Instant data analysis and context-aware anchor broadcasting method, device and medium
By using real-time game data analysis and context-aware technology, the system automatically determines the observer's perspective and generates professional broadcast content, solving the problems of data integration delay and insufficient tactical understanding in existing technologies, and achieving more real-time and realistic virtual e-sports broadcasts.
Patent Information
- Application Number
- CN202610261652.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-04
- Publication Date
- 2026-06-05
AI Technical Summary
Current esports broadcasting methods rely on manual operation, which cannot integrate multi-dimensional data in real time. This results in broadcast delays, errors, and missed opportunities for key tactics. It also lacks tactical insight and situational understanding, affecting the viewer experience.
By combining real-time game data analysis, action event prediction, tactical intent inference, and deep contextual models with natural language processing technology, virtual e-sports anchor broadcasts with temporal continuity and tactical relevance are generated, automatically determining the viewer's perspective and adjusting the emotional rendering effects.
It has improved the ability to capture key tactical events, optimized the continuity of the virtual anchor's emotional expression and the professionalism of the broadcast, and enhanced the immediacy and realism of the broadcast.
Smart Images

Figure CN122153321A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to an e-sports event broadcasting technology, specifically combining real-time game data acquisition technology, action event prediction model, tactical intent inference model, deep context model, and natural language processing technology to realize a virtual e-sports anchor broadcasting method, computer equipment, and computer-readable recording medium based on real-time game data analysis and context awareness. Background Technology
[0002] With the rapid development of the esports industry, the demand for live streaming and broadcasting of esports events is increasing. Current esports broadcasting methods largely rely on human announcers and observers switching between camera angles and providing tactical commentary. However, this method heavily depends on the experience and judgment of announcers and directors, making it impossible to integrate large amounts of data in real time, such as character status, skill cooldown times, equipment differences, character movement routes, and tactical deployments. Because the amount of data during an event is massive and changes rapidly, it is difficult for humans to complete multi-dimensional data integration and analysis in a short period, potentially leading to delays, judgment errors, or omissions of key tactical shifts in the broadcast.
[0003] In addition, the existing observer perspective footage is mostly switched by human operators using camera angles. When multiple potentially important events occur simultaneously in multiple areas of the competition field, such as team battles, resource contests, or ambushes, it is difficult for humans to judge the most critical focus of the footage in time. This often results in the footage switching being too slow or missing key tactical moments, which in turn affects the audience's viewing experience and the integrity of the event.
[0004] On the other hand, existing virtual anchors or voice broadcasting systems mostly rely on preset speech synthesis rules or fixed scripts for broadcasting, lacking a deep understanding of real-time tactical situations and character behavior. This approach fails to reflect the tension of the competition situation and the trend of tactical changes, resulting in a lack of temporal continuity and hierarchical variation in the emotional presentation of the broadcast, making it difficult to create a realistic sense of presence.
[0005] Furthermore, existing broadcast generation technologies mostly use simple template replacement or keyword matching to generate broadcast content, failing to integrate and analyze semantic levels by combining tactical intentions, multi-role interaction relationships, and real-time data advantages and disadvantages. As a result, the generated broadcast statements often remain superficial descriptions and lack the tactical insight and situational reasoning ability required for professional e-sports broadcasting.
[0006] In summary, existing technologies have long suffered from a lack of features such as automatically determining the viewer's perspective position based on real-time game data, accurately analyzing tactical intentions and combining them with emotional indicators to generate professional broadcast content, and achieving virtual anchor emotional rendering effects with temporal continuity and tactical relevance. Therefore, it is necessary to propose an improved solution that integrates real-time game data analysis, action event prediction, tactical intention inference, emotional indicator calculation, and natural language generation technologies to enhance the immediacy, accuracy, and immersive experience of esports event broadcasts. Summary of the Invention
[0007] In view of the shortcomings of existing technologies, such as the lack of ability to automatically determine the viewer's perspective position based on real-time game data, accurately analyze tactical intentions and combine them with emotional indicators to generate professional broadcast content, and achieve virtual anchor emotional rendering effects with temporal continuity and tactical relevance, this invention discloses a virtual e-sports anchor broadcasting method, computer equipment, and computer-readable recording medium based on real-time game data analysis and context awareness, wherein: This invention discloses a virtual esports broadcasting method based on real-time game data analysis and context awareness, which involves performing the following steps using a computer device: The system obtains real-time game data and observer-view footage from the competitive game server via the game's application programming interface (API). Based on the observer-view footage, it filters out the real-time game data of the corresponding character, and parses the character's status, position, movement path, and action instructions to extract action events and tactical intentions. It then uses a deep contextual model to obtain emotion indicators from the filtered character's real-time game data, the extracted action events, and tactical intentions. Natural language processing technology is used to combine the tactical intentions and the filtered character's real-time game data with a competitive terminology database to generate real-time broadcast content. A virtual esports commentator is integrated into the match stream. The virtual esports commentator is controlled to deliver voice broadcasts and expressive emotions based on the emotion indicators, providing esports event commentary.
