence performed during the interaction between the virtual object and the virtual item to retain both the object's actions and the item's actions, thereby improving the diversity and adaptability of actions during the interaction between the virtual object and the virtual item and enhancing the player's experience. Attached Figure Description

[0022] Figure 1 is a schematic diagram of the architecture of the virtual scene interactive processing system 100 provided in an embodiment of this application;

[0023] Figure 2 is a structural schematic diagram of the terminal 400 provided in an embodiment of this application;

[0024] Figure 3A is a schematic diagram of the first process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0025] Figure 3B is a schematic diagram of the second process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0026] Figure 3C is a schematic diagram of the third process of the virtual scene interaction processing method provided in the embodiments of this application;

[0027] Figure 3D is a schematic diagram of the fourth process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0028] Figure 3E is a schematic diagram of the fifth process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0029] Figure 3F is a schematic diagram of the sixth process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0030] Figure 3G is a schematic diagram of the seventh process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0031] Figure 3H is a schematic diagram of the eighth process of the interactive processing method for virtual scenes provided in the embodiments of this application;

[0032] Figure 3I is a ninth flowchart illustrating the interactive processing method for virtual scenes provided in the embodiments of this application;

[0033] Figure 4 is a first schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0034] Figure 5 is a second schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0035] Figure 6A is a third schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0036] Figure 6B is a fourth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0037] Figure 6C is a fifth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0038] Figure 6D is a sixth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0039] Figure 6E is a seventh schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0040] Figure 6F is the eighth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application;

[0041] Figure 7A is a first schematic diagram of the picking process provided in an embodiment of this application;

[0042] Figure 7B is a second schematic diagram of the picking process provided in an embodiment of this application;

[0043] Figure 7C is a third schematic diagram of the picking process provided in an embodiment of this application;

[0044] Figure 8A is a schematic diagram of an attack-defensible embodiment provided in this application;

[0045] Figure 8B is a schematic diagram of an undefendable attack provided in an embodiment of this application;

[0046] Figure 9A is a first schematic diagram of a virtual scene provided in an embodiment of this application;

[0047] Figure 9B is a second schematic diagram of the virtual scene provided in the embodiment of this application;

[0048] Figure 9C is a third schematic diagram of the virtual scene provided in the embodiment of this application;

[0049] Figure 10A is a fourth schematic diagram of the virtual scene provided in the embodiments of this application;

[0050] Figure 10B is a fifth schematic diagram of the virtual scene provided in the embodiments of this application;

[0051] Figure 11A is a sixth schematic diagram of the virtual scene provided in the embodiment of this application;

[0052] Figure 11B is a seventh schematic diagram of the virtual scene provided in the embodiments of this application;

[0053] Figure 11C is an eighth schematic diagram of the virtual scene provided in the embodiments of this application;

[0054] Figure 12 is a schematic flowchart of the picking process provided in an embodiment of this application;

[0055] Figure 13 is a schematic diagram of action fusion provided in an embodiment of this application;

[0056] Figure 14A is a ninth schematic diagram of the virtual scene provided in the embodiment of this application;

[0057] Figure 14B is a tenth schematic diagram of the virtual scene provided in the embodiment of this application;

[0058] Figure 14C is the eleventh schematic diagram of the virtual scene provided in the embodiments of this application.

[0059] It should be noted that the terms "first" and "second" mentioned above are only used to distinguish between different options and do not represent the degree of superiority or inferiority of the options or their priority in the implementation process. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0062] In the following description, the terms "first / second / third" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first / second / third" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0063] In this application embodiment, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0064] Unless otherwise specified, "at least one" as used below refers to one or more cases, and "multiple" can refer to two or more cases.

[0065] Unless otherwise defined, all technical and scientific terms used in the embodiments of this application have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in the embodiments of this application is for the purpose of describing the embodiments of this application only and is not intended to limit this application.

[0066] In the implementation of this application, the collection and processing of relevant data should strictly comply with the requirements of relevant laws and regulations, obtain the informed consent or separate consent of the personal information subject, and carry out subsequent data use and processing within the scope of laws and regulations and the authorization of the personal information subject.

[0067] Before providing a further detailed description of the embodiments of this application, the nouns and terms involved in the embodiments of this application will be explained, and the nouns and terms involved in the embodiments of this application shall be interpreted as follows.

[0068] 1) Responding to: used to indicate the conditions or states on which the operation is performed depends. When the conditions or states on which it depends are met, one or more operations can be performed in real time or with a set delay. Unless otherwise specified, there is no restriction on the order in which the multiple operations are performed.

[0069] 2) Human-Computer Interaction Interface: An interface used to provide human-computer interaction functions. Examples include graphical user interfaces (GUIs), augmented reality (AR) interfaces, virtual reality (VR) interfaces, voice user interfaces (VUIs), interactive projection interfaces (using projection technology to display information on a flat surface), eye-tracking interfaces (interfaces controlled by detecting the user's gaze), holographic interfaces (three-dimensional holograms formed by projecting images using holographic projection technology, allowing viewing of stereoscopic images without special glasses), multimodal interfaces (interfaces combining multiple interaction methods, such as tactile, visual, and auditory interaction), and brain-machine interfaces (BMIs).

[0070] 3) Virtual Scene: This is the scene displayed (or provided) by the application when it runs on the terminal device. This scene can be a simulation of the real world, a semi-simulated / semi-fictional virtual environment, or a purely fictional virtual environment. Users can control virtual objects to move within this virtual scene. For example, in a game scene, the virtual scene could be the setting of a game match.

[0071] 4) Virtual Objects: Images of various people and objects that interact within a virtual scene, or movable objects within the virtual scene. These movable objects can be virtual characters, virtual animals, virtual buildings, etc. A virtual scene can include multiple virtual objects, each with its own shape and volume, occupying a portion of the space within the virtual scene. For example, in a game scene, a virtual object can be a player character (PC) controlled by the user, or a non-player character (NPC) automatically generated by the game system.

[0072] 5) Usage Mode: The mode in which virtual objects use virtual items to achieve interactive goals. For example, it can be an attack mode, where virtual objects use virtual items to attack the enemy; it can also be a defense mode, where virtual objects use virtual items to defend against the enemy's attack; or it can be a counter-attack mode, where virtual objects use virtual items to defend against the enemy's attack and launch a counter-attack.

[0073] 6) Representative Actions: Actions that embody the character traits of a virtual object. These can also be called unique or specialized actions. A representative action is tied to a specific type of virtual object and is not applicable to other types of virtual objects. For example, a brave knight's representative action might be to step forward, draw his sword, and point it at the enemy, demonstrating fearlessness and a sense of justice. A cunning enemy's representative action might be to stealthily move about or quickly move from one place to another, showcasing their agility.

[0074] 7) General Actions: Actions applicable to virtual objects with different character characteristics in a virtual scene. These actions are general and do not represent the character characteristics of the virtual object. Examples include movement actions (such as walking, running, or jumping), interaction actions (such as picking up or placing items, opening or closing, etc.), basic facial expressions (such as crying, laughing, etc.), and social actions (such as talking, nodding, shaking one's head, etc.).

[0075] 8) Action Sequence: A sequence of actions arranged in a specific order, used to describe the execution order of actions in a complex behavior or task, determining the behavior and logical flow of a virtual object during interaction with virtual props. For example, the action sequence for a character attack is: preparatory stance, attack action, and subsequent action; the action sequence for a character interaction is: approaching the target, performing the interaction action, and leaving the target; the action sequence for a character skill release is: energy gathering action (the character gathers energy to prepare for skill release), skill release, and recovery action.

[0076] 9) Object Actions: These are actions that conform to the characteristics and behavioral patterns of virtual objects. They are specific actions designed for interaction with virtual objects in the virtual environment, and are usually associated with the object's attributes, functions, and the interaction methods between the player and the object. They can include any interactive behavior such as operating, using, maintaining, or interacting with other virtual objects in the virtual scene. For example, operating the control panel in the virtual scene, interacting with other virtual objects, driving virtual vehicles, using daily necessities (such as pouring coffee, brushing teeth, playing music, adjusting lights), exploring the environment (such as opening treasure chests, climbing over walls), and interacting with the virtual environment (such as picking up stones, planting plants).

[0077] 10) Item Actions: These refer to actions pre-designed based on the function, purpose, and operation of virtual items. For example, when a virtual item is a virtual weapon (such as a sword), the actions could be drawing the sword, swinging the sword, and sheathing the sword; when a virtual item is a tool, the actions could be using a hammer to strike an object, turning a key, etc.; when a virtual item is a consumable item, the actions could be eating food, drinking water, etc.; when a virtual item is a magical item, such as a magic potion, the actions could be drinking (triggering a healing effect) or throwing (using it as a throwing virtual weapon).

[0078] In related technologies, the interaction methods between virtual objects and virtual props are limited and relatively simple, which cannot adapt to the increasingly diverse characteristics of virtual props in virtual scenes and cannot meet the players' needs for diverse interaction methods, resulting in a poor player experience.

[0079] Based on the above analysis, the applicant found that the interactive processing methods of virtual scenes in related technologies cannot guarantee the diversity and adaptability of actions during the interaction between virtual objects and virtual props. In response to the above problems, this application provides an interactive processing method, device, electronic device, computer-readable storage medium, and computer program product for virtual scenes, which can improve the diversity and adaptability of actions during the interaction between virtual objects and virtual props.

[0080] The following describes exemplary applications of the electronic devices provided in the embodiments of this application. These electronic devices can be implemented as various types of terminals such as laptops, tablets, desktop computers, set-top boxes, smartphones, smart speakers, smartwatches, smart TVs, and in-vehicle terminals, or as servers. Exemplary applications of the electronic devices as terminals will be described below.

[0081] Referring to Figure 1, which is a schematic diagram of the architecture of the virtual scene interactive processing system 100 provided in the embodiment of this application, in order to realize the interactive processing application supporting a virtual scene, the terminal 400 (exemplarily showing the human-computer interaction interface 411) connects to the server 200 through the network 300. The network 300 can be a wide area network or a local area network, or a combination of both.

[0082] Terminal 400 is used to display a virtual scene on human-computer interaction interface 411. When the picking conditions are met, it controls the first virtual object to pick up the first virtual prop. When a trigger operation of the first virtual object on the first virtual prop is detected, it controls the first virtual object to perform an action sequence based on the first virtual prop. The action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

[0083] Referring to Figure 2, which is a schematic diagram of the structure of a terminal 400 provided in an embodiment of this application, the terminal 400 shown in Figure 2 includes at least one processor 410, a memory 450, at least one network interface 420, and a user interface 430. The various components in the terminal 400 are coupled together via a bus system 440. It is understood that the bus system 440 is used to implement communication between these components. In addition to a data bus, the bus system 440 also includes a power bus, a control bus, and a status signal bus. However, for clarity, all buses are labeled as bus system 440 in Figure 2.

[0084] The processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0085] User interface 430 includes one or more output devices 431 that enable the presentation of media content, including one or more speakers and / or one or more visual displays. User interface 430 also includes one or more input devices 432, including user interface components that facilitate user input, such as a keyboard, mouse, microphone, touch screen display, camera, other input buttons and controls.

[0086] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state storage, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices physically located away from the processor 410.

[0087] The memory 450 may include volatile memory or non-volatile memory, or both. The non-volatile memory may be read-only memory (ROM), and the volatile memory may be random access memory (RAM). The memory 450 described in this application embodiment is intended to include any suitable type of memory.

[0088] In some embodiments, memory 450 is capable of storing data to support various operations, examples of which include programs, modules, and data structures or subsets or supersets thereof, as illustrated below.

[0089] Operating system 451 includes system programs for handling various basic system services and performing hardware-related tasks, such as the framework layer, core library layer, driver layer, etc., for implementing various basic business functions and handling hardware-based tasks;

[0090] The network communication module 452 is used to reach other electronic devices via one or more (wired or wireless) network interfaces 420, exemplary network interfaces 420 including: Bluetooth, WiFi, and Universal Serial Bus (USB), etc.

[0091] Presentation module 453 is configured to enable the presentation of information (e.g., a user interface for operating peripheral devices and displaying content and information) via one or more output devices 431 (e.g., a display screen, a speaker, etc.) associated with user interface 430;

[0092] The input processing module 454 is used to detect and translate one or more user inputs or interactions from one or more input devices 432.

[0093] In some embodiments, the apparatus provided in this application can be implemented in software. FIG2 shows an interactive processing apparatus 455 for a virtual scene stored in memory 450, which can be software in the form of programs and plug-ins, including the following software modules: display module 4551, picking module 4552, and control module 4553. These modules are logically related, and therefore can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be described below.

[0094] In some embodiments, the terminal or server can implement the interactive processing method of the virtual scene provided in this application embodiment by running various computer-executable instructions or computer programs. For example, computer-executable instructions can be microprogram-level commands, machine instructions, or software instructions. Computer programs can be native programs or software modules in an operating system; they can be native applications (APPs), i.e., programs that need to be installed in the operating system to run, such as game APPs; or they can be applets that can be embedded in any APP, i.e., programs that only need to be downloaded to a browser environment to run. In summary, the aforementioned computer-executable instructions can be any form of instruction, and the aforementioned computer programs can be any form of application, module, or plugin.

[0095] The interactive processing method for virtual scenes provided in this application will be described by referring to the exemplary applications and implementations of the terminals provided in the embodiments of this application.

[0096] Referring to Figure 3A, which is a first flowchart of the interactive processing method for a virtual scene provided in an embodiment of this application, the steps shown in Figure 3A will be explained using the terminal as the execution subject as an example.

[0097] In step 101, a virtual scene is displayed, wherein the virtual scene includes at least one virtual object.

[0098] In some embodiments, for scenarios where virtual objects and virtual props interact in a virtual scene, the virtual objects in the virtual scene can be in a state where they have not picked up virtual props or in a state where they have picked up virtual props.

[0099] In step 102, in response to the fulfillment of the picking condition, the first virtual object is controlled to pick up the first virtual prop, wherein the first virtual object is any virtual object in the virtual scene.

[0100] In some embodiments, the operation of picking up the first virtual prop can be performed when any virtual object in the virtual scene meets the picking conditions.

[0101] Here, the picking conditions include any of the following: receiving a trigger operation for a picking control in the virtual scene; the distance between the first virtual object and the first virtual prop is less than a threshold (for example, the distance between the first virtual object and the first virtual prop is 5 meters, and the threshold is 10 meters), and the first virtual object is not currently holding the first virtual prop (at this time, the first virtual object may hold other virtual props, and when the picking conditions are met, other virtual props can be replaced with the first virtual prop); the distance between the first virtual object and the first virtual prop is less than the threshold, and the performance parameters of the currently held second virtual prop are lower than those of the first virtual prop; the distance between the first virtual object and the first virtual prop is less than the threshold, and the first virtual prop is compatible with the currently interacting second virtual object; the distance between the first virtual object and the first virtual prop is less than the threshold, and the first virtual prop is compatible with the environment in which the first virtual prop is currently located, that is, the attributes of the environment are consistent with the attributes of the first virtual prop, that is, the environment can enhance the capabilities of the first virtual prop. For example, the first virtual item is the Storm Sword. When a player wields the Storm Sword in the glacial region to attack an enemy, it will release a cold wind that slows down the enemy's movement, making it easier for the player to attack or gather teammates to surround and attack.

[0102] Environmental attributes can be static or dynamic environmental features in a virtual scene, including element types, terrain features, scene theme, enemy distribution, lighting conditions, and weather conditions. Virtual items have inherent functions or usage characteristics, including element type, applicable terrain, style theme, target type, usage conditions, and special effects / buffs.

[0103] In some embodiments, the attributes of the environment are consistent with the attributes of the first virtual item, which can be the same elemental attribute, that is, the environment and the item have the same "elemental type". The elemental characteristics of the environment (such as natural / magical elements such as ice, fire, lightning, wind, etc.) are consistent with the elemental effects of the item to achieve the "elemental buff" effect.

[0104] Example environment: "Northern Icefield" (Environmental attributes: Ice, scene covered with snow and icicles, ground frozen);

[0105] Example item: "Frost Sword" (Item attribute: Ice type, the blade emits cold air, the skill is "Ice Cone");

[0106] Adaptation Effects: Item Skill Enhancement: The range of "Ice Cone" is increased from 5 meters to 8 meters (the low temperature environment of the Ice Plains enhances the diffusion of ice element); Environmental Interaction: When swinging the sword, "Ice Marks" will be left on the ice surface (resonating with the ice element in the environment), and enemies will slip when they step on the ice marks (additional control effect); Strategic Advantage: Ice items deal double damage to "fire-attribute enemies" (such as "fire element creatures" in the Ice Plains) (elemental advantage + environmental benefit).

[0107] In some embodiments, the attributes of the environment are consistent with the attributes of the first virtual prop, which may be consistent with the terrain attributes, that is, the environment matches the "applicable terrain" of the prop.

[0108] The terrain features of the environment (such as swamps, mountains, deserts, narrow alleyways, etc.) are consistent with the usage scenarios of the props, thus solving the limitations brought about by the terrain.