[0008] In one embodiment of the present invention, the character state, position, and movement route of each character in the real-time game data are used to determine the viewpoint position of the observer's game view.
[0009] In one embodiment of the present invention, when the position of each character in the real-time game data is close to a preset important resource position, or when the positions of more than or equal to a preset number of game characters in the real-time game data are close to a preset important resource position, the viewpoint position of the observer's game view screen is set to the important resource position.
[0010] In one embodiment of the present invention, a deep context model is trained using real-time game data pre-labeled with corresponding emotion scores and action events pre-labeled with corresponding emotion scores.
[0011] In one embodiment of the present invention, the emotional index is accumulated and calculated over time within the time range of the action event, so that the emotional state of the virtual e-sports anchor's e-sports event broadcast has the effect of continuous change with time correlation.
[0012] In one embodiment of the present invention, the tone, speed and volume of the real-time broadcast content are adjusted based on the emotion corresponding to the emotion index, so as to present different emotion rendering effects.
[0013] This invention discloses a computer device, which includes: The storage device stores multiple computer-readable instructions; and One or more hardware processors are electrically connected to a storage device and execute multiple computer-readable instructions to enable a computer device to implement the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness, as described above.
[0014] This invention discloses a computer-readable recording medium storing a computer program. When the computer program is executed by one or more hardware processors of a computer device, it causes the computer device to execute a virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness.
[0015] The method, computer equipment, and computer-readable recording medium disclosed in this invention are as described above. By integrating multiple levels of technologies such as real-time game data acquisition, character behavior feature analysis, action event prediction, tactical intent inference, deep context model emotion calculation, and natural language generation, it can not only identify the behavior of a single character, but also understand the collaborative tactical relationship of multiple characters. Furthermore, it combines a deep context model to generate temporally continuous emotion indicators, enabling the voice broadcast and facial expression rendering of virtual e-sports anchors to be dynamically adjusted according to the game situation, thereby enhancing the realism and immersiveness of the broadcast.
[0016] Through the aforementioned technical means, this invention can achieve the technical effects of enhancing the ability to capture key tactical events, optimizing the continuity of virtual anchors' emotional expressions, and improving the semantic professionalism of real-time broadcasts. Attached Figure Description
[0017] Figure 1 The diagram illustrates a flowchart of the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness according to the present invention.
[0018] Figure 2 The illustration is a schematic diagram of the observer's game perspective screen for the virtual e-sports anchor broadcast based on real-time game data analysis and context awareness according to the present invention.
[0019] Figure 3 The illustration shows the real-time broadcast content of the virtual e-sports anchor broadcast based on real-time game data analysis and context awareness according to the present invention.
[0020] Figure 4 The illustration shows a schematic diagram of the integrated virtual e-sports anchor broadcasting of the present invention, which is based on real-time game data analysis and context awareness.
[0021] Figure 5 The diagram illustrates the computer equipment system architecture of the virtual e-sports anchor broadcasting based on real-time game data analysis and context awareness according to the present invention.
[0022] Explanation of reference numerals in the attached figures: Step 101: Obtain real-time game data and observer's game perspective from the competitive game server through the game application interface. Step 102: Based on the observer's game perspective, filter out the real-time game data of the corresponding character, and analyze the action events and tactical intentions from the character's status, position, movement route, and action instructions in the real-time game data. Step 103: Use a deep contextual model to obtain sentiment indicators from the real-time game data of the selected characters, the parsed action events, and tactical intentions. Step 104: Use natural language processing technology to combine tactical intentions and real-time game data of the selected characters with a competitive terminology database to generate real-time broadcast content. Step 105: Integrate virtual esports streamers into the match streaming screen. Step 106: Control the virtual esports anchor to deliver real-time broadcasts with voice and facial expressions based on the emotions corresponding to the emotion indicators, in order to provide esports event broadcasts. 20: Observer's game perspective screen 21: First-party online hero characters 22: First-party soldier role 23: Second-tier top-tier hero characters 24: Second-party soldier role 25: Second Roaming Hero Character 26: Object in the grass above 27: Grassy object below 30: Real-time broadcast content 40: Match Streaming Footage 41: Virtual e-sports streamer 500: Computer System 501: CPU 502: ROM 503: RAM 504: Bus 505: I / O Interface 506: Input section 507: Output Section 508: Storage Section 509: Communications Section 510: Driver 511: Removable media Detailed Implementation The following will describe in detail the implementation of the present invention with reference to the accompanying drawings and embodiments, so as to fully understand how the present invention uses technical means to solve technical problems and achieve technical effects and to implement it accordingly.