[0109] Example environment: "Poisonous Fog Swamp" (Environmental attributes: muddy, many obstacles, you will get stuck in the mud when moving, vines obstruct your view);

[0110] Example of a prop: "Long-handled Thorn Scythe" (Props: Suitable for swampy terrain, the long handle prevents hands from touching the mud, and the blade can cut vines);

[0111] Adaptation Effects: Movement Optimization: When holding the long-handled sickle, the character will not sink into the mud (the long handle supports the body weight), and the movement speed remains at 80% (only 50% when not holding it); Combat Adaptation: Slashing can cut vines in front (clearing obstacles in the field of vision), and the "poison mist immunity" coating on the sickle blade can resist the poison gas in the swamp (no need to wear an additional gas mask); Scene Integration: The "vine entanglement" decoration on the sickle handle is consistent with the style of the swamp environment, and there is no visual incongruity.

[0112] In some embodiments, the attributes of the environment are consistent with the attributes of the first virtual prop, which can be a consistent scene theme, that is, the "style and tone" of the environment and the prop match. The theme style of the environment (such as a medieval castle, a futuristic sci-fi base, an ancient town, etc.) is consistent with the appearance / setting of the prop, enhancing the character's "sense of scene integration".

[0113] Example environment: "Medieval Knight's Castle" (Environment attributes: Retro cold weapon style, buildings are stone walls and towers, NPCs are knights and squires);

[0114] Example of an item: "Knight's Greatsword" (Item attributes: Retro cold weapon style, the sword is tall and heavy, and the hilt is decorated with a knight's emblem);

[0115] Adaptation Effects: NPC Interaction: Castle guards will salute characters wielding greatswords (considering them "friendly knights") and will not attack them without cause; Quest Adaptation: The "Knight's Trial" quest in the castle requires the use of a "Knight's Greatsword" to complete (the quest cannot be triggered without this item); Visual Harmony: The "rust texture" of the greatsword matches the "patina of the stone walls" of the castle, and the sound effects of actions (such as "drawing the sword" and "sheathing the sword") (the sound of metal clashing) match the "hollow echo" of the castle, enhancing the sense of immersion.

[0116] In some embodiments, the attributes of the environment are consistent with the attributes of the first virtual prop, such as lighting / climate attributes: the environment matches the "use conditions" of the prop, and the dynamic conditions of the environment (such as darkness, heavy rain, strong wind) are consistent with the triggering conditions of the prop, thus solving the "operational restrictions" caused by the environment.

[0117] Example environment: "Midnight Forest" (Environmental attributes: insufficient light, weak moonlight, and a line of sight of only 3 meters);

[0118] Example of an item: "Night-Shining Elf Bow" (Item attributes: Trigger condition is "dark environment", the bow is inlaid with night-shining crystals, and the arrows have a fluorescent trajectory);

[0119] Adaptation effects: Vision optimization: The bow's glow can illuminate a range of 5 meters in front (compensating for the lack of vision in the dark); Aiming assistance: The glowing trajectory of the arrow can track the enemy in real time (even if the enemy moves, the aiming point can be adjusted through the trajectory); Concealment advantage: The glow is only visible to the player (enemies cannot detect it), avoiding the exposure of position (unlike the "torch" item which will attract enemies).

[0120] In this embodiment, the design of "environmental attributes and item attributes being consistent" allows the item's function to complement the environmental characteristics, achieving three core effects: Enhanced practicality: Items overcome environmental limitations (such as the slipperiness of the ice field or the darkness of the forest), becoming "scene-specific tools"; Enhanced strategic depth: Players need to select items based on the environment (such as choosing the Holy Silver Sword in the Tomb of the Dead), rather than using "universal items for everything"; Enhanced immersion: The visual / functional linkage between items and the environment (such as the ice sword leaving ice marks on the ice field) makes players feel that "items truly belong to this scene".

[0121] Here, the performance parameters of the virtual item can be damage value, attack range, and durability, etc. The distance between the first virtual object and the first virtual item is less than a threshold, and the first virtual item is compatible with the currently interacting second virtual object, which can be any of the following: 1) Item compatibility, that is, the first virtual item is compatible with the item held by the second virtual object, for example, spear with shield, gun with knife; 2) Distance compatibility between the first virtual item and both (first virtual object and second virtual object), that is, the radius of the attack range of the first virtual item exceeds the distance between the first virtual object and the second virtual object; 3) Skill compatibility between the first virtual item and the second virtual object, for example, if the skill of the second virtual object is freeze control, then the first virtual item can be a flame spear.

[0122] In some embodiments, the picking condition may also be receiving a trigger operation of a shortcut key in a terminal device (e.g., a mouse, keyboard, etc.) that controls the first virtual object; or it may be receiving a voice command or a motion-sensing operation command to trigger the first virtual object to pick up the first virtual item.

[0123] This application provides a scheme for players to actively pick up items, and for automatically triggering the pickup of virtual items when certain conditions are met. Players can intuitively pick up virtual items through triggering operations, increasing their sense of control and immersion. When specific conditions are met, virtual items can be automatically picked up without additional player intervention, simplifying the game process, reducing tedious operations for players, and improving the user experience. Specifically, the distance between the virtual object and the virtual item is less than a threshold ensures that pickup only occurs when the player character is close to the virtual item. When the performance of the currently held virtual item is lower than that of the available virtual item, an automatic pickup operation can be triggered, improving the game's playability and strategic depth. Adaptation to the currently interacting virtual object ensures that the picked-up virtual item is suitable for the virtual object the player is currently interacting with. For example, if facing multiple enemies, the player may automatically pick up a virtual item with a wider attack range. Adaptation to the environment makes the picked-up virtual item more harmonious with the game environment. For example, in a concealed environment, a virtual sniper rifle can be automatically picked up, increasing the game's strategic depth and realism. Overall, the above solutions improve the game's ease of use, immersion, realism, and strategic depth, providing players with a richer gaming experience.

[0124] In some embodiments, step 102, “controlling the first virtual object to pick up the first virtual item”, can be implemented by performing the following process: starting from the state where the first virtual object has not picked up the first virtual item, controlling the first virtual object to perform multiple intermediate actions of the picking process, and after performing the intermediate actions, controlling the first virtual object to be in the state where it has picked up the first virtual item.

[0125] As an example, refer to Figure 4, which is a first schematic diagram of the interaction process of a virtual scene provided in an embodiment of this application. As shown in Figure 4, Figure 4 shows the start action 461 when the first virtual object 471 is in a state where it has not picked up the first virtual prop 472, the end action 467 when the first virtual object 471 has picked up the first virtual prop 472, and a number of intermediate actions, including a first intermediate action 462, a second intermediate action 463, a third intermediate action 464, a fourth intermediate action 465, and a fifth intermediate action 466.

[0126] This application embodiment controls the first virtual object to perform multiple intermediate actions during the pickup process, between the state where the first virtual object is not picking up the first virtual item and the state where the first virtual object has picked up the first virtual item. This better showcases the pickup process, allowing players to experience a more natural interaction and enhancing their immersion in the game. The addition of intermediate actions makes the pickup process not just an instantaneous event, but a transitional process, increasing the richness and complexity of the interaction. Through these intermediate actions, players can more clearly understand the game mechanics, improving the intuitiveness of the user interface and thus enhancing the overall user experience. The animation effects during the pickup process increase the game's visual appeal, making it look more professional and engaging.

[0127] In some embodiments, the aforementioned intermediate actions conform to the character characteristics of the first virtual object; the preceding action of each intermediate action is represented by the skeletal data of the third bone (the pose of multiple third bone points in the third bone), and the following action of each intermediate action is represented by the skeletal data of the fourth bone (the pose of multiple fourth bone points in the fourth bone).

[0128] Here, the action preceding the first intermediate action in a series of intermediate actions is the action in which the first virtual object is in a state where it has not picked up the first virtual item, and the action following the last intermediate action in a series of intermediate actions is the action in which the first virtual object is in a state where it has picked up the first virtual item.

[0129] Referring to Figure 3B, which is a second flowchart of the virtual scene interaction processing method provided in the embodiment of this application. Before performing the above-mentioned "controlling the first virtual object to perform intermediate actions of multiple picking processes", the skeletal data of at least one intermediate action can be determined by performing steps 201 to 203 of Figure 3B, as described in detail below.

[0130] In step 201, the third bone point of the third bone is matched with the fourth bone point of the fourth bone to obtain the second matching result.

[0131] Skeleton joint matching is a technical process in virtual motion interaction scenarios that establishes a one-to-one or many-to-one correspondence between skeletal points (joints) in two or more skeletal structures by aligning their geometric features with functional semantics. Its core objective is to solve compatibility issues between different skeletal structures (such as the character's native skeleton and prop skeleton, or the skeletons of frames before and after an action sequence), ensuring the naturalness, rationality, and reusability of interactive actions (such as gripping or swinging) or transitions (such as standing → jumping) between virtual objects (such as characters) and virtual props.

[0132] Here, the second matching result represents the correspondence between each third bone point and each fourth bone point.

[0133] In some embodiments, a second matching result can be obtained by performing the following processes: First, extract the spatial coordinates of the third bone point from the data structure of the third bone, and extract the spatial coordinates of the fourth bone point from the data structure of the fourth bone. If the data sources or coordinate systems of the bones are inconsistent, coordinate transformation is required to ensure that they are in the same coordinate system. Second, determine the error between the third and fourth bone points, which may include Euclidean distance, orientation differences, etc. Then, based on the error, adjust the position, rotation, or scaling of the third bone to make the coordinates of the third bone point as close as possible to the coordinates of the fourth bone point. For example, the third bone can be moved along the X, Y, and Z axes until the coordinates of the third bone point and the coordinates of the fourth bone point are as close as possible in space, until they are aligned; if only the position is close enough, the orientation of the bone can also be adjusted by rotating the third bone to align the orientation vector of the third bone with the orientation vector of the fourth bone point; if the sizes of the third and fourth bones are inconsistent, scaling adjustments can also be made. Finally, the quality of the matching is evaluated using a metric method (such as minimizing the sum of squared errors) to determine whether a preset threshold has been reached. If the matching result does not meet the requirements, iterative fine-tuning can be performed until the second matching result is achieved.

[0134] Skeleton Structure: A hierarchical structure composed of skeleton joints (such as the "palm" and "elbow" of a character, and the "hilt" and "blade" of an item) and skeleton chains (such as the joint connections from "shoulder → elbow → wrist → palm"). Skeleton joints are the basic units of the skeletal structure, containing pose information (Position: 3D spatial coordinates (x, y, z); Rotation: rotational state described by Euler angles / quaternions) and functional semantics (such as the "gripping" function of the "palm" and the "being gripped" function of the "hilt").

[0135] Source bone and target bone: two bone structures to be matched (e.g., the "third bone" in the previous frame of the action is the source bone, and the "fourth bone" in the next frame of the action is the target bone).

[0136] By combining geometric feature matching (spatial location, distance, direction) and functional semantic matching (functional equivalence), we find pairs of skeletal points in two skeletal structures that are "similar in location and functionally equivalent" and establish a correspondence mapping: one-to-one correspondence: such as the character's "palm" corresponding to the prop's "sword hilt"; many-to-one correspondence: such as the character's "thumb," "index finger," and "middle finger" collectively corresponding to the prop's "sword hilt" (simulating the fingers wrapping around when holding).

[0137] The result of skeleton matching is a joint correspondence table, which may include: a mapping between source skeleton ID and target skeleton ID (e.g., character "hand" ID = 10 → prop "sword hilt" ID = 5); matching accuracy indicators (e.g., spatial distance error ≤ 5cm, functional semantic similarity ≥ 90%); and adjusted skeleton pose (e.g., the coordinates of the character "hand" are adjusted from (x1, y1, z1) to (x1', y1', z1'), coinciding with the prop "sword hilt").

[0138] In some embodiments, taking a "character-prop holding scene" as an example, specifically a scene where a virtual character (source skeleton: first skeleton) holds a virtual prop (target skeleton: second skeleton), the specific steps for skeleton point matching are as follows:

[0139] First, we analyze the skeletal structure.

[0140] Extract the first skeletal structure of the character: parse the bone chain of "shoulder → elbow → wrist → palm → fingers" and obtain the pose of each bone point (e.g., the coordinates of the palm are (0.5, 0.2, 1.0), and the rotation is a quaternion (q = 0.1, w = 0.8, e = 0.1)).

[0141] Extract the second skeletal structure of the prop: Analyze the skeletal chain of "hilt → blade → tip" and obtain the pose of each skeletal point (e.g., the coordinates of the hilt are (0.55, 0.22, 1.05), and the rotation is a quaternion (q = 0.05, w = 0.85, e = 0.1)).

[0142] Secondly, feature extraction.

[0143] Two types of features are extracted for each skeletal point for matching:

[0144] Geometric features: the position of a skeletal point in the skeletal chain (e.g., the "palm" is the end of the "arm chain"), the distance to adjacent skeletal points (e.g., the distance from the "palm" to the "wrist" is 0.15m), and the spatial direction (e.g., the "palm" is facing "forward").

[0145] Functional semantic features: Identify the function of skeletal points through predefined labels or deep learning models (e.g., the label for a character's "hand" is "holding", and the label for a prop's "sword hilt" is "being held").

[0146] Next, the skeleton matching algorithm is executed.

[0147] The optimal matching pair is found using a hybrid algorithm that combines functional semantic filtering with geometric distance matching.

[0148] ① Functional semantic filtering: Filter out points in the character skeleton whose function is "to hold" (such as "palm" and "fingers") and points in the prop skeleton whose function is "to be held" (such as "sword hilt").

[0149] ② Geometric distance calculation: Calculate the spatial Euclidean distance between the bone points after filtering (e.g., the distance between the character's "hand" and the item's "sword hilt" is: [(0.5-0.55)] 2 +(0.2-0.22) 2 +(1.0-1.05) 2 ]≈0.07m);

[0150] ③ Threshold determination: If the distance is ≤0.1m (predefined “reasonable grip distance”), then establish the correspondence between “palm → hilt”.

[0151] Again, position adjustment

[0152] Based on the matching results, adjust the pose of the character's skeletal points to ensure they coincide with the prop's skeletal points:

[0153] Adjust the coordinates of the character's "hand" from (0.5, 0.2, 1.0) to (0.55, 0.22, 1.05) (to match the coordinates of the item "sword hilt");

[0154] Adjust the rotation of the character's "hand" to match the rotation of the item "sword hilt" (change the quaternion from (q=0.1, w=0.8, e=0.1) to (q=0.05, w=0.85, e=0.1)) to simulate the posture of "hand gripping the sword hilt".

[0155] As an example, bone matching algorithms include:

[0156] K-Nearest Neighbors (KNN) geometric matching: Calculate the spatial distance between the source skeleton point and the target skeleton point, select the K nearest points as candidates, and then filter them through functional semantics;

[0157] PointNet++ deep learning matching: Treat skeletal points as 3D point clouds, train the model to learn "functionally equivalent" skeletal point features (e.g., input the point cloud of a character's "palm" and the point cloud of a prop's "sword hilt", and the model outputs the "matching" probability).

[0158] Inverse Kinematics (IK) Assisted Matching: If the geometric distance slightly exceeds the threshold, the pose of the character's skeletal chain (such as "elbow → wrist → palm") is adjusted through IK to bring the "palm" closer to the "hilt".

[0159] Typical application scenarios for skeletal matching are as follows: Scenario 1: Motion fusion: When a character picks up an item, the "hand → sword hilt" skeletal points are matched to ensure a natural "hand holding the item" motion; Scenario 2: Motion transition: In the transition from "standing → jumping and swinging the sword", the "hand" skeletal points of the preceding and following frames are matched, and a smooth transition from "bending the knee → jumping → swinging the sword" is generated through interpolation; Scenario 3: Multiple characters sharing items: By matching the "hand" skeletal points of different characters with the "sword hilt" skeletal points of the same item, "one-click adaptation of item actions" is achieved, eliminating the need to create animations for each character repeatedly.

[0160] In this embodiment, the following technical effects are achieved through skeletal matching: Naturalness: By precisely aligning skeletal points, unreasonable appearances such as "hands through molds" and "props floating" are avoided;

[0161] Compatibility: Supports matching different characters (such as "fat characters" and "thin characters") with the same prop; Reusability: Enables cross-character redirection of prop actions (such as adapting the "samurai swinging a sword" action to the "assassin's" skeletal structure), reducing animation production costs; Real-time: Geometric feature-based matching can be completed within 5ms, meeting the real-time interaction requirements of games.

[0162] In step 202, based on the second matching result, the bone data of the fourth bone is mapped onto the third bone.

[0163] In some embodiments, the bone data of each fourth bone point in the fourth bone can be mapped to the third bone point corresponding to the fourth bone point in the third bone, based on the second matching result.

[0164] In step 203, based on the pose of the fourth bone point of the fourth bone, the third bone point of the third bone is adjusted at least once according to the set adjustment range to obtain bone data of at least one intermediate action, wherein the bone data of at least one intermediate action is used to control the first virtual object to perform at least one intermediate action of multiple picking processes.

[0165] In some embodiments, the set adjustment range can be determined in the following way: First, determine the number of intermediate actions to be performed by the first virtual object during the picking process, and determine the total range of pose adjustment required to adjust the third bone point of the third bone to the fourth bone point of the fourth bone during the picking process from the state of not picking up the first virtual prop to the state of controlling the first virtual object to pick up the first virtual prop; then, determine the ratio of the total range to the number of intermediate actions as the adjustment range for each adjustment operation.

[0166] This application embodiment maps skeletal data based on the matching of skeletal nodes, and adjusts the positions of skeletal points in the preceding action according to the pose of the skeletal points to obtain the intermediate action. The intermediate action conforms to the character characteristics of the first virtual object, ensuring the naturalness and consistency of the screen during the interaction between the virtual object and virtual props, and improving the character's recognizability and the game's immersion. Matching the third and fourth bones ensures the consistency of the two skeletal points in spatial position and direction, providing a foundation for subsequent animation adjustments. Skeletal data mapping allows for the mutual conversion and adaptation of skeletal data from different characters, improving the reusability of animation resources and reducing animation production costs. Precise pose adjustment and amplitude control enable the creation of delicate intermediate actions that meet game design requirements, enhancing the realism of the animation and player feedback. Generating skeletal data for controlling the intermediate action of the first virtual object during the pickup process ensures that character movements are not only visually coherent but also logically interactive. In summary, the above solution improves the quality of game animation, enhances the player's gaming experience, and also increases game development efficiency and flexibility.

[0167] In some embodiments, referring to FIG3C, FIG3C is a third flowchart of the virtual scene interaction processing method provided in the embodiments of this application. After executing "controlling the first virtual object to pick up the first virtual prop" in step 102 of FIG3A, the processing of steps 204 to 206 of FIG3C can be executed, which will be described in detail below.

[0168] In step 204, in response to the fulfillment of the discard condition, the first virtual object is controlled to discard the first virtual item.

[0169] In some embodiments, the discarding conditions include any one of the following: receiving a discard trigger operation for the first virtual item; the first virtual object holding the first virtual item completing a predetermined interactive task (e.g., exchanging blows with an opponent for more than 10 moves, or defeating the opponent within 20 moves); the first virtual object holding the first virtual item for a predetermined duration; and a second virtual item existing at a distance less than a threshold from the first virtual object, and the performance parameters (such as lethality, lethal radius) of the second virtual item being greater than the performance parameters of the first virtual item.

[0170] Here, the discarding condition can also be that a second virtual item exists at a distance of less than a threshold from the first virtual object, and the performance parameters of the second virtual item are greater than those of the first virtual item. For example, if the kill radius of the first virtual item is 5 meters and the kill radius of the second virtual item is 10 meters, then the first virtual object will be automatically triggered to discard the first virtual item.

[0171] In some embodiments, controlling the second virtual object to perform the action sequence based on the first virtual prop can be achieved by adjusting the weight of the prop action according to the character characteristics (such as body size, skill type, and action style) of the second virtual object: if the second virtual object is a heavy character (such as a tank), increase the weight of "defensive stance" in the prop action; if the second virtual object is an agile character (such as an assassin), increase the weight of "rapid slash" in the prop action.

[0172] The character characteristics (body type, skill type, and movement style) of the second virtual object determine its core functional positioning (such as "damage resistance and defense" for heavy characters and "rapid output" for agile characters). The essence of adjusting the weight of item actions is to strongly bind the functional actions of items with the core functions of characters—by increasing the weight of "item actions that match the core functions of characters," the expressiveness of character positioning is enhanced; at the same time, the weight of "item actions that do not have core functions" is reduced to avoid conflicts between actions and character positioning.

[0173] As an example, core strategies for adjusting weights can include: Functional alignment: If a character's core function is "defense" (such as a heavy character), increase the weight of "defense-related actions" in item actions (such as "raising a shield" and "barrier"); Style adaptation: If a character's action style is "fast" (such as an agile character), increase the weight of "high-frequency / low-amplitude" actions in item actions (such as "rapid slash" and "thrust"); Body type compatibility: If a character's body type is "robust" (such as a heavy character), increase the weight of "large-amplitude / high-stability" actions in item actions (such as "raising a shield for defense" having a larger coverage area); If a character's body type is "slender" (such as an agile character), increase the weight of "small-amplitude / high-flexibility" actions (such as "rapid slash" having a higher frequency).

[0174] Using a medieval battlefield as a backdrop, this paper explains how character traits drive the adjustment of item action weights.

[0175] Example 1: Heavy character (Iron Wall Knight) + defensive item (Dawn Heavy Shield)

[0176] 1. Basic information about the characters and items is as follows:

[0177] Second virtual character: Ironclad Knight (Heavyweight Character)

[0178] Character characteristics: Sturdy build (2m tall, shoulder bone radius is 20% larger than normal characters); skill type is "defensive" (core skills "Hold the line" and "Group Shield Wall"); movement style is "fierce but slow" (action frame interval is 15% longer than normal characters, emphasizing "stability").

[0179] First virtual item: Dawn Shield (defensive item)

[0180] Item attributes: Weight 50kg, defense range covers the upper body of the character (related to the skeletal points "shoulder" and "chest");

[0181] Item actions: ① "Shield Defense": The character raises the shield to their chest, covering 80% of their upper body (initial weight 0.5); ② "Shield Bash": The character strikes the enemy with the shield, causing a knockback effect (initial weight 0.5).

[0182] 2. During the weight adjustment process, the weights of item actions are adjusted based on the strategy of "functional alignment + body type compatibility":

[0183] First, character feature recognition: based on the physical parameters of the skeletal points (shoulder radius > 0.15m) and the skill tag (defensive), it is automatically identified as a "heavy character".

[0184] Secondly, weight calculation: The attention encoding model was called, and the "defense features of the Iron Wall Knight" and the "action features of the Dawn Heavy Shield" were input. The results showed that the weight of "shield defense" was increased from 0.5 to 0.7 (matching the "core defense function" + "high stability of the robust body"); the weight of "shield bash" was decreased from 0.5 to 0.3 (non-core function, to avoid imbalance under "slow action").

[0185] Secondly, the range of motion has been adapted: the range of motion of the skeletal points for "shield defense" has been adjusted in sync – the rotation angle of the Iron Wall Knight's "shoulder" skeletal points has been increased from 60° to 75°, and the shield coverage area has been increased from 80% to 90% (to accommodate the defensive needs of "robust builds").

[0186] 3. The effect of adjusting the weights:

[0187] Naturalness of movement: The range of motion of "raising the shield to defend" matches the "broad shoulders" of the Iron Wall Knight, avoiding the incongruity of "small shield with big shoulders";

[0188] Functional enhancements: The success rate of defense has increased from 70% to 85% (due to the expanded coverage area), and the damage taken has decreased (in line with the "damage resistance" positioning);

[0189] Consistent style: The frame interval of the "shield defense" animation is consistent with the "slow style" of the Iron Wall Knight, improving the smoothness.

[0190] This application embodiment achieves action sequence adaptation of "second virtual object - first virtual prop" by binding the character characteristics with the prop actions: the action sequence of heavy characters emphasizes "defensive stability" (increased weight of raising shield), which is in line with the "damage resistance" positioning;

[0191] The agile character's action sequence emphasizes "attack speed" (increased weight for rapid slashes), aligning with the "harvesting" role; both action smoothness and functional expressiveness are significantly improved, validating the effectiveness of "character characteristics driving item action weight adjustment." This embodiment clearly demonstrates the feasibility of the technical solution—from "character feature recognition" to "weight calculation" to "action adaptation," each step has clear technical logic and quantitative indicators (such as weight values, frequency, and amplitude), providing reproducible verification cases for the technical details in the manual.

[0192] In step 205, in response to a pickup trigger operation for a first virtual prop in the virtual scene, the second virtual object is controlled to pick up the first virtual prop.

[0193] In some embodiments, the second virtual object may be triggered to pick up the first virtual item in response to a triggering operation of a pick-up control in a virtual scene. Alternatively, the second virtual object may be automatically triggered to pick up the first virtual item when the following automatic pick-up conditions are met: the distance between the second virtual object and the first virtual item is less than a threshold (e.g., the distance between the second virtual object and the first virtual item is 5 meters, and the threshold is 10 meters), and the second virtual object is not currently holding the first virtual item (in this case, the second virtual object may be holding other virtual items, which can be replaced with the first virtual item when the pick-up conditions are met); the distance between the second virtual object and the first virtual item is less than the threshold, and the performance parameters of the currently held second virtual item are lower than those of the first virtual item; the distance between the second virtual object and the first virtual item is less than the threshold, and the first virtual item is compatible with the currently interacting first virtual object; the distance between the second virtual object and the first virtual item is less than the threshold, and the first virtual item is compatible with the environment in which the first virtual item is currently located.

[0194] In step 206, in response to the triggering operation on the first virtual prop, the second virtual object is controlled to perform an action sequence based on the first virtual prop.

[0195] In some embodiments, in response to a triggering operation on a first virtual prop, a second virtual object is controlled to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the second virtual object.

[0196] This application embodiment allows the transfer of the first virtual prop's actions to any virtual object in the virtual scene after the first virtual object discards the first virtual prop. This mechanism provides more interactive options, allowing players to use props creatively, increasing the game's interactivity and exploratory nature. Action transfer allows players to try different gameplay each time they play, increasing the game's playability and replay value. By transferring the actions of one virtual prop to another virtual object, the consumption of game resources can be reduced, avoiding the need to create animations or behaviors separately for each virtual object. Transferring specific actions to different objects can add novel visual elements and performance effects to the game.

[0197] Please refer to Figure 3A for a continuation of step 102 above.

[0198] In step 103, in response to a trigger operation on the first virtual prop, the first virtual object is controlled to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

[0199] Here, object actions adapted to the first virtual object refer to actions that conform to the characteristics and behavior patterns of the virtual object itself. These can be native actions of the virtual object, representative actions that are only adapted to the first virtual object, and general actions that are adapted to multiple virtual objects, including the first virtual object. For example, the object actions of a virtual samurai character can be swinging a sword, blocking, and charging. Item actions adapted to the first virtual item refer to actions pre-designed based on the function, purpose, and operation method of the virtual item. For example, when the virtual item is a virtual weapon (such as a sword), the item actions can be drawing the sword, swinging the sword, and sheathing the sword; when the virtual item is a tool, the item actions can be using a hammer to strike an object, turning a key, etc.; when the virtual item is a consumable item, the item actions can be eating food, drinking water, etc.; when the virtual item is a magical item, such as a magic potion, the item actions can be drinking (triggering a healing effect) or throwing (using it as a throwing virtual weapon).

[0200] In some embodiments, the action sequence includes multiple action combinations, an action combination includes one or more actions, an action combination is used to implement a skill, and the type of the skill is adapted to the characteristics of the first virtual prop.

[0201] Here, in addition to the matching of skill type with the characteristics of the first virtual item, there can also be an overall matching standard for the characteristics of skills and the first virtual item. For example, the range of motion of the skill is negatively correlated with the appearance parameters of the first virtual item, where the appearance parameters include the weight and size of the first virtual item.

[0202] In some embodiments, the way the type of the skill is adapted to the characteristics of the first virtual item includes any of the following: when the skill type is a normal attack skill, the attack power of the normal attack skill is positively correlated with the attack parameters of the first virtual item (e.g., attack range, damage value per attack), and the normal attack skill is a skill whose damage value is below a damage threshold; when the skill type is a defensible attack skill, the attack power of the defensible attack skill is positively correlated with the defense parameters of the first virtual item (e.g., the first virtual item is an umbrella, and in response to the player character's control operation on the umbrella, it can reflect enemy flying items back; when defending...). When the parameter is the area of ​​the umbrella, the larger the area of ​​the umbrella, the stronger the attack power of the skill; when the type of skill is an undefendable attack skill, the attack power of the undefendable attack skill is positively correlated with the attack parameter of the first virtual item (for example, if the first virtual item is a hammer, the greater the weight of the hammer, the stronger the attack power of the skill); when the type of skill is a defensive skill, the evasion ability of the defensive skill is positively correlated with the protection range of the first virtual item (for example, if the first virtual item is a long stick, the player character can use the long stick to jump onto the long stick to avoid enemy attacks. The longer the long stick, the larger the protection range, and the stronger the evasion ability of the defensive skill).

[0203] Here, the attack parameters of the first virtual item can include any of the following: attack power (i.e., the direct damage value caused by the first virtual item each time), attack range, attack speed, attack effect (such as slowing, stunning, etc.). The defense parameters of the first virtual item can include any of the following: defense power (i.e., the amount of damage reduced when attacked, for example, a heavy armor shield provides 30 points of defense power and reduces damage by 30 points each time it is attacked), defense range, defense success rate, defense effect, and durability (e.g., the number of attacks it can withstand, or the total amount of damage it can withstand).

[0204] As examples, in a game scenario, basic attack skills can include: sword swing, where the virtual object performs a rapid sword slash; fireball, where a virtual mage summons a fireball and throws it forward, dealing basic damage to enemies; and arrow shot, where an archer draws their bow and shoots an arrow, dealing damage to a single target. Defensive attack skills can include: heavy attack, where the virtual object performs a high-damage attack, but the attack animation has obvious warning signs that enemies can recognize and use defensive skills to block; sword dance, where a swordsman performs multiple rapid attacks in succession, but each attack can have its damage reduced or eliminated through defensive actions; and spell barrage, where a mage-type virtual object unleashes a series of spell attacks that enemies can avoid by defending or dodging. Undefendable attack skills can include: area-of-effect skills, which deal continuous damage to a region; and instant-kill attacks, which can directly kill enemies.

[0205] In some embodiments, referring to FIG3D, FIG3D is a fourth flowchart of the interactive processing method for virtual scenes provided in the embodiments of this application. The step 103 in FIG3A, "controlling the first virtual object to perform an action sequence based on the first virtual prop", can be implemented by executing steps 1031 to 1032 in FIG3D, which will be described in detail below.

[0206] In step 1031, multiple usage mode controls for the first virtual prop are displayed, wherein one usage mode control is used to characterize a usage mode of the first virtual prop, and a usage mode includes a sequence of actions for using the first virtual prop.

[0207] Here, the usage mode of the first virtual item includes any of the following: using the first virtual item to attack, using the first virtual item to defend, and using the first virtual item to defend and counterattack.

[0208] In some embodiments, a virtual tool has multiple usage modes, each usage mode corresponds to an action sequence, and the actions in an action sequence are used to perform one or more skills, and each skill conforms to the usage mode.

[0209] As an example, refer to Figure 5, which is a second schematic diagram of the interaction process of a virtual scene provided in the embodiments of this application. The upper view of Figure 5 shows a first virtual object 511, a first virtual prop 512, a usage mode control 513, a usage mode control 514, and a usage mode control 515. The usage mode control 513 represents the usage mode of the first virtual prop 512 as an attack mode, the usage mode control 514 represents the usage mode of the first virtual prop 512 as a defense mode, and the usage mode control 515 represents the usage mode of the first virtual prop 512 as a defensive counterattack mode.

[0210] In step 1032, in response to a trigger operation on the first usage mode control, the first virtual object is controlled to perform the action sequence included in the triggered first usage mode based on the first virtual prop.

[0211] As an example, continuing to refer to Figure 5, in response to a trigger operation on the mode control 513, the action sequence including the first virtual object 511 implementing the triggered attack mode based on the first virtual prop 512 is shown in the lower bitmap of Figure 5, which shows the action sequence 516 including the first virtual object 511 implementing the triggered attack mode based on the first virtual prop 512.

[0212] This application embodiment allows for the selection of different usage modes for the same virtual item. For each usage mode, the virtual object can be controlled to perform a corresponding action sequence, providing multiple functions and usage methods for the virtual item, increasing the playability and diversity of the game. This allows players to choose the most suitable usage mode according to different game scenarios, increasing the strategic depth of the game. Different action sequences can make the virtual object exhibit different behaviors and actions, increasing the expressiveness and personalization of the character. Players can obtain a richer interactive experience and a sense of satisfaction by controlling the virtual object to perform specific action sequences.

[0213] In some embodiments, an object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item. Before executing step 103 above, "controlling the first virtual object to perform an action sequence based on the first virtual item", the following process can be performed: combining at least one object action in the object action template and at least one item action in the item action template into an action sequence.

[0214] Here, multiple object action templates associated with multiple usage modes can be pre-set for the first virtual object, and multiple prop action templates associated with multiple usage modes can be pre-set for the first virtual prop. Before executing step 103 above, "controlling the first virtual object to perform an action sequence based on the first virtual prop", the following processing is performed on the object action template and prop action template of the same usage mode: at least one object action in the object action template and at least one prop action in the prop action template are combined into an action sequence in a specific order.

[0215] In some embodiments, the specific order of at least one object action in the object action template and at least one item action in the item action template may be object action first and item action second, or it may be according to the order set by the player.

[0216] As an example, refer to Figure 6A, which is a third schematic diagram of the interaction process of a virtual scene provided in an embodiment of this application. In the human-computer interaction interface shown in Figure 6A, the upper view of Figure 6A shows five object actions in the object action template, namely object action 611, object action 612, object action 613, object action 614, and object action 615; the middle view of Figure 6A shows five prop actions in the prop action template, namely prop action 621, prop action 622, prop action 623, prop action 624, and prop action 625. Combining prop actions 621, prop actions 622, and prop actions 623 in the prop action template with object actions 614 and object actions 615 in the object action template yields the action sequence shown in the lower view of Figure 6A.