[0023] The following section will first explain the virtual esports broadcasting method based on real-time game data analysis and context awareness disclosed in this invention, and please refer to [the relevant documentation / reference]. Figure 1 As shown, Figure 1 The diagram illustrates a flowchart of the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness according to the present invention.
[0024] This invention discloses a virtual esports broadcasting method based on real-time game data analysis and context awareness, which involves performing the following steps using a computer device: The game application interface obtains real-time game data and the observer's game perspective from the competitive game server (step 101); based on the observer's game perspective, the real-time game data of the corresponding character is selected, and the action events and tactical intentions are parsed from the character's status, position, movement route, and action instructions in the real-time game data (step 102); the real-time game data of the selected corresponding character and the parsed action events and tactical intentions are used with a deep context model to obtain emotion indicators (step 103); natural language processing technology is used to combine the tactical intentions and the real-time game data of the selected corresponding character with a competitive terminology database to generate real-time broadcast content (step 104); a virtual e-sports anchor is integrated into the match streaming screen (step 105); and the virtual e-sports anchor is controlled to broadcast the real-time broadcast content with voice and facial expression based on the emotion corresponding to the emotion indicators to provide e-sports event broadcasts (step 106).
[0025] The computer device obtains real-time game data and the observer's game perspective from the competitive game server through a game application interface. The aforementioned computer device can be implemented as a single computer or a distributed processing architecture. The game application interface enables different systems to communicate with each other without disclosing the internal implementation details through predefined communication standards, data formats, and request-response mechanisms. The computer device further standardizes the real-time game data of the competitive game to provide a standardized structured data format (e.g., JSON). This is only an example and does not limit the scope of application of the present invention.
[0026] Please refer to the aforementioned observer game perspective screenshot 20. Figure 2 As shown, Figure 2 The illustration is a schematic diagram of the observer's game perspective view of the virtual e-sports anchor broadcast based on real-time game data analysis and context awareness according to the present invention. Figure 2 The game features a first-team lane hero character 21, multiple first-team minion characters 22, a second-team top lane hero character 23, multiple second-team minion characters 24, and a second roaming hero character 25. The second roaming hero character 25 moves from the upper bush object 26 to the lower bush object 27. The second-team top lane hero character 23 attempts to lure the first-team lane hero character 21 to approach the lower bush object 27, thus achieving the tactical intention of the second-team top lane hero character 23 and the second roaming hero character 25 to cooperate in attacking the first-team lane hero character 21.
[0027] The computer device filters out the real-time game data of the corresponding characters based on the real-time game data of the competitive game in the observer's game perspective screen 20. Specifically, it filters out the real-time game data of the first-party online hero character 21, the real-time game data of multiple first-party minion characters 22, the real-time game data of the second-party online hero character 23, the real-time game data of multiple second-party minion characters 24, and the real-time game data of the second roaming hero character 25. The aforementioned real-time game data includes, but is not limited to, character status, character position, movement route, and action instructions. In this embodiment, the character status of the second roaming hero character 25 includes, but is not limited to, the health and mana of the second roaming hero character 25. Value / energy value, equipment of the second roaming hero character 25, skills of the second roaming hero character 25, skill cooldown time of the second roaming hero character 25, etc. In this embodiment, the character position of the second roaming hero character 25 is located at the lower bush object 27. In this embodiment, the movement route of the second roaming hero character 25 is from the upper bush object 26 to the lower bush object 27. In this embodiment, the action instruction operation issued by the player operating the second roaming hero character 25 is the action instruction operation of the player operating the second roaming hero character 25 to notify the player operating the second lane upper hero character 23 to prepare to act. This is only an example for illustration and does not limit the application scope of the present invention.
[0028] It is worth noting that the aforementioned observer game perspective screen 20 is further determined by the computer device using an action event prediction model to determine the position of the observer game perspective screen based on the character status, position, and movement route of each character in the real-time game data.