[0217] In some embodiments, the above-mentioned "combining at least one object action in the object action template and at least one prop action in the prop action template into an action sequence" can be achieved by performing the following process: when the object action template is used to implement multiple first skills (not dependent on virtual props) and the prop action template is used to implement multiple second skills (dependent on virtual props), combine (all) object actions in the object action template used to implement at least one first skill and (all) prop actions in the prop action template used to implement at least one second skill into an action sequence.

[0218] As an example, the object action template is used to perform skills 1, 2 and 3, and the prop action template is used to perform skills 4, 5 and 6. Skills 2 and 3 in the object action template and skill 4 in the prop action template can be combined into an action sequence.

[0219] Here, the action sequence can involve multiple skills, each skill corresponds to multiple actions, multiple actions are used consecutively for a single skill, or releasing a skill requires performing a specific combination of multiple actions (such as performing a specific combination of multiple actions to trigger an ultimate skill). For example, in the action sequence above, skill 2 includes actions 1 and 2, skill 3 includes actions 3, 4, and 5, and skill 4 includes actions 6 and 7.

[0220] In some embodiments, the first and second skills used for combination can be determined in any of the following ways: selected by the player controlling the first virtual object; or predicted by invoking a pre-trained machine learning model based on the features of the first and second skills in each combination to determine the skill combination with the highest probability of interaction.

[0221] In other embodiments, the above-mentioned "combining at least one object action in the object action template and at least one prop action in the prop action template into an action sequence" can be achieved by performing the following process: when the object action template is used to implement a third skill (not dependent on virtual props) and the prop action template is used to implement a fourth skill (dependent on virtual props), the at least one object action in the object action template used to implement the third skill and the at least one prop action in the prop action template used to implement the fourth skill are combined into an action sequence.

[0222] As an example, the object action template is used to perform skill A, and the object actions in the object action template include action 11, action 12 and action 13. The prop action template is used to perform skill B, and the prop actions in the prop action template include action 21, action 22 and action 23. Then, action 12 and action 13 in the object action template and action 21 in the prop action template can be combined into an action sequence.

[0223] In other words, the combination method of combining at least one object action in the object action template and at least one item action in the item action template into an action sequence can be based on skills or on actions.

[0224] This application embodiment combines at least one object action from the object action template and at least one prop action from the prop action template into an action sequence. This combination can be based on skills or actions, allowing players to choose different combinations according to their preferences and strategies to create unique action sequences. Different combinations can address different game scenarios, increasing the diversity of game strategies. Different combinations of action sequences can bring more layers of expression to the characters and interaction processes, providing players with a more flexible and creative game environment, while also enhancing the game's playability and player immersion.

[0225] In other embodiments, the types of object actions in the object action template include representative actions and general actions. Representative actions are actions that embody the character characteristics of the first virtual object, while general actions are actions that are applicable to multiple virtual objects (e.g., all virtual objects in the virtual scene).

[0226] As examples of representative actions, for virtual characters with different personality traits, for instance, a brave knight might step forward, draw his sword, and point it at the enemy, displaying fearlessness and a sense of justice; a cunning thief might stealthily move about or quickly shift from one place to another, showcasing agility and shrewdness. For virtual characters with different attributes or skills, for example, a mage might have specific incantations or hand gestures when casting spells. For virtual characters with different physical characteristics, for example, a dwarf, due to their small stature, might frequently swing their virtual weapon, demonstrating their fighting spirit and strength; an elf might have light steps and elegant gestures.

[0227] As examples of general actions, the general actions of virtual objects can be movement actions (such as walking, running or jumping), interaction actions (such as picking up or placing items, opening or closing, etc.), basic facial expressions (such as crying, laughing, etc.), and social actions (such as talking, nodding, shaking one's head, etc.).

[0228] The above-mentioned "combining at least one object action in the object action template and at least one prop action in the prop action template into an action sequence" can be achieved by performing the following process: removing at least one object action that is not a representative action from the object action template, and inserting a prop action from the prop action template that is in the same position as the removed object action at the position where the object action was removed, to obtain the action sequence.

[0229] As an example, refer to Figure 6B, which is a fourth schematic diagram of the interaction process of the virtual scene provided in the embodiments of this application. In the human-computer interaction interface shown in Figure 6B, the upper view of Figure 6B shows five object actions in the object action template, namely object action 611, object action 612, object action 613, object action 614 and object action 615. Among them, object action 611, object action 612 and object action 613 are general actions, and object action 614 and object action 615 are representative actions. The middle of Figure 6B shows five prop actions in the prop action template, namely prop action 621, prop action 622, prop action 623, prop action 624 and prop action 625. Then, object actions 611, 612, and 613 in the object action template can be removed from the prop action template, and prop actions 621, 622, and 623 in the prop action template can be inserted in the corresponding positions. At this time, the newly inserted prop actions 621, 622, and 623 in the prop action template are combined with object actions 614 and 615 in the object action template to obtain the action sequence shown in the lower part of Figure 6B.

[0230] Here, the type of object action can be determined by calling a pre-trained first machine learning model. The training process of the first machine learning model can be achieved by performing the following steps: acquiring training samples, which are multiple sample object actions of sample virtual objects, and label information identifying the type of each sample object action; calling the initialized first machine learning model to predict the type of multiple sample object actions, obtaining predicted type information; determining the loss value using a loss function based on the difference between the label information and the predicted type information, and updating the parameters of the initialized machine learning model through the backpropagation algorithm, thus obtaining the pre-trained first machine learning model.

[0231] As an example, the first machine learning model could be a Support Vector Machine (SVM), Random Forest, K-Nearest Neighbors (KNN), Multilayer Perceptron (MLP), Convolutional Neural Network (CNN), Gradient Boosting Machine (GBM), Deep Neural Network (DNN), Long Short-Term Memory (LSTM), etc. The loss function used to determine the loss value could be a Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Cross-Entropy Loss, etc.

[0232] In some embodiments, when the skeletal points of the object action and the prop action overlap (e.g., the object action of a character "crossing his arms" conflicts with the action of a prop "raising a shield"), the following rules apply: if the object action is a representative action (e.g., a character's exclusive "taunting pose"), the object action is retained, and the skeletal position of the prop action is adjusted; if the prop action is a functional action (e.g., a prop's "defense skill"), the prop action is retained, and the amplitude of the object action is adjusted.

[0233] The core principle of the conflict resolution rules is that when the skeletal points of an object's action (a character's native action, such as "crossing arms" or "taunting pose") overlap with those of an item's action (an item-specific action, such as "raising a shield" or "swinging a sword"), the core logic of the resolution rules is to balance "character recognizability" and "item functionality."

[0234] As an example of preserving object actions, priority can be given to retaining the character's "representative actions." Representative actions are the core identifying features of a character (such as an assassin's "taunting after stealth" or a knight's "victorious sword swing"), which are directly related to the player's perception of the character. If the object action is a representative action, it should still be retained even if it conflicts with an item action—by adjusting the skeletal position of the item action (such as changing the height of the shield) to avoid overlapping skeletal points.

[0235] As an example of preserving object actions, priority can be given to preserving the "functional actions" of props. Functional actions are the core value carriers of props (such as the "raise shield to defend" of a shield, or the "rapid slash" of a sword), and directly determine the effectiveness of the prop. If the prop action is a functional action, even if it conflicts with the object action, the action should still be preserved—by adjusting the amplitude of the object action (such as reducing the angle of the arms when "crossing arms") to avoid overlapping bone points.

[0236] Below, using a fantasy adventure virtual scenario as a background, we will illustrate a specific example of the "skeletal point overlap conflict resolution rule".

[0237] The target action is a representative action (assassin's "taunt after stealth" + item "raise shield").

[0238] 1. Conflict Scenario and Basic Information

[0239] Character and Target Actions: Nightblade Assassin (Agility-type character); Character Traits: Signature action "Taunt After Stealth" - While stealthy, the assassin crosses their arms in front of their chest, slightly raising their head (related to the "Arm," "Chest," and "Head" bone points), emphasizing "confidence and intimidation." Item and Item Actions: Iron Spine Shield (Defensive item); Item Action: "Shield Raise for Defense" - Raises the shield to chest level, covering the "Chest" and "Shoulder" bone points (the initial shield height overlaps with the arm position of the "Cross-Chest" action).

[0240] Conflict Description: When the assassin picks up the Iron Spine Shield while in stealth and taunting (chest-crossing action), the arm bone points of the "chest-crossing" action overlap with the shield bone points of the "shield-raising" action (arm passes through the shield), resulting in an illogical action.

[0241] 2. The conflict resolution process, based on the rule of "prioritizing the preservation of representative actions," resolves conflicts as follows:

[0242] First, action type determination: "Taunting after stealth" is identified as a "representative action" (associated with the character's unique ID) by action tags, and "raising shield to defend" is identified as a "non-core functional action" (the position can be adjusted without affecting the defensive effect).

[0243] Secondly, the prop animation was adjusted: the bone point matching algorithm was called to move the bone position of "raising the shield" from "directly in front of the chest" to "above the shoulder" (adjustment range: 0.2m along the Y axis) to avoid overlapping with the arm bone points of "crossing the chest".

[0244] Furthermore, after the adjustment, the assassin retains the "taunting after stealth" chest-crossing pose, while raising the shield above the shoulder—preserving the character's signature pose (maintaining recognizability) and achieving the defensive function of the item (the shield still covers the key areas of the upper body).

[0245] The above solution achieves the following technical effects: Character recognition: Players can still identify the "Nightblade Assassin" through the "chest-crossing taunt" action, without losing the character's characteristics due to the adjustment of the item action; Item functionality: The coverage area of ​​"shield defense" has been expanded from "chest" to "shoulder + chest", and the success rate of defense has increased due to the increased coverage area; Action naturalness: The arm and shield do not overlap, and the smoothness of the action is improved.

[0246] In summary, the representative action-first approach preserves the core identifying features of the character (such as the assassin's "taunt after stealth") and adjusts the skeletal position of item animations, maintaining character recognizability while ensuring item functionality. Conversely, the functional action-first approach preserves the core value of items (such as the Holy Shield's "shield defense") and adjusts the range of motion of the target, resolving conflicts and preventing functional malfunctions. This approach balances character uniqueness and item practicality, ensuring the core functions of items are fulfilled without disrupting player character cognition, while also maintaining the naturalness of the animations.

[0247] In other embodiments, referring to FIG3E, FIG3E is a fifth flowchart of the virtual scene interaction processing method provided in the embodiments of this application. The above-mentioned "combining at least one object action in the object action template and at least one prop action in the prop action template into an action sequence" can be achieved by executing steps 301 to 302 of FIG3E, which will be described in detail below.

[0248] In step 301, in response to a marking operation for at least one object action in the object action template, a marking result is displayed, wherein the marking result indicates whether the object action is in a retained state or not.

[0249] In some embodiments, a marking operation for at least one object action in an object action template may either treat the marked object action as retained or treat it as not retained. For example, a retained marking control (or a non-retained marking control) may be displayed in a human-computer interaction interface for each object action, and the marked object action may be treated as retained (or not retained) in response to a triggering operation for the retained marking control (or non-retained marking control).

[0250] As an example, refer to Figure 6C, which is the fifth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application. In the human-computer interaction interface shown in Figure 6C, the upper view of Figure 6C shows five object actions in the object action template, namely object action 611, object action 612, object action 613, object action 614, and object action 615. For each object action, a corresponding reserved marker control is displayed, including reserved marker control 631 for object action 611, reserved marker control 632 for object action 612, reserved marker control 633 for object action 613, reserved marker control 634 for object action 614, and reserved marker control 635 for object action 615; the middle of Figure 6B shows five prop actions in the prop action template, namely prop action 621, prop action 622, prop action 623, prop action 624, and prop action 625. In response to a trigger operation on the reserved marker control 634 and the reserved marker control 635, object actions 614 and 615 are considered to be reserved, while object actions 611, 612 and 613 are not reserved.

[0251] In step 302, in response to the composition operation, the object action in the object action template that is not retained is replaced with the prop action in the prop action template that is in the same position as the object action in the non-retained state, thus obtaining the action sequence.

[0252] As an example, continuing to refer to Figure 6C, in response to the trigger operation of the composition control 640 in the human-computer interaction interface, the object actions 611, 612 and 613 in the object action template that are not retained can be replaced with the corresponding prop actions 621, 622 and 623 in the prop action template to obtain the action sequence.

[0253] In some embodiments, referring to FIG3F, FIG3F is a sixth flowchart of the virtual scene interaction processing method provided in the embodiments of this application. The above-mentioned "combining at least one object action in the object action template and at least one prop action in the prop action template into an action sequence" can be achieved by executing steps 303 to 305 of FIG3F, which will be described in detail below.

[0254] In step 303, in response to a selection operation for at least one object action in the object action template, the at least one selected object action is controlled to be in a selected state.

[0255] As an example, refer to Figure 6D, which is a sixth schematic diagram of the interaction process of the virtual scene provided in the embodiments of this application. In the human-computer interaction interface shown in Figure 6D, the upper view of Figure 6D shows five object actions in the object action template, namely object action 611, object action 612, object action 613, object action 614, and object action 615, wherein a selection control corresponding to each object action is displayed below each object action; the middle of Figure 6B shows five prop actions in the prop action template, namely prop action 621, prop action 622, prop action 623, prop action 624, and prop action 625, wherein a selection control corresponding to each prop action is displayed below each prop action. In response to the triggering operation of the selection control corresponding to object action 614 and object action 615 in the object action template, the selected object action 614 and object action 615 are controlled to be in the selected state.

[0256] In step 304, in response to a selection operation for at least one item action in the item action template, the selected at least one item action is controlled to be in a selected state.

[0257] As an example, continuing to refer to Figure 6D, in response to the triggering operation of the selection controls corresponding to item actions 621, 622 and 623 in the item action template, the selected item actions 621, 622 and 623 are controlled to be in the selected state.

[0258] In step 305, in response to the compositing operation, the selected at least one prop action and at least one object action are composited to obtain an action sequence.

[0259] As an example, continuing to refer to Figure 6D, in response to the trigger operation of the composition control 640 in the human-computer interaction interface, the prop actions 621, 622 and 623 in the selected prop action template and the object actions 614 and 615 in the selected object action template can be composed to obtain an action sequence.

[0260] In some embodiments, at least one object action in the object action template and at least one prop action in the prop action template can be randomly combined; then, based on the features of each combination, a pre-trained machine learning model is invoked to predict the success probability of the interaction corresponding to each combination (for example, the probability that the first virtual object can defeat the enemy in the adversarial scenario by performing the combined actions; or the probability that the first virtual object can complete the set task in the collaborative scenario by performing the combined actions); finally, the combination with the highest probability is taken as the action sequence.

[0261] This application provides a variety of combination methods for combining at least one object action from an object action template and at least one prop action from an prop action template into an action sequence. Through real-time marking and selection operations, players can receive immediate feedback, increasing the game's interactivity and player immersion. Players can select and combine different object actions and prop actions according to their own style and strategy, thereby achieving a personalized gaming experience and increasing the variability and flexibility of gameplay. Players can customize action sequences, making the game experience richer and more interesting, and improving player participation and satisfaction.

[0262] In some embodiments, an object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item. Before executing step 103 above, "controlling the first virtual object to perform an action sequence based on the first virtual item", the following processing can be performed: each object action in the object action template is merged with the item action in the item action template that is in the same position as the object action to obtain an action.

[0263] Here, multiple object action templates associated with multiple usage modes can be pre-set for the first virtual object, and multiple prop action templates associated with multiple usage modes can be pre-set for the first virtual prop. Before executing step 103 above, "controlling the first virtual object to implement an action sequence based on the first virtual prop", the following processing is performed on the object action template and prop action template of the same usage mode: each object action in the object action template is merged with the prop action in the prop action template that is in the same position as the object action to obtain the action in the action sequence.

[0264] In some embodiments, the trigger operation may be a fusion operation of a fusion control in a human-computer interaction interface, triggering the fusion of object actions and prop actions; or it may be a trigger operation received from a shortcut key in a terminal device (e.g., a mouse, keyboard, etc.) that controls the first virtual object; or it may be a fusion operation of an object action and prop action received from a voice command or a motion-sensing operation command.

[0265] As an example, refer to Figure 6E, which is a seventh schematic diagram of the interaction process of a virtual scene provided in the embodiments of this application. In the human-computer interaction interface shown in Figure 6E, object action 651 and prop action 652 are displayed. In response to the trigger operation of the fusion control 654 in the human-computer interaction interface, the object action 651 and prop action 652 are fused to obtain the fused action 653.

[0266] In some embodiments, object actions are represented by first skeletal data, and prop actions are represented by second skeletal data. Referring to Figure 3G, which is a seventh flowchart illustrating the virtual scene interaction processing method provided in this application embodiment, the aforementioned "merging each object action in the object action template with the prop action in the prop action template that is in the same position as the object action to obtain an action" can be achieved by executing steps 401 to 403 of Figure 3G, as detailed below.

[0267] In step 401, the object motion features of the first skeleton data and the prop motion features of the second skeleton data are extracted.

[0268] In some embodiments, pose parameters of the first skeletal data of the object's motion and pose parameters of the second skeletal data of the prop's motion can be extracted. Taking the extraction of pose parameters of the first skeletal data of the object's motion as an example, this can be achieved as follows: First, the positions of key points of the human body are estimated from the image frames of the object's motion using a second machine learning model. Key points typically include major joints of the body, such as the head, shoulders, elbows, wrists, waist, knees, and ankles. Second, based on the estimated key point positions, the detected key points are connected into a skeletal chain. A line segment between every two key points represents a bone. The direction and length of each bone are calculated based on the position of the key points, thereby constructing a skeletal model of the object's motion. The skeletal model contains the connection relationships between bone nodes. Then, pose parameters, such as rotation angles and translation vectors, are extracted using the bone directions and lengths in the skeletal chain.

[0269] As an example, the second machine learning model could be a convolutional neural network, a recurrent neural network (RNN), a graph neural network (GNN), or the like.

[0270] In step 402, the first weight of the object's action features and the second weight of the prop's action features are determined.

[0271] In some embodiments, determining the first weight of the object action feature and the second weight of the prop action feature can be achieved by performing the following process: performing attention encoding based on the character features of the first virtual object and the prop features of the first virtual prop to obtain the first weight of the object action feature and the second weight of the prop action feature (wherein the sum of the first weight and the second weight is 1, for example, the first weight is 0.3 and the second weight is 0.7); or, displaying a weight editing control, and in response to an editing operation on the weight editing control, displaying the first weight of the object action feature and the second weight of the prop action feature.

[0272] In this embodiment, the attention encoding adopts a Multi-Head Self-Attention (MHSA) model based on the Transformer architecture. This model is a mainstream solution for handling multimodal feature interactions (character, prop, and scene features belong to different modalities). Its core advantages are: capturing complex relationships: it can simultaneously handle implicit relationships across features such as "character size - prop weight" and "prop function - scene terrain"; dynamic weight allocation: it quantifies the degree of influence between features through "attention scores" rather than relying on fixed rules (such as if-else judgments); parallel efficiency: it supports parallel computation of multiple features and adapts to the low latency requirements of game scenarios (real-time generation of action weights is required).

[0273] The basic principle of attention encoding is as follows.

[0274] Taking "a bulky tank character (character feature) + a heavy hammer item (item feature) + open terrain (scene feature)" as an example, this illustrates how attention encoding transforms input features into fusion weights for object actions and item actions:

[0275] First, feature preprocessing—from “semantic description” to “computable vector”. The input to the attention mechanism needs to be a numerical vector, so the discrete / semantic features (such as “fat body type”, “heavy weight”) must first be converted into high-dimensional vectors (i.e., “feature embedding”).

[0276] The feature classification and embedding rules are as follows:

[0277] The input features are divided into three categories (characters, props, and scenes). The sub-features under each category are converted into vectors through word embedding or numerical normalization: discrete features (such as "body type fat / thin" and "skill type: attack / defense") are converted into fixed-dimensional vectors (such as 32-dimensional vectors) using One-Hot encoding or pre-trained word vectors (such as Word2Vec).

[0278] Continuous features (such as "weight: 10kg / 20kg" and "number of enemies: 5 / 10") are mapped to the [0,1] interval using Min-Max normalization and then expanded into a high-dimensional vector.

[0279] Composite features (e.g., "action style: aggressive / lightweight") are described using multi-dimensional vectors (e.g., "aggressive" corresponds to [0.8, 0.1, 0.1], and "lightweight" corresponds to [0.1, 0.8, 0.1]).

[0280] Example feature embedding results:

[0281] Secondly, multi-head self-attention computation—quantifying the "correlation strength" between features.

[0282] The embedding vectors of the character (R), prop (P), and scene (S) are input into the multi-head self-attention layer, and each "attention head" independently calculates the contribution of the feature to the final weight (i.e., the "attention score").

[0283] The core formula for multi-head attention is as follows: For each attention head, calculate three vectors: Query, Key, and Value (obtained by a linear transformation of the embedding vector), and calculate the attention score using the following formula:

[0284] Where Q (query) represents "the feature that needs to be focused on" (such as "character size");

[0285] K (key): Represents "matchable features" (such as "item weight");

[0286] V (value): Represents the "actual value of a feature" (e.g., "being overweight requires heavy props for support");

[0287] The square root of the vector dimension is used for normalization to prevent the softmax gradient from vanishing due to an excessively large dot product.

[0288] softmax: maps scores to the [0,1] interval, representing the "relative importance" of features.

[0289] Example: The "Character Size - Item Weight" correlation calculation in Head 1 assumes that Head 1 focuses on the correlation between "Character Size" and "Item Weight":

[0290] Extract the "body size" subvector from the character embedding vector (R) as the query (Q = [0.9, 0.1, 0.0]);

[0291] Extract the "weight" subvector from the prop embedding vector (P) as the key (K = [0.9, 0.05, 0.05]);

[0292] Calculate the dot product: QK T =0.9×0.9+0.1×0.05+0.0×0.05=0.815;

[0293] Normalization:

[0294] Softmax processing: Since only the "body size-weight" relationship is considered, the score after softmax is approximately 0.9 (indicating that "fat body size" is highly correlated with "heavy weight").

[0295] Furthermore, multi-faceted results are integrated—comprehensive attention perspectives from multiple dimensions.

[0296] Each attention head outputs a local weight vector (representing the importance of each feature under that head). The results of all heads are concatenated and compressed into global attention weights (i.e., the total contribution of all features to action fusion) through a linear layer.

[0297] Example: Fusion of 3 attention heads. This assumes the use of 3 attention heads, each focusing on different combinations of features:

[0298] The results from the three heads are concatenated ([0.8,0.1,0.1,0.7,0.0,0.3,0.0,0.6,0.4]), and the input to the linear layer is compressed into global weights: prop feature weight = 0.75, character feature weight = 0.2, scene feature weight = 0.05 (total ≈ 1).

[0299] Again, action weight calculation—from “feature weight” to “action fusion ratio”.

[0300] Based on the global attention weights and the feature's "action relevance" (a predefined hyperparameter representing the degree of influence of the feature on the action), the final fusion ratio of the object's action and the prop's action is calculated.

[0301] Action relevance definition: Assign an "action relevance coefficient" (between 0 and 1) to each feature, representing the degree of influence of that feature on the "object action" or "prop action":

[0302] Example: Calculating the final action weight = global feature weight × action correlation coefficient, then normalizing: Item action total weight = (0.75 × (0.2 + 0.1 + 0.8) + 0.05 × (0.7 + 0.9 + 0.5)) = 0.75 × 1.1 + 0.05 × 2.1 = 0.825 + 0.105 = 0.93; Object action total weight = (0.2 × (0.8 + 0.9 + 0.7) + 0.05 × (0.3 + 0.1 + 0.5)) = 0.2 × 2.4 + 0.05 × 0.9 = 0.48 + 0.045 = 0.525;

[0303] After normalization (ensuring the sum is 1): Item action weight = 0.93 / (0.93+0.525)≈0.64 (approximately 0.7); Object action weight = 0.525 / (0.93+0.525)≈0.36 (approximately 0.3).

[0304] Finally, motion fusion—weight-driven final motion generation.

[0305] Based on the final weights, object actions (such as a tank's "horse stance") are merged with item actions (such as a hammer's "ground slam attack").

[0306] Prop actions account for 70%: the amplitude of the hammer hitting the ground (such as increasing the angle of the arm swing from 60° to 90°) and the force (such as enhancing the effect of ground indentation) are both amplified according to weight;

[0307] The target's actions account for 30%: retain the tank's basic "horse stance" posture (such as keeping the width of the feet apart unchanged and lowering the body's center of gravity) to avoid conflicts between the actions and the character's characteristics.

[0308] The final generated animation not only fits the character setting of "Fat Tank" (stable horse stance), but also highlights the item function of "Hammer" (powerful ground slam), while adapting to the scene requirements of "open terrain" (wide-range attack covering more enemies).

[0309] In summary, attention encoding implemented through the Transformer multi-head self-attention model has the following core advantages compared to traditional rule matching (such as if-else judgment): Dynamic adaptability: It can automatically adapt to different combinations of "character-item-scene" (e.g., "slender assassin + light sword + narrow alleyway combat" will generate completely different results with "item action weight 0.3 and object action weight 0.7"); Strong generalization ability: It supports adding new features (such as the character's "equipment level" and the item's "durability") without modifying the core logic; Good interpretability: The "weight source" can be traced back through the attention score (e.g., the "body size-weight" association of the first head contributes 70% of the item weight), which is convenient for debugging and optimization; High performance: The parallel computing characteristics of Transformer can control the generation time of action weights in a short time, meeting the real-time requirements of games.

[0310] Here, the characteristics of the first virtual object can be its size, strength, intelligence, skills, attack methods (such as melee, ranged, magic, etc.), combat style (such as defensive, offensive, healing, etc.), equipment preferences (types of virtual weapons and armor it commonly uses), character level, etc.; the characteristics of the first virtual item can be its durability, weight, type (virtual weapons, armor, consumables, materials, etc.), attack power, defense power, enhancement attributes (increasing attack power, defense power, hit rate, etc.).

[0311] In step 403, based on the first weight and the second weight, the object action is merged with the prop action in the prop action template that is in the same position as the object action to obtain the action.

[0312] As an example, if the first weight is 0.3 and the second weight is 0.7, then the object action and the prop action in the prop action template that are in the same position as the object action will be merged in a ratio of 3:7 to obtain the action.

[0313] This application embodiment can fuse object actions and item actions according to different weights to obtain actions in the action sequence. The fusion ratio of object actions and item actions can be set by the player or determined through attention encoding. Here, allowing players to set the fusion ratio provides a highly personalized gaming experience, enabling players to adjust action effects according to their own game style and strategy. During the game, players may need to explore different fusion ratios and action combinations to find the most effective strategy, which increases the strategic depth and diversity of the game. Determining the fusion ratio through attention encoding makes the action sequence more in line with the player's intentions, improves the coherence and accuracy of actions, and helps players better allocate resources and attention, such as using stronger actions or items at critical moments. In summary, this flexible setting of the fusion ratio and technical optimization brings a richer and deeper gaming experience, while also providing more possibilities for game development.

[0314] In some embodiments, when performing the above step 103 of "controlling the first virtual object to perform an action sequence based on the first virtual prop", the following processing can be performed: controlling the first virtual object to perform multiple actions included in the action sequence, wherein, for any two adjacent actions, if the change in the action between the two actions is greater than the amplitude threshold, then between the two actions, controlling the first virtual object to perform at least one transitional action.

[0315] For example, if the first action is oriented south and the first virtual object is located on the ground, and the second action is oriented north and the first virtual object is located in mid-air, then between the two actions, the first virtual object can be controlled to perform multiple transitional actions to reflect the process of the first virtual object changing from facing south to north and jumping from the ground to mid-air, so as to ensure the smoothness of the actions in the action sequence.

[0316] In some embodiments, object actions are represented by the skeletal data of a first bone (the poses of multiple first bone points in the first bone), and prop actions are represented by the skeletal data of a second bone (the poses of multiple second bone points in the second bone). Referring to Figure 3H, which is an eighth flowchart illustrating the virtual scene interaction processing method provided in this application embodiment, before performing the above-described "controlling the first virtual object to perform at least one transitional action between two actions," the skeletal data for at least one transitional action can be determined by performing steps 501 to 503 of Figure 3H, as detailed below.

[0317] In step 501, the first bone point of the first bone is matched with the second bone point of the second bone to obtain the first matching result.

[0318] Here, the first matching result represents the correspondence between each first bone point and each second bone point.

[0319] In some embodiments, a first matching result can be obtained by performing the following processes: First, extract the spatial coordinates of the first bone point from the data structure of the first bone, and extract the spatial coordinates of the second bone point from the data structure of the second bone. If the data sources or coordinate systems of the bones are inconsistent, coordinate transformation is required to ensure that they are in the same coordinate system. Second, determine the error between the first bone point and the second bone point, where the error may include Euclidean distance, orientation difference, etc. Then, based on the error, adjust the position, rotation, or scaling of the first bone to make the coordinates of the first bone point as close as possible to the coordinates of the second bone point. For example, the first bone can be moved along the X, Y, and Z axes until the coordinates of the first bone point and the coordinates of the second bone point are as close as possible in space; if only the position is close, the orientation of the bone point also needs to be considered, so the first bone is rotated to align the orientation vector of the first bone with the orientation vector of the second bone point; if the sizes of the first bone and the second bone are inconsistent, scaling adjustments may also be required. Finally, evaluate the quality of the matching using a metric method (such as minimizing the sum of squared errors) to determine whether a preset threshold has been reached. If the matching result does not meet the requirements, iterative fine-tuning can be performed until the first matching result is achieved.

[0320] In step 502, the bone data of the second bone is mapped onto the first bone based on the first matching result.

[0321] In some embodiments, the bone data of each second bone point of the second bone can be mapped to the first bone point corresponding to the second bone point of the first bone, based on the first matching result.

[0322] In step 503, based on the pose of the second bone point of the second bone, the first bone point of the first bone is adjusted at least once according to the set adjustment range to obtain bone data for at least one transitional action, wherein the bone data for at least one transitional action is used to control the first virtual object to perform at least one transitional action.

[0323] In some embodiments, the set adjustment range can be determined by: first, determining the number of transitional actions that the first virtual object will perform between two actions, and determining the total range that needs to be adjusted between the two actions to adjust the pose of each first bone point of the first bone to each second bone point of the second bone; then, determining the ratio of the total range to the number of transitional actions as the adjustment range for each adjustment operation.

[0324] This application's embodiments map skeletal data based on the matching of skeletal nodes and adjust the positions of skeletal points for the previous action according to their poses to obtain transitional actions. This ensures the naturalness and consistency of the visuals during the interaction between virtual objects and virtual props, improving character recognizability and game immersion. Matching the first and second skeletons ensures consistency in spatial position and direction, providing a foundation for subsequent animation adjustments. Skeletal data mapping allows for the conversion and adaptation of skeletal data from different characters, improving the reusability of animation resources and reducing animation production costs. Precise pose adjustment and amplitude control enable the creation of delicate intermediate actions that meet game design requirements, enhancing the realism of the animation and player feedback. Generating skeletal data for controlling the transitional actions of the first virtual object during the pickup process ensures that character actions are not only visually coherent but also logically interactive. In summary, the above solution improves the quality of game animation, enhances the player's gaming experience, and also increases game development efficiency and flexibility.

[0325] In some embodiments, at least one transitional action is generated by invoking a pre-trained machine learning model (such as LSTM or GAN) to generate a transitional action that conforms to the action logic based on the skeletal data of two adjacent actions, the character features of the first virtual object, and the prop features of the first virtual prop.

[0326] The core principle of transitional motion generation is that when the change range of two adjacent actions (such as "standing" and "jumping and swinging a sword") exceeds a threshold (such as the displacement of bone points between action frames > 20cm), a transitional motion needs to be inserted to connect the preceding and following actions to avoid "frame skipping" or "motion discontinuity".

[0327] The core logic of pre-trained machine learning models (such as LSTM for processing the temporal relationship of action sequences and GAN for generating realistic action frames) is as follows: Learning action logic: By pre-training the model with massive amounts of action data (such as the complete sequence of "standing → bending knees → jumping → swinging sword"), the model grasps the natural laws of "action flow" (such as "bending knees before jumping" and "raising the sword before swinging"); Integrating multimodal features: Inputting skeletal data of adjacent actions (skeletal point positions / rotations of frames before and after), character features (body shape, action style), and prop features (weight, function) allows the model to generate "transitional actions that adapt to the character and prop"; Ensuring action consistency: The generated transitional actions must match the skeletal points of the preceding and following actions (such as the "knee" skeletal point of "standing" → "bending knees" in the transitional action → "raising knees" in "jumping and swinging sword"), ensuring the continuity of the action flow.

[0328] Using a martial arts-themed virtual scene as a background, the effectiveness of "pre-trained model generating transitional actions" was verified:

[0329] Example: Standing → Jumping and swinging the sword (Swordsman + Qinggang Sword)

[0330] 1. Basic Scenarios and Input Features

[0331] Action sequence requirement: Generate a transitional action (Action T) between "standing" (Action A) and "jumping and swinging the sword" (Action B) to resolve the sense of disjointedness from "standing directly to jumping and swinging the sword".

[0332] Character and Traits: Swordsman (Agile Character), Character Traits: Slender build (height 1.8m, wrist bone flexibility +30%); Movement style "light and graceful" (action frame interval -20%, emphasizing "speed"); Skeletal data for Action A (standing): Knee bone point position (0,0.5,0) (Y-axis is height), rotation 0°; Hand bone point position (0.3,0.8,0), rotation 0°.

[0333] Item and characteristics: Qinggang Sword (light sword, offensive item), item characteristics: weight 1.5kg, attack speed 1.8 times / second (related bone points "hand" "sword blade"); bone data for action B (jumping and swinging the sword): knee bone point position (0,1.2,0) (jump height 1.2m), rotation 30° (knee bends when jumping); hand bone point position (0.5,1.5,0), rotation 90° (swinging the sword).

[0334] 2. Model Processing Procedure

[0335] Using an LSTM+GAN combined model to generate transitional actions:

[0336] First, input feature encoding is used to convert the following features into vectors that the model can process:

[0337] ① Skeletal data of adjacent actions: the pose (position + rotation) of 17 key skeletal points (head, shoulder, elbow, wrist, knee, ankle, etc.) of action A (standing) and action B (jumping and swinging sword), a total of 34-dimensional vectors (2-dimensional for each skeletal point: position + rotation); ② Character characteristics: slender body (1-dimensional), light movement style (1-dimensional), a total of 2-dimensional vectors; ③ Item characteristics: light sword (1-dimensional), fast attack speed (1-dimensional), a total of 2-dimensional vectors; Total input vector dimensions: 34 + 2 + 2 = 38 dimensions.