[0029] Specifically, the computer device can first convert the real-time game data of each character into a corresponding feature vector. Character status features may include the percentage of health points, the percentage of mana / energy / rage points, skill cooldown time, equipment strength rating, whether there are buffs or debuffs, etc.; the character position features may include the character's real-time coordinates in the map coordinate system, the distance to important resource locations (e.g., dragons, barons, etc.), and the distance to other characters; the movement route features may include the movement trajectory sequence, speed vector, rate of change of movement direction angle, and acceleration trend within a preset time window, etc. The computer device can set a sliding time window (e.g., 2 seconds to 5 seconds), establish a time series feature matrix for each character within this time window, and integrate the feature matrices of multiple characters into a global contextual feature vector to reflect the overall state of the current game.
[0030] Action event prediction models can be, for example, time series classification models or multi-role association models, such as Long Short-Term Memory Networks (LSTM), Gated Recurrent Units (GRU), Temporal Convolutional Networks (TCN), Graph Neural Networks (GNN), or combinations thereof. These models can be trained on historical match data to learn the correspondence between different combinations of role behaviors and potentially important events (e.g., team fights, ambushes, resource contention, or tower pushing).
[0031] The action event prediction model can output an event importance score or a focus weight score for each map region or character. For example, when the action event prediction model predicts that a certain area will have a high probability of a team fight in a short period of time, the focus weight score for that area will be significantly increased.
[0032] When the second roaming hero 25 and the second laner hero 23 move towards the bush object 27 below at the same time, and the skill cooldown status of the second roaming hero 25 and the second laner hero 23 shows that they can be used immediately, and at the same time the health of the first laner hero 21 is lower than the preset threshold, the action event prediction model can predict that a high-importance event (such as a cooperative attack or kill event) will occur in the bush object 27 below in a short period of time. At this time, the action event prediction model can output that the importance score of the bush object 27 below is higher than that of other areas.
[0033] The computer device can select the character or map area corresponding to the highest score based on the importance score, and automatically switch or pan the view of the observer's game view 20 to that character or area. When the number of character positions in the real-time game data that is greater than or equal to a preset number are close to the preset important resource positions (e.g., Dragon, Baron, High Ground Tower, etc.), the action event prediction model can increase the focus weight score of the important resource position, so that the observer's game view automatically focuses on the important resource position.
[0034] Furthermore, to avoid viewing discomfort caused by frequent screen switching, computer devices can be equipped with a viewpoint stabilization mechanism. This could involve setting a minimum dwell time threshold or a smooth weighted transition function, allowing the viewpoint position to smoothly change with focus weight rather than jumping abruptly. Through these techniques, the viewpoint position in the game is no longer dependent on manual switching, but is automatically determined based on the analysis results of multi-character behavior data and action event prediction models. This improves the immediacy, accuracy, and tactical precision of the visual presentation.
[0035] The computer device can parse action events and tactical intentions from the character's status, position, movement route, and action instructions in real-time game data. Based on a pre-set sliding time window (e.g., 2 to 5 seconds, which is only an example and does not limit the scope of application of this invention), the computer device performs feature aggregation on the real-time game data of the corresponding character within the time window, generating a corresponding feature vector. The aforementioned feature vector includes, but is not limited to, position change rate, speed vector, orientation angle change, area dwell time, skill cooldown transition status, health change trend, and whether there are action instructions, etc., so that the computer device can determine action events based on the feature vector through an event parsing model. Continuing with the above embodiment, when the second roaming hero character 25 moves from the upper grass object 26 to the lower grass object 27 within the time window, and enters the lower grass... When the position change of object 27 is lower than a preset threshold and remains there for more than a preset time, and an action instruction to prepare for action appears within the time window, the computer device can parse the action event as an "ambush and positioning event". When the hero character 23 above the second line moves towards the hero character 21 on the first line within the time window, and exhibits a movement pattern of pressing forward for a short time and then retreating slightly, and the skill cooldown status of the hero character 23 above the second line and the second roaming hero character 25 shows that the cooldown is complete, and the cumulative health and cumulative mana / energy / rage of the hero character 23 above the second line and the second roaming hero character 25 are greater than the health and mana / energy / rage of the hero character 21 on the first line, the computer device can parse the action event as an "enemy-luring and pressing event". This is only an example and is not intended to limit the application scope of the present invention.
[0036] After the computer device completes the analysis of action events, it can further analyze the tactical intent by performing temporal and spatial correlation analysis on multiple action events through a tactical intent inference model. As in the above embodiment, when there are "ambush positioning events" and "luring the enemy forward events" in the same time window, and the predicted path of the first-party online hero character 21 shows that it will approach the grass object 27 below, the computer device can infer that the tactical intent is "the tactical intent of cooperating to attack the first-party online hero character 21". This is only an example and does not limit the application scope of the present invention.