[0338] Secondly: the timing relationship of LSTM learning.

[0339] LSTM (Long Short-Term Memory) models process the temporal dependencies of action sequences: ① Input the skeletal data of action A and action B, and learn the temporal logic of "standing → jumping and swinging a sword" (such as "knee bending from 0° to 30°" and "hand rotating from 0° to 90°").

[0340] ② Output the "intermediate state of the motion flow": such as the trend of bone point changes in "knee flexion 30° → jump 60° → sword raising 45°".

[0341] Next, GAN generates realistic transition actions.

[0342] GAN (Generative Adversarial Network) models generate transitional action frames that conform to action logic based on the output of LSTM:

[0343] ① Generator: Generates skeletal data for 3 frames of transitional movements based on the intermediate states of the LSTM (e.g., Frame 1: Knee bent at 30°, palm rotated at 45°; Frame 2: Knee raised at 60°, palm rotated at 60°; Frame 3: Knee raised at 90°, palm rotated at 80°); ② Discriminator: Compares the generated transitional movements with the real movement sequence (e.g., real data of "standing → knee bent → jump → sword swing") to optimize the generator's output; ③ Final output: 3 frames of transitional movements (Frame 1: knee bent in preparation; Frame 2: jump and rise; Frame 3: sword raised and ready to swing), connecting movement A and movement B.

[0344] The above solution achieves the following technical effects: Temporal logic: LSTM learns the skeletal point change rules of the preceding and following actions, ensuring the "reasonableness" of the transition actions; Realism: The transition action frames generated by GAN are highly similar to real actions, avoiding the "mechanical feeling"; Adaptability: By integrating the characteristics of the character and props, the generated transition actions conform to the positioning of "agile swordsman + light sword" (light and fast).

[0345] In some embodiments, referring to FIG3I, FIG3I is a ninth flowchart of the virtual scene interaction processing method provided in the embodiments of this application. Before executing step 103, "controlling the first virtual object to perform an action sequence based on the first virtual prop", the processing of steps 601 to 602 of FIG3I can be executed, which will be described in detail below.

[0346] In step 601, an action configuration control for the first virtual prop is displayed. The action configuration control includes multiple actions included in the action sequence, and the actions are any one of prop actions and object actions.

[0347] As an example, refer to Figure 6F, which is the eighth schematic diagram of the interaction process of the virtual scene provided in the embodiment of this application. In the human-computer interaction interface shown in Figure 6F, the upper view of Figure 6F shows five actions in the action sequence, namely action 661, action 662, action 663, action 664 and action 665. At the same time, below each action, corresponding configuration controls are displayed, namely configuration control 671, configuration control 672, configuration control 673, configuration control 674 and configuration control 675.

[0348] In step 602, in response to a configuration operation for any action, the updated action sequence is displayed, wherein the updated action sequence replaces the previous action sequence.

[0349] In some embodiments, configuration operations for any action can be performed by the player in the human-computer interaction interface or by maintenance personnel in the game system's backend interface. Configuration operations can be implemented in the following ways: displaying the skeletal data for each action, and in response to a data modification operation, displaying the action corresponding to the modified skeletal data, i.e., visually modifying the action. Alternatively, it can display candidate actions for each action, and in response to a selection operation for a candidate action, replacing the original action in the action sequence with the selected candidate action; and in response to a save operation, displaying the updated action sequence.

[0350] As an example, referring to Figure 6F, in response to a trigger operation on configuration control 674, the human-computer interaction interface shown in the middle of Figure 6F is displayed. The human-computer interaction interface in the middle of Figure 6F shows four candidate actions from the candidate action template: candidate action 681, candidate action 682, candidate action 683, and candidate action 684, as well as a save control 695. Simultaneously, below each candidate action, corresponding confirmation controls are displayed: confirmation controls 691, 692, 693, and 694. In response to a selection operation on confirmation control 692, candidate action 682 corresponding to confirmation control 692 is used as the replacement action for action 664 in the original action sequence. In response to a trigger operation on save control 695, the updated action sequence is displayed in the lower-level image of Figure 6F.

[0351] This application supports configuring actions in an action sequence to update the action sequence, allowing players to configure action sequences according to their preferences and game strategies, creating a unique gaming experience. Players can choose the most suitable action combination to deal with different game situations, which can not only provide players with a more flexible and interesting gaming experience, but also improve the scalability and adaptability of game design, while bringing new possibilities to fields such as education, training and media production.

[0352] In summary, in this embodiment of the application, when the picking conditions are met, the first virtual object in the virtual scene can be controlled to pick up the first virtual item. When the first virtual object picks up the first virtual item, the first virtual object can be triggered to perform an action sequence based on the first virtual item. This allows the action sequence performed during the interaction between the virtual object and the virtual item to retain both the object action of the virtual object and the item action of the virtual item, thereby improving the diversity and adaptability of actions during the interaction between the virtual object and the virtual item, and thus enhancing the player's experience.

[0353] In some embodiments, when displaying the action configuration control for the first virtual prop, an action preview window may also be displayed to show the action effect corresponding to the configuration operation in real time; in response to the preview confirmation operation, the updated action sequence is saved.

[0354] When users adjust action parameters (such as speed, amplitude, and effects) through action configuration controls (such as sliders and drop-down menus), a real-time preview window provides feedback on the adjustment effect. Its core logic is "What You See Is What You Get," addressing the pain point of "blindly modifying": Real-time feedback: The user's configuration operations (such as dragging the "speed slider" from 1× to 2×) are mapped to action effects in real time (such as doubling the sword swing speed), which are then visually displayed through the preview window; Logic verification: The preview window verifies whether the adjusted action conforms to the logic of "character characteristics + item characteristics" (such as whether a "lightweight" sword swing speed remains after being adjusted to 2× for an "agile character"); Confirmation and saving: After the user confirms through the preview that the adjustment effect meets expectations, the updated action sequence is saved, avoiding invalid modifications.

[0355] Using a martial arts-themed virtual scene as a background, this section explains the speed configuration of a swordsman's sword-wielding actions.

[0356] 1. Basic Scenarios and Objects

[0357] First Virtual Object: Swordsman (Agile Character): Character Characteristics: Movement Style "Light and Graceful" (Default sword swing speed 1×, action frame interval 0.1s); First Virtual Item: Qinggang Sword (Light Sword, default sword swing action "Quick Horizontal Slash", associated bone points "Wrist" and "Sword Body"); Configuration Requirements: The user wants to adjust the speed of "Quick Horizontal Slash" from 1× to 1.5× to enhance the "Agile" style.

[0358] 2. User Interface and Controls

[0359] Action configuration controls: ① "Speed ​​adjustment slider": range 0.5×~2× (0.5× is slow motion, 2× is fast motion), default value 1×; ② "Amplitude adjustment slider": range 50%~150% (adjusts the arm opening angle of the sword swing); ③ "Effect selection drop-down menu": options include "No effect", "Flame trajectory" and "Frost trajectory".

[0360] Real-time preview window: Located to the right of the configuration controls, it synchronously displays the adjusted action effects—the window displays the 3D model of the swordsman and plays the action frames of "rapid horizontal slash" in real time.

[0361] 3. Configuration and Preview Process

[0362] First, parameter adjustment

[0363] In response to the user dragging the "speed adjustment slider" from 1× to 1.5×, the action frame interval is calculated in real time: the original interval is 0.1s → the adjusted interval is 0.066s (speed increased by 50%).

[0364] Secondly, real-time preview, with the preview window updating the action effects synchronously:

[0365] ① The rotation frequency of the swordsman's "wrist" bone points has been increased from 10 times / second to 15 times / second (corresponding to a sword swing speed of 1.5×);

[0366] ② The action frame changed from "swinging the sword in 10 frames" to "swinging the sword in 7 frames", but the "light" style is still maintained (the angle of the arm opening remains unchanged, which is consistent with the character's characteristics);

[0367] ③ The preview window displays a message at the bottom: "Current speed: 1.5×", which is convenient for users to confirm.

[0368] Secondly, logical verification automatically verifies whether the adjusted actions conform to logic:

[0369] ① A speed of 1.5× does not exceed the maximum speed threshold (2×) for "Agile Characters" and will not cause "motion blur";

[0370] ② The displacement of the skeletal points during the sword swing (the wrist rotates from 0° to 90°) is still within the reasonable range for "light sword" (the sword will not "pass through the body" due to excessive speed).

[0371] Finally, confirm and save.

[0372] In response to the user clicking the "Preview Confirm" button, the adjusted action sequence (speed 1.5×) is saved and replaced with the default sword-swinging action of the swordsman.

[0373] Through the above interactive design of "action configuration control + real-time preview window", the following technical effects are achieved: Efficiency: Real-time feedback allows users to quickly complete action adjustments and avoids ineffective trial and error; Reasonableness: The system automatically verifies whether the adjusted action conforms to the logic of "character + prop" to ensure that the action is natural; Personalization: Users can customize the speed, amplitude and special effects of the action to meet the needs of different game styles.

[0374] The following will describe an exemplary application of the embodiments of this application in a real game scenario.

[0375] The interactive processing method for virtual scenes provided in this application can be applied to any game scene that emphasizes attack experience, operation, and real-time feedback, and is characterized by action, such as action games (ACT) and action role-playing games (ARPG).

[0376] In games of the aforementioned types, player combat is mostly tied to a default initial state. Combat styles typically don't change as game content expands, and player character action templates are fixed. Over time, this leads to a loss of replayability and enjoyment for the player character. Furthermore, related technologies prevent players from experiencing different combat styles by picking up virtual weapons (equivalent to virtual items mentioned above). For example, in current games, picking up virtual weapons usually responds to the player's action at the weapon's location; the weapon automatically appears in the player's hand. This process is simplistic and lacks logical consistency. After picking up a virtual weapon, the player's original action template is directly replaced with another, resulting in a significant difference in action style between before and after weapon pickup, with almost no correlation. Moreover, the mechanics for different virtual weapons lack diversity, making the differences in attributes and functions insufficiently apparent.

[0377] The virtual scene interaction processing method provided in this application embodiment can enhance the diversity and smoothness of the interaction between the player character controlled by the player and the virtual weapon while maintaining the player's original distinctive combat style.

[0378] First, when a player character controlled by a terminal device approaches a collectible virtual weapon, an interactive "Pick Up" control is displayed on the terminal device's human-computer interaction interface for the player character to pick up the virtual weapon. In response to the player clicking the "Pick Up" control, the player character can use the picked-up virtual weapon to engage in combat. Simultaneously, multiple intermediate actions are inserted between the actions of not picking up a virtual weapon and picking it up to ensure smooth action flow. For example, in the current image frame where the virtual weapon is not picked up, an "action fusion transition" technique is used; that is, the actions of not picking up a virtual weapon and picking it up are merged. After the fusion transition is complete, the action switches to the state where the virtual weapon is picked up.

[0379] For example, to make the blending of two actions smoother, an animation of approximately 20 frames per second can be used to represent the blending process. After the blending transition is complete, the action of not picking up the short stick switches to the action of picking up the short stick. Using this method, the transition between the two actions is smoother, making the interaction between the player-controlled character and the virtual weapon during the pickup process more free and fluid, consistent with real-world logic, and enhancing the player's immersion in the game.

[0380] As an example, referring to Figures 7A to 7C, Figure 7A is a first schematic diagram of the picking process provided in an embodiment of this application. The human-computer interaction interface in Figure 7A displays a first virtual object 701, a first virtual prop 702, and a pickable marker 703. At this time, the first virtual object 701 is in a state where it has not picked up the first virtual prop 702. In the human-computer interaction interface on the left side of Figure 7A, the first virtual object 701 does not perform any action to pick up the first virtual prop 702. In the human-computer interaction interface on the right side of Figure 7A, in response to the operation of controlling the first virtual object 701 to raise its right leg, the human-computer interaction interface on the left side of Figure 7B is displayed. Figure 7B is a second schematic diagram of the picking process provided in an embodiment of this application. In the human-computer interaction interface on the left side of Figure 7B, one end of the first virtual prop 702 leaves the ground. Then, the human-computer interaction interface in the middle of Figure 7B is displayed. At this time, both ends of the first virtual prop 702 leave the ground, and the height of the first virtual prop 702 rises to the shoulder position of the first virtual object 701. Then, the human-computer interaction interface on the right side of Figure 7B is displayed. At this time, the first virtual object 701 is controlled to raise its right arm, and the arm moves closer to the position of the first virtual prop 701. Finally, the human-computer interaction interface shown in Figure 7C is displayed. Figure 7C is a third schematic diagram of the picking process provided in an embodiment of this application. In the human-computer interaction interface on the left side of Figure 7C, the first virtual object 701 holds the first virtual prop 701 in its hand, and the hand is still at the shoulder height. Then, the human-computer interaction interface on the right side of Figure 7C is displayed. At this time, the hand of the first virtual object 701 is at the waist height.

[0381] Secondly, the interactive processing method for virtual scenes provided in this application embodiment sets different mechanisms for different virtual weapons, so that different virtual weapons correspond to completely different combat styles, which can bring players a richer combat experience and allow players to have different ways to deal with the same enemy. For example, when facing a group of enemies, players can choose to use a short stick to quickly defeat a single enemy, or use a hammer to attack and eliminate enemies in a wide area; they can also use the characteristics of a long stick to jump onto the stick, dodge enemy attack skills, and then launch a follow-up attack on the enemy from the stick; or they can use umbrella-type virtual weapons for defense to reflect projectiles thrown by enemies back.

[0382] Here, to enrich the gaming experience, different mechanics are designed for different virtual weapons. For example, unique, varied skills are designed for different virtual weapons, such as basic attacks (equivalent to defensible attack skills mentioned above) and undefensible attacks (equivalent to undefensible attack skills mentioned above). Furthermore, to highlight the characteristics of virtual weapons, the proportion of different skills can be allocated to make the experience of using virtual weapons more aligned with their performance. For example, to emphasize the powerful and unstoppable nature of hammer-type virtual weapons, all skills of hammer-type virtual weapons can be set as undefensible attacks.

[0383] As an example, refer to Figure 8A, which is a schematic diagram of a defensible attack provided by an embodiment of this application. In the human-computer interaction interface on the left side of Figure 8A, a first virtual object 801, a first virtual item 802, and a "normal attack" effect 803 are displayed, wherein the first virtual item 802 is a short stick. In response to controlling the first virtual object 801 to release the "normal attack" skill, the human-computer interaction interface on the right side of Figure 8A is displayed, showing that the virtual object 804 is defeated. Refer to Figure 8B, which is a schematic diagram of an undefensible attack provided by an embodiment of this application. In the human-computer interaction interface on the left side of Figure 8B, a first virtual object 801, a first virtual item 812, and a "dangerous attack" effect 813 are displayed, wherein the first virtual item 812 is a hammer. In response to the control of the first virtual object 801 to release the "risky move" skill, the human-computer interaction interface shown on the right side of Figure 8B is displayed. As shown in the human-computer interaction interface on the right side of Figure 8B, the virtual object 804 is knocked down, and a dent 814 is displayed on the ground.

[0384] Meanwhile, the virtual scene interaction processing method provided in this application embodiment can utilize a more flexible and adaptable "redirection" technology to migrate the actions in the source action template of a virtual weapon to the action template of a specified player character, replacing some of the original functional actions of the player character. This allows the player character controlled by the player to use both the replaced virtual weapon skills and the original skills that were not replaced. This retains the signature actions that express the player character's personality without severing ties with the original character style, while also incorporating the unique mechanics and technical characteristics of the virtual weapon's actions. This allows the player character to experience a completely new combat experience when wielding a virtual weapon, making the player's gaming experience both unique and unified. Here, redirection refers to the technology in game development of migrating the actions in the source action template of a virtual item (equivalent to the item action template mentioned above) to a target virtual object. This combines the characteristics of the action templates of virtual items and virtual objects, thereby increasing the game's playability and diversity.

[0385] The following, with reference to the illustrations, details the implementation process of the virtual scene interaction processing method provided in the embodiments of this application.

[0386] When there are virtual weapons that can be picked up in the virtual scene, they will be highlighted in the human-computer interaction interface. At the same time, an interactive icon will be displayed above the virtual weapon to prompt the player to interact with the currently displayed virtual weapon.

[0387] As an example, referring to Figure 9A, Figure 9A is a first schematic diagram of a virtual scene provided in an embodiment of this application. Figure 9A shows a first virtual object 901, a first virtual prop 902 (i.e., a short stick), a pickable marker 903, a pickable range 904, and a skill combination, wherein the skill combination includes a standard move skill control 905 and a risky move skill control 906. In response to the first virtual object 901 moving into the pickable range 904, the human-computer interaction interface shown in Figure 9B is displayed. Figure 9B is a second schematic diagram of a virtual scene provided in an embodiment of this application. As shown in Figure 9B, a pick-up control 907 is displayed in the human-computer interaction interface. In response to the first virtual object 901 triggering the pick-up control 907, the first virtual object 901 is controlled to pick up the first virtual prop 902, and the human-computer interaction interface shown in Figure 9C is displayed. Figure 9C is a third schematic diagram of the virtual scene provided in the embodiment of this application. As shown in the upper view of Figure 9C, the first virtual object 901 is in a state of picking up the first virtual prop 902. At this time, the pickable identifier 903, pickable range 904, and pick-up control 907 are canceled. The risky move skill control 906 in Figure 9B is updated to risky move skill control 908, which can be used to perform the operation of throwing virtual weapons at enemies. In response to the player controlling the first virtual object 901 triggering the risky move skill control 908, the player can control the first virtual object 901 to perform risky move skills based on the short stick of the first virtual prop 902. As shown in the human-computer interaction interface in the middle of Figure 9C, the action of the first virtual object 901 releasing the risky move skill to the virtual object 909 is displayed. At the same time, the risky move skill effect 910 is displayed in the human-computer interaction interface. Then, as shown in the lower view of Figure 9C, the first virtual object 901 throws the first virtual prop 902 to attack the virtual object 909.

[0388] The mechanisms of different types of virtual weapons will be explained below.

[0389] For example, in related technologies, short stick-type virtual weapons are simple and crude, lacking design specific to the movements and performance of short sticks. The embodiments of this application focus on short stick-type virtual weapons in games, with the main design features being flexibility, lightness, and dense attack points. They allow for the use of normal attacks to connect with skill attacks; for example, during a dash attack, a normal attack can be seamlessly integrated to achieve a rapid and swift assault on the enemy. Simultaneously, the player character's original grab and throw skills will be replaced with short stick-specific grab and throw skills, and the action combinations performed during the entire grab and throw interaction will better suit the characteristics of short stick movements.

[0390] As an example, refer to Figure 10A, which is a fourth schematic diagram of a virtual scene provided in the embodiments of this application. Figure 10A shows three attack actions performed by the first virtual object 1001 based on the first virtual prop 1002 (i.e., a short stick) within a preset time period (e.g., 3 seconds). The left side of Figure 10A shows the attack action performed by the first virtual object 1001 based on the first virtual prop 1002 in a jumping state. The middle side of Figure 10A shows the attack action performed by the first virtual object 1001 based on the first virtual prop 1002 in a rushing state. The right side of Figure 10A shows the attack action performed by the first virtual object 1001 based on the first virtual prop 1002 in a standing state.

[0391] Referring again to Figure 10B, which is the fifth schematic diagram of the virtual scene provided in the embodiment of this application, Figure 10B shows four attack actions launched by the first virtual object 1001 against the virtual object 1003 within a preset time period (e.g., 5 seconds) based on the first virtual prop 1002 (i.e., short stick). In the human-computer interaction interface in the upper left, lower left, upper right and lower right corners of Figure 10B, the attack actions of the first virtual object 1001 using the short stick to strike different parts of the virtual object 1003 are shown respectively.

[0392] In games, related technologies often design hammer-type virtual weapons as long and cumbersome to emphasize their weight. This application, however, does not prioritize bulkiness in its design. Instead, it emphasizes the weapon's solidity and power by designing its attack skills as "risky moves." When using a hammer to attack, the player character gains a state of invulnerability, making them less vulnerable to enemy attacks. Furthermore, the player can use these hammer skills to interrupt enemy attacks, gaining an advantage.

[0393] As an example, referring to Figures 11A to 11C, Figure 11A is a sixth schematic diagram of a virtual scene provided in an embodiment of this application. The human-computer interaction interface in Figure 11A shows a first virtual object 1101, a first virtual prop 1102 (i.e., a hammer), and a risky move skill control 1103. The risky move skill control 1103 is associated with the spinning attack skill of the first virtual prop 1102. In response to a click operation on the risky move skill control 1103, the first virtual object 1101 is triggered to perform a spinning attack using the hammer, displaying the interface shown in Figure 11B. Figure 11B is a seventh schematic diagram of a virtual scene provided in an embodiment of this application. The human-computer interaction interface in Figure 11B displays four attack actions of the first virtual object 1101 performing risky moves. From left to right, the diagrams show the four stages of the four attack actions, with the risky move skill effect 1104 displayed in the first stage diagram of Figure 11B. After completing the four attack actions of the risky move skill, the player can click the risky move skill control 1103 again according to the current environment of the virtual scene to quickly execute the finishing attack skill, displaying the interface shown in Figure 11C. This allows the player to have a smooth and flexible feel and experience even when operating heavy virtual weapons. Figure 11C is the eighth schematic diagram of the virtual scene provided in this application embodiment. It shows the action diagrams of four stages in sequence from left to right. In the first stage of the schematic diagram in Figure 11C, the risky move skill effect 1104 is displayed. In the second stage of the schematic diagram in Figure 11C, the first virtual object 1101 raises the hammer high. In the third stage of the schematic diagram in Figure 11C, the first virtual object 1101 holds the hammer and leaps up. In the fourth stage of the schematic diagram in Figure 11C, the first virtual object 1101 smashes the hammer to the ground.

[0394] For virtual weapons like staffs in games, this application's embodiments design a dodging mechanism. When the player character is on the ground, the skills usable with the staff are all "horizontal attacks." When the player character uses the staff's dodging mechanism, in response to the player character's side-stepping operation, such as responding to the player's triggering of the "dodge" control in the human-computer interaction interface, the player character can jump onto the staff while pushing off, and then launch subsequent attacks on the enemy from the staff. Here, the attack skills launched by the player character while on the staff are all wide-range "risky moves." In other words, the staff's dodging mechanism is a high-risk, high-reward approach for the player character; that is, as long as the player character can successfully jump onto the staff using the "dodge" mechanism, they can subsequently gain greater benefits through "risky moves," thereby quickly defeating the enemy.

[0395] Regarding the umbrella-type virtual weapon in the game, the feature of this application's embodiment lies in its defensive mechanism. When the player character is attacked by a flying object, the umbrella's defensive mechanism will control the umbrella to switch from a closed state to an open state, covering the area in front of the player character that was attacked by the flying object and reflecting the flying object back. Simultaneously, if the player is hit by an enemy's melee attack while the umbrella is in "defense" mode, they can use the umbrella to perform a wide-range "risky move" attack, instantly reversing the offensive and defensive rhythm and turning defense into offense.

[0396] The following describes the process of controlling a player character to pick up virtual weapons according to an embodiment of this application. Referring to Figure 12, which is a flowchart illustrating the picking-up process provided in an embodiment of this application, the process will be explained in detail using the terminal device controlling the player character as an example, in conjunction with steps 1201 to 1208 shown in Figure 12.

[0397] In step 1201, the player character in the virtual scene is displayed.

[0398] In some embodiments, the player character in the virtual scene may be displayed in the human-computer interaction interface of the terminal device controlled by the player.

[0399] In step 1202, it is determined whether there are any pickable virtual weapons in the current virtual scene. If the determination result is yes, then the processing in step 1203 is executed; if the determination result is no, then the processing in step 1202 continues.

[0400] In step 1203, it is determined whether the player character is within the pickup range of the virtual weapon. If the result is yes, then the following step 1204 is executed; if the result is no, then the process returns to step 1202.

[0401] In step 1204, it is determined whether the current player character is in a pick-upable state. If the result is yes, the following step 1204 is executed; if the result is no, the process returns to step 1202.

[0402] In step 1205, in response to receiving a pick-up operation for the virtual weapon, the player character is controlled to pick up the virtual weapon and the virtual weapon is controlled to release an attack skill.

[0403] In step 1206, during the player character's attack using the virtual weapon, it can be determined whether the virtual weapon's durability is greater than 0. If the result is yes, the battle continues; if the result is no, the virtual weapon is considered damaged. Here, the virtual weapon's durability gradually decreases with the accumulation of usage time or number of uses.

[0404] In step 1207, the player character is controlled to throw a virtual weapon.

[0405] In some embodiments, in response to a player's triggering of a throwing skill control, the player character can be controlled to throw a virtual weapon at a location. During the process of controlling the player character to perform the throwing skill, the following steps 1208 can be executed.

[0406] In step 1208, it is determined whether the durability value of the virtual weapon is greater than 0. If the result is yes, the battle can continue; if the result is no, the virtual weapon is considered to be damaged, and the process in step 1203 above will continue.

[0407] This application utilizes a more flexible and adaptable redirection technique to transfer the source motion of a virtual weapon to a designated virtual object, enabling the virtual object to perform the same actions. Furthermore, inverse kinematics (IK) is applied to key limb positions in the source motion template to ensure that player characters of different body types can perfectly match the virtual weapon's movements. For example, when a regular virtual object holds a hammer, the hammer is positioned in front of the abdomen at the torso level. If the hammer's source motion is transferred to a character with a large belly, the IK method can be used to adjust the arm position in the source motion template, making the distance between the arm holding the hammer and the body greater than in a regular character, ensuring that the hammer does not overlap with the large belly. Simultaneously, for moving objects such as the virtual object's hair and clothing, physics can be used to ensure that they do not affect the combat performance of the virtual object holding the virtual weapon. For example, when the virtual object's hair, clothes, and accessories sway with the virtual object's movements, these moving objects can be blocked from intersecting with the virtual weapon, preventing unreasonable intermingling of moving objects and virtual weapons. The redirection technology provided in the embodiments of this application will be described in detail below.

[0408] First, skeletal matching is performed to ensure the skeletal structure of the original motion model of the virtual object and the motion model of the virtual weapon are compatible. This means identifying and matching the corresponding bone nodes of the original motion structure of the virtual object and the motion structure of the virtual weapon. Second, skeletal mapping is performed, mapping the skeletal data of the virtual weapon's bone nodes onto the adapted virtual object's bone nodes. Then, interpolation or inverse kinematics techniques are used to adjust the position and rotation of the bone nodes in the virtual object to ensure a more natural-looking motion transition. Finally, manual correction and refinement are performed, manually correcting any unnatural-looking movements and refining the overall design.

[0409] When the player character picks up a virtual weapon, this embodiment of the application provides an action fusion and retention scheme based on the use of redirection technology. It retains some of the action performances with the characteristics and memorable points of the original character. Based on the similarity between the performance of the virtual weapon and the original character's actions, it drives a new combat style to achieve a better interactive effect.

[0410] For example, when the virtual weapon is a short stick, if the virtual character is a street thug, the action of controlling the thug to pick up the stick could be a simple bending down to pick it up, with a simplistic and brutal attack lacking in performance. However, if the virtual character is a master, the action of controlling the master to pick up the stick could be the master using their foot to lift the stick and catching it in mid-air, which is more expressive. The attack animation could be based on martial arts techniques using the stick, and there would also be a unique finishing move after the attack, allowing the same virtual weapon to reflect the characteristics of different characters.

[0411] The following describes the motion fusion retention scheme provided in this application embodiment. Referring to Figure 13, which is a schematic diagram of motion fusion provided in this application embodiment, Figure 13 shows the original character motion module, the virtual weapon motion module, and the redirected motion module. Different motion fusion ratios are designed for basic motions, special motions, attack motions, and other motions. Specifically, this can be achieved in the following way: First, the motion of a virtual object can be divided into two types based on its performance and functionality: representative (i.e., unique) and universal. Uniqueness refers to actions that highlight the character's personality, such as the action of picking up a stick, or the performance action after a stick attack, as mentioned above. Universality refers to actions that highlight the stick attack, actions that any character can use after picking up a stick. Then, when redirecting the motion, all unique actions can be marked first, ensuring that these actions are not changed during motion redirection fusion. Finally, after the universal actions are fused, they will be replaced with the new virtual weapon motion module.

[0412] As an example, refer to Figures 14A to 14C. Figure 14A is the ninth schematic diagram of a virtual scene provided in an embodiment of this application. The action types shown in Figure 14A are all general actions of the player character. After being merged with the actions of the virtual weapon, they will be replaced with the actions shown in Figure 14B. Figure 14B is the tenth schematic diagram of a virtual scene provided in an embodiment of this application. Figure 14C is the eleventh schematic diagram of a virtual scene provided in an embodiment of this application. The action types shown in Figure 14C are all representative actions of the player character. Among them, the actions on the left and in the middle are representative actions during the attack process, and the actions on the right side of the screen are finishing actions at the end of the battle.

[0413] The motion fusion preservation scheme provided in this application can retain the original motion characteristics of the player character, and has the following technical effects: 1) Authentic experience: It can retain the original style and charm of the player character's original motion, making the final fused motion more meaningful and layered. 2) Enhanced recognizability: Each motion of the retained player character has its unique characteristics and rhythm. Retaining these characteristics makes the fused motion easier to recognize, improving its visual appeal and attractiveness. 3) Dynamic expression: Retaining these characteristics during the fusion process makes the new combined motion more precise in expressing emotions and intentions. 4) Technical depth: Maintaining the characteristics of the original motion can better showcase the technical content of different motions, making the result of motion fusion more professional and visually appealing.

[0414] The following beneficial technical effects can be achieved during the technical implementation process: 1) In related technologies, different motion models need to be established for different virtual objects. When multiple virtual objects use the same virtual weapon, multiple animation files for different virtual objects need to be created and exported, resulting in an exponential increase in the workload of artists and the time cost of project development. The interactive processing method for virtual scenes provided in this application can effectively reduce the workload of artists. 2) It facilitates motion maintenance. Compared with related technologies that require the maintenance of a large number of character models, which leads to human error, the above-mentioned solution provided in this application only needs to maintain the source motion model of one virtual weapon, which can be adapted to multiple virtual objects through redirection technology. When modification or optimization is required, it can effectively reduce human error during the maintenance process. 3) The above-mentioned technical solution provided in this application can effectively reduce the system resource occupation of the game package during the development process.

[0415] The following description continues to illustrate the exemplary structure of the virtual scene interaction processing device 455 provided in the embodiments of this application as a software module. In some embodiments, as shown in FIG2, the software module stored in the virtual scene interaction processing device 455 in the memory 450 may include:

[0416] Display module 4551 is used to display a virtual scene, wherein the virtual scene includes at least one virtual object.

[0417] The picking module 4552 is used to control the first virtual object to pick up the first virtual prop in response to the fulfillment of the picking conditions, wherein the first virtual object is any virtual object in the virtual scene.

[0418] The control module 4553 is used to respond to a trigger operation on the first virtual prop and control the first virtual object to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

[0419] In some embodiments, the action sequence includes multiple action combinations, an action combination includes one or more actions, an action combination is used to implement a skill, and the type of the skill is adapted to the characteristics of the first virtual prop.

[0420] In some embodiments, the adaptation method includes: when the skill type is a normal attack skill, the attack power of the normal attack skill is positively correlated with the attack parameters of the first virtual item, and the normal attack skill is a skill whose damage value is lower than the damage threshold; when the skill type is a defensible attack skill, the attack power of the defensible attack skill is positively correlated with the defense parameters of the first virtual item; when the skill type is an undefensible attack skill, the attack power of the undefensible attack skill is positively correlated with the defense parameters of the first virtual item; when the skill type is a defensive skill, the evasion ability of the defensive skill is positively correlated with the protection range of the first virtual item.

[0421] In some embodiments, the picking module 4552 is further configured to, after controlling the first virtual object to pick up the first virtual prop, in response to satisfying a discard condition, control the first virtual object to discard the first virtual prop; in response to a picking trigger operation for the first virtual prop in the virtual scene, control the second virtual object to pick up the first virtual prop; and in response to a trigger operation for the first virtual prop, control the second virtual object to perform an action sequence based on the first virtual prop.

[0422] In some embodiments, the discarding conditions include any one of the following: receiving a discard trigger operation for the first virtual item; the first virtual object holding the first virtual item completing a predetermined interaction task; the first virtual object holding the first virtual item for a predetermined duration; and a second virtual item existing at a position less than a distance threshold from the first virtual object.

[0423] In some embodiments, the control module 4553 is further configured to display a plurality of usage mode controls for the first virtual prop, wherein a usage mode control is used to characterize a usage mode of the first virtual prop, and a usage mode includes a sequence of actions for using the first virtual prop; in response to a trigger operation on the first usage mode control, the control module 4553 controls the first virtual object to perform the sequence of actions included in the triggered first usage mode based on the first virtual prop.

[0424] In some embodiments, the usage mode includes any of the following: using the first virtual prop to attack, using the first virtual prop to defend, and using the first virtual prop to both defend and counterattack.

[0425] In some embodiments, an object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item; the control module 4553 is further configured to combine at least one object action in the object action template and at least one item action in the item action template into an action sequence.

[0426] In some embodiments, the control module 4553 is further configured to perform any one of the following processes: when the object action template is used to implement multiple first skills and the prop action template is used to implement multiple second skills, combining the object action in the object action template used to implement at least one first skill and the prop action in the prop action template used to implement at least one second skill into an action sequence; when the object action template is used to implement a third skill and the prop action template is used to implement a fourth skill, combining at least one object action in the object action template used to implement the third skill and at least one prop action in the prop action template used to implement the fourth skill into an action sequence.

[0427] In some embodiments, the types of object actions in the object action template include representative actions and general actions. The representative action is an action that reflects the character characteristics of the first virtual object, and the general action is an action that is applicable to multiple virtual objects. The control module 4553 is also used to remove at least one object action that is not a representative action from the object action template, and insert a prop action in the prop action template that is in the same position as the object action that was removed at the position where the object action was removed, to obtain an action sequence.

[0428] In some embodiments, the control module 4553 is further configured to, in response to a marking operation for at least one object action in the object action template, display a marking result, wherein the marking result indicates whether the object action is in a retained state or a non-retained state; and, in response to a composition operation, replace the object action in the object action template that is in a non-retained state with a prop action in the prop action template that is in the same position as the object action in a non-retained state, to obtain an action sequence.

[0429] In some embodiments, the control module 4553 is further configured to, in response to a selection operation for at least one object action in the object action template, control the selected at least one object action to be in a selected state; in response to a selection operation for at least one prop action in the prop action template, control the selected at least one prop action to be in a selected state; and in response to a synthesis operation, synthesize the selected at least one prop action and at least one object action to obtain an action sequence.

[0430] In some embodiments, an object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item; the control module 4553 is further configured to merge each object action in the object action template with the item action in the item action template that is in the same position as the object action to obtain an action.

[0431] In some embodiments, object actions are represented by first skeletal data, and prop actions are represented by second skeletal data; the control module 4553 is further configured to extract object action features from the first skeletal data and prop action features from the second skeletal data; determine a first weight of the object action features and a second weight of the prop action features; and based on the first weight and the second weight, fuse the object action with prop actions in the prop action template that are in the same position as the object action to obtain an action.

[0432] In some embodiments, the control module 4553 is further configured to perform attention encoding based on the role characteristics of the first virtual object and the prop characteristics of the first virtual prop to obtain a first weight of the object action characteristics and a second weight of the prop action characteristics; or, to display a weight editing control, and in response to an editing operation on the weight editing control, to display the first weight of the object action characteristics and the second weight of the prop action characteristics.

[0433] In some embodiments, the control module 4553 is further configured to control the first virtual object to perform multiple actions included in the action sequence, wherein, for any two adjacent actions, if the magnitude of the action change between the two actions is greater than a magnitude threshold, then between the two actions, the first virtual object is controlled to perform at least one transitional action.

[0434] In some embodiments, object actions are represented by the skeletal data of a first bone, and prop actions are represented by the skeletal data of a second bone; the control module 4553 is further configured to match the first bone point of the first bone with the second bone point of the second bone to obtain a first matching result; based on the first matching result, map the skeletal data of the second bone onto the first bone; and based on the pose of the second bone point of the second bone, adjust the first bone point of the first bone at least once according to a set adjustment range to obtain skeletal data of at least one transitional action, wherein the skeletal data of at least one transitional action is used to control the first virtual object to perform at least one transitional action.

[0435] In some embodiments, the picking module 4552 is further configured to control the first virtual object to perform multiple intermediate actions of the picking process starting from the state where the first virtual object is not picking up the first virtual item, and after performing the intermediate actions, control the first virtual object to be in the state where it has picked up the first virtual item.

[0436] In some embodiments, the intermediate actions conform to the character characteristics of the first virtual object; the preceding action of each intermediate action is represented by the bone data of the third bone, and the following action of each intermediate action is represented by the bone data of the fourth bone; the picking module 4552 is further configured to match the third bone point of the third bone with the fourth bone point of the fourth bone to obtain a second matching result; according to the second matching result, map the bone data of the fourth bone onto the third bone; according to the pose of the fourth bone point of the fourth bone, adjust the third bone point of the third bone at least once according to a set adjustment range to obtain bone data of at least one intermediate action, wherein the bone data of at least one intermediate action is used to control the first virtual object to perform at least one intermediate action of multiple picking processes.

[0437] In some embodiments, the control module 4553 is further configured to display an action configuration control for a first virtual prop, wherein the action configuration control includes a plurality of actions included in the action sequence, and the actions are any one of prop actions and object actions; in response to a configuration operation for any one action, the updated action sequence is displayed, wherein the updated action sequence is used to replace the previous action sequence.

[0438] In some embodiments, the picking conditions include any one of the following: receiving a trigger operation for a picking control in a virtual scene; the distance between the first virtual object and the first virtual prop is less than a threshold, and the first virtual object does not currently hold the first virtual prop; the distance is less than a threshold, and the performance parameters of the currently held second virtual prop are less than those of the first virtual prop; the distance is less than a threshold, and the first virtual prop is compatible with the currently interacting second virtual object; the distance is less than a threshold, and the first virtual prop is compatible with the environment in which the first virtual prop is currently located.

[0439] This application provides a computer program product, which includes a computer program or computer-executable instructions stored in a computer-readable storage medium. The processor of an electronic device reads the computer-executable instructions from the computer-readable storage medium and executes the computer-executable instructions, causing the electronic device to perform the virtual scene interaction processing method described above in this application.

[0440] This application provides a computer-readable storage medium storing computer-executable instructions or computer programs. When the computer-executable instructions or computer programs are executed by a processor, the processor will execute the virtual scene interaction processing method provided in this application, such as the virtual scene interaction processing method shown in FIG3A.

[0441] In some embodiments, the computer-readable storage medium may be a memory such as RAM, ROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or it may be a variety of devices including one or any combination of the above-mentioned memories.

[0442] In some embodiments, computer-executable instructions may take the form of programs, software, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.

[0443] As an example, computer-executable instructions may, but do not necessarily, correspond to files in a file system. They may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a Hyper Text Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple co-located files (e.g., files that store one or more modules, subroutines, or code sections).

[0444] As an example, computer-executable instructions can be deployed to execute on a single electronic device, or on multiple electronic devices located at one location, or on multiple electronic devices distributed across multiple locations and interconnected via a communication network.

[0445] In summary, in this embodiment of the application, when the picking conditions are met, the first virtual object in the virtual scene can be controlled to pick up the first virtual item. When the first virtual object picks up the first virtual item, the first virtual object can be triggered to perform an action sequence based on the first virtual item. This allows the action sequence performed during the interaction between the virtual object and the virtual item to retain both the object action of the virtual object and the item action of the virtual item, thereby improving the diversity and adaptability of actions during the interaction between the virtual object and the virtual item, and thus enhancing the player's experience.

[0446] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, and improvements made within the spirit and scope of this application are included within the scope of protection of this application.

Claims

1. A method for interactive processing of a virtual scene, executed by an electronic device, the method comprising: Displaying a virtual scene, wherein the virtual scene includes at least one virtual object; In response to the fulfillment of the picking conditions, the first virtual object is controlled to pick up the first virtual prop, wherein the first virtual object is any one of the virtual objects in the virtual scene; In response to a trigger operation on the first virtual prop, the first virtual object is controlled to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

2. The method according to claim 1, wherein, The action sequence includes multiple action combinations, each action combination including one or more actions, and each action combination is used to implement a skill, the type of which is adapted to the characteristics of the first virtual prop.

3. The method according to claim 2, wherein, The adaptation methods include: When the type of the skill is a normal attack skill, the attack power of the normal attack skill is positively correlated with the attack parameters of the first virtual item, and the normal attack skill is a skill whose damage value is lower than the damage threshold; When the type of the skill is a defensible attack skill, the attack power of the defensible attack skill is positively correlated with the defense parameters of the first virtual item; When the type of the skill is an undefendable attack skill, the attack power of the undefendable attack skill is positively correlated with the attack parameters of the first virtual item; When the skill type is a defensive skill, the evasion capability of the defensive skill is positively correlated with the protection range of the first virtual item.

4. The method according to any one of claims 1 to 3, wherein, After controlling the first virtual object to pick up the first virtual item, the method further includes: In response to the fulfillment of the discard condition, the first virtual object is controlled to discard the first virtual item; In response to a pickup trigger operation for the first virtual prop in the virtual scene, control the second virtual object to pick up the first virtual prop; In response to a trigger operation on the first virtual prop, the second virtual object is controlled to perform the action sequence based on the first virtual prop.

5. The method according to claim 4, wherein, The discard conditions include any of the following: A discard trigger operation was received for the first virtual item; The first virtual object, holding the first virtual item, completed the predetermined interactive task; The first virtual object holds the first virtual item for a predetermined duration; A second virtual object exists at a position less than a distance threshold from the first virtual object.

6. The method according to claim 4 or 5, wherein, The control of the second virtual object to perform the action sequence based on the first virtual prop includes: Adjust the weight of the prop actions based on the character characteristics of the second virtual object: If the second virtual object is a heavy character, increase the weight of the defensive stance in the prop action; If the second virtual object is an agile character, increase the weight of the rapid slashing action in the prop.

7. The method according to any one of claims 1 to 6, wherein, The control of the first virtual object to perform an action sequence based on the first virtual prop includes: Display multiple usage mode controls for the first virtual item, wherein one of the usage mode controls is used to characterize a usage mode of the first virtual item, and the usage mode includes a sequence of actions for using the first virtual item; In response to a trigger operation on a first usage mode control, the first virtual object is controlled to perform the action sequence included in the triggered first usage mode based on the first virtual item.

8. The method according to claim 7, wherein, The usage mode includes any of the following: using the first virtual item to attack, using the first virtual item to defend, and using the first virtual item to both defend and counterattack.

9. The method according to claim 1 or 8, wherein, An object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item; Before controlling the first virtual object to perform an action sequence based on the first virtual prop, the method further includes: The action sequence is formed by combining at least one object action from the object action template and at least one prop action from the prop action template.

10. The method according to claim 9, wherein, The step of combining at least one object action from the object action template and at least one prop action from the prop action template into the action sequence includes: Perform any of the following processes: When the object action template is used to perform multiple first skills and the prop action template is used to perform multiple second skills, the object action in the object action template used to perform at least one first skill and the prop action in the prop action template used to perform at least one second skill are combined into the action sequence. When the object action template is used to perform a third skill and the prop action template is used to perform a fourth skill, at least one object action in the object action template used to perform the third skill and at least one prop action in the prop action template used to perform the fourth skill are combined into the action sequence.

11. The method according to claim 9, wherein, When combining at least one object action from the object action template and at least one prop action from the prop action template into the action sequence, the method further includes: When the skeletal points of the object's action and the prop's action overlap, the following rules apply: If the object's action is representative, retain the object's action and adjust the bone position of the prop's action; If the prop action is a functional action, retain the prop action and adjust the amplitude of the object action.

12. The method according to claim 10, wherein, The object action template includes representative actions and general actions. The representative actions are those that reflect the character characteristics of the first virtual object, and the general actions are those that are applicable to multiple virtual objects. The step of combining at least one object action from the object action template and at least one prop action from the prop action template into the action sequence includes: Remove at least one object action from the object action template that is not the representative action, and insert a prop action from the prop action template that is in the same position as the removed object action at the position where the object action was removed, to obtain the action sequence.

13. The method according to claim 10, wherein, The step of combining at least one object action from the object action template and at least one prop action from the prop action template into the action sequence includes: In response to a marking operation for at least one of the object actions in the object action template, a marking result is displayed, wherein the marking result indicates whether the object action is in a retained state or a non-retained state; In response to the composition operation, the object action in the object action template that is in the non-retained state is replaced with the prop action in the prop action template that is in the same position as the object action in the non-retained state, to obtain the action sequence.

14. The method of claim 10, wherein, The step of combining at least one object action from the object action template and at least one prop action from the prop action template into the action sequence includes: In response to a selection operation for at least one object action in the object action template, control the at least one selected object action to be in a selected state; In response to a selection operation for at least one item action in the item action template, control the at least one selected item action to be in a selected state; In response to the compositing operation, at least one of the selected prop actions and at least one of the selected object actions are combined to obtain the action sequence.

15. The method according to any one of claims 1 to 14, wherein, An object action template is pre-set for the first virtual object, and an item action template is pre-set for the first virtual item; Before controlling the first virtual object to perform an action sequence based on the first virtual prop, the method further includes: Each object action in the object action template is merged with the prop action in the prop action template that is in the same position as the object action to obtain the action.

16. The method according to claim 15, wherein, The object's actions are represented by first skeletal data, and the prop's actions are represented by second skeletal data; The step of fusing each object action in the object action template with the prop action in the prop action template that is in the same position as the object action to obtain the action includes: Extract the object motion features from the first skeletal data and the prop motion features from the second skeletal data; Determine the first weight of the object's motion features and the second weight of the prop's motion features; Based on the first weight and the second weight, the object action is merged with the prop action in the prop action template that is in the same position as the object action to obtain the action.

17. The method according to claim 16, wherein, The determination of the first weight of the object's motion features and the second weight of the prop's motion features includes: Based on the character characteristics of the first virtual object and the item characteristics of the first virtual prop, attention encoding is performed to obtain a first weight for the object's action characteristics and a second weight for the prop's action characteristics; or... Display a weight editing control, and in response to an editing operation on the weight editing control, display a first weight of the object's action feature and a second weight of the prop's action feature.

18. The method according to any one of claims 1 to 15, wherein, The control of the first virtual object to perform an action sequence based on the first virtual prop includes: The first virtual object is controlled to execute multiple actions included in the action sequence. For any two adjacent actions, if the change in the action between the two actions is greater than a threshold, then the first virtual object is controlled to perform at least one transitional action between the two actions.

19. The method of claim 16, wherein, The at least one transitional action is generated in the following way: A pre-trained machine learning model (such as LSTM or GAN) is invoked to generate the transitional action that conforms to the action logic, based on the skeletal data of the two adjacent actions, the character features of the first virtual object, and the prop features of the first virtual prop.

20. The method according to claim 18, wherein, The object's actions are represented by the skeletal data of the first skeleton, and the prop's actions are represented by the skeletal data of the second skeleton. Before controlling the first virtual object to perform at least one transitional action between the two actions, the method further includes: Match the first bone point of the first bone with the second bone point of the second bone to obtain the first matching result; Based on the first matching result, the skeletal data of the second bone is mapped onto the first bone; Based on the pose of the second bone point of the second bone, the first bone point of the first bone is adjusted at least once according to a set adjustment range to obtain bone data for at least one transitional action, wherein the bone data for at least one transitional action is used to control the first virtual object to perform at least one transitional action.

21. The method according to any one of claims 1 to 20, wherein, The control of the first virtual object to pick up the first virtual item includes: Starting from the state where the first virtual object has not picked up the first virtual item, the system controls the first virtual object to perform multiple intermediate actions in the picking process, and after performing the intermediate actions, the system controls the first virtual object to be in the state where it has picked up the first virtual item.

22. The method according to claim 21, wherein, The intermediate actions conform to the character characteristics of the first virtual object; the preceding action of each intermediate action is represented by the skeletal data of the third bone, and the following action of each intermediate action is represented by the skeletal data of the fourth bone. Before the intermediate actions of controlling the first virtual object to perform multiple picking processes, the method further includes: The third bone point of the third bone is matched with the fourth bone point of the fourth bone to obtain a second matching result; Based on the second matching result, the bone data of the fourth bone is mapped onto the third bone; Based on the pose of the fourth bone point of the fourth bone, the third bone point of the third bone is adjusted at least once according to a set adjustment range to obtain bone data of at least one intermediate action, wherein the bone data of at least one intermediate action is used to control the first virtual object to perform at least one intermediate action of multiple picking processes.

23. The method according to any one of claims 1 to 22, wherein, Before controlling the first virtual object to perform an action sequence based on the first virtual prop, the method further includes: Displays an action configuration control for the first virtual prop, wherein the action configuration control includes multiple actions included in the action sequence, and the action is any one of the prop action and the object action; In response to a configuration operation for any of the actions, an updated action sequence is displayed, wherein the updated action sequence replaces the previous action sequence.

24. The method according to claim 23, wherein, When displaying the action configuration control for the first virtual prop, the method further includes: Display an action preview window to show the action effects corresponding to the configured operation in real time; in response to the preview confirmation operation, save the updated action sequence.

25. The method according to any one of claims 1 to 24, wherein, The picking conditions include any one of the following: A trigger operation was received for the pickup control in the virtual scene; The distance between the first virtual object and the first virtual item is less than a threshold, and the first virtual object does not currently possess the first virtual item; The distance is less than the threshold, and the performance parameters of the currently held second virtual item are less than those of the first virtual item; The distance is less than the threshold, and the first virtual prop is compatible with the currently interacting second virtual object; The distance is less than the threshold, and the first virtual item is compatible with the environment in which the first virtual item is currently located.

26. An interactive processing device for a virtual scene, the device comprising: A display module is used to display a virtual scene, wherein the virtual scene includes at least one virtual object; A picking module is used to control a first virtual object to pick up a first virtual prop in response to the fulfillment of picking conditions, wherein the first virtual object is any one of the virtual objects in the virtual scene; A control module is configured to respond to a trigger operation on the first virtual prop and control the first virtual object to perform an action sequence based on the first virtual prop, wherein the action sequence includes prop actions adapted to the first virtual prop and object actions adapted to the first virtual object.

27. An electronic device, the electronic device comprising: Memory is used to store executable instructions or computer programs. A processor, when executing computer-executable instructions or computer programs stored in the memory, implements the interactive processing method of the virtual scene as described in any one of claims 1 to 26.

28. A computer-readable storage medium storing computer-executable instructions or a computer program, wherein the computer-executable instructions or the computer program, when executed by a processor, implement the interactive processing method of the virtual scene according to any one of claims 1 to 25.

29. A computer program product comprising computer-executable instructions or a computer program, wherein the computer-executable instructions or the computer program, when executed by a processor, implement the interactive processing method of the virtual scene according to any one of claims 1 to 25.

Images