[0037] The computer device will filter out the real-time game data of the corresponding character and the parsed action events and tactical intentions, and use a deep contextual model to obtain emotion indicators. Specifically, the computer device can use the timestamp of the real-time game data to convert the action events and tactical intentions into semantic embedding vectors and input them into the deep contextual model. The deep contextual model may include a time series encoding layer (e.g., Long Short-Term Memory Network LSTM, Gated Recurrent Unit GRU, and Temporal Convolutional Network TCN, etc., which are only examples and do not limit the application scope of this invention), a multi-head attention mechanism layer, and a fully connected output layer, so that the deep contextual model can simultaneously capture real-time behavioral changes in a short period of time and tactical evolution trends in a long period of time.
[0038] Following the above embodiments, when a computer device detects both an "ambush and positioning event" and an "enemy-luring and advancing event" within the same time window and infers a "cooperative attack tactical intention," the deep context model can output corresponding emotion indicators based on features such as the type weight of the tactical intention, the urgency of the event, the difference in character health points, skill cooldown status, and the predicted success rate of the kill. The aforementioned emotion indicators can be continuous emotion scores (e.g., emotion scores between 0 and 10) or multi-dimensional vectors (e.g., including tension, excitement, and expectation indicators, etc.) to represent the intensity and type of the broadcast emotion.
[0039] It is worth noting that the emotion index is not an instantaneous single-point value, but rather is calculated cumulatively over time within the time range of the action event. That is, the computer device updates the emotion index at each time window and uses the emotion index of the previous time window as part of the recursive input, so that the emotion change has temporal continuity. Specifically, for example, when the cooperative attack tactical intention persists and the enemy character gradually approaches the grass object, the tension in the emotion index can gradually increase; when an actual kill occurs, the excitement index can rapidly increase; if the action fails or retreats, the emotion index can change to a state of loss or relief. This is only an example and is not intended to limit the application scope of the present invention.
[0040] It is worth noting that the training data for the deep context model consists of real-time game data with pre-labeled emotion scores and action events with pre-labeled emotion scores. The emotion scores are generated by professional esports commentators or by statistically analyzing audience reaction data from a large number of historical events (e.g., combinations of bullet screen density, chat room message frequency, and audience volume changes in decibels). Through supervised learning training, the deep context model can learn the correspondence between different tactical situations and emotion intensity.
[0041] The computer device uses natural language processing technology to combine tactical intentions and real-time game data of the selected characters with a competitive terminology database to generate real-time broadcast content. The aforementioned competitive terminology database pre-stores broadcast statement templates, character titles, skill names, game nicknames, and map area names, etc., corresponding to various tactical intentions. The computer device converts tactical intentions into semantic tags. For example, the tactical intention of "cooperatively attacking the first-party lane hero character 21" is converted into semantic structure data. This semantic structure data includes at least: intention type (e.g., gank, team fight, resource steal, counter-attack), initiating character set (e.g., the second-party lane hero character 23 and the second-party roaming hero character 25), target character (e.g., the first-party lane hero character 21), and key area location (e.g., the object 27 in the bush below), etc. This is only an example and does not limit the application scope of the present invention. At the same time, the computer device converts the real-time game data of the selected characters into structured descriptive information, such as: the character's current health percentage, skill cooldown status, equipment advantage difference, movement direction, and distance difference between the two sides, etc.
[0042] Specifically, the computer device selects the corresponding statement template from the competitive terminology database based on the tactical intent type: "{Initiating character} is preparing to launch a {tactical type} attack in {key area}!" Real-time game data is then filled into the variable fields of the statement template. For example, the initiating character's name is replaced with "Second lane top hero and second roaming hero," and the key area is replaced with "bottom bush." The device also determines whether to add strength adjectives based on the real-time game data. For example, when all skills are off cooldown and the health advantage is greater than a preset threshold, adjectives such as "powerful" or "extremely threatening" can be added. The resulting real-time broadcast content 30 is: "The second lane top hero is pushing forward! The second roaming hero is already ambushd in the bottom bush; this is a carefully planned cooperative attack!" Please refer to the example of real-time broadcast content 30. Figure 3 As shown, Figure 3 The illustration is a schematic diagram of the real-time broadcast content of the virtual e-sports anchor broadcast based on real-time game data analysis and context awareness of the present invention. If the real-time game data shows that the initiating party has a significant equipment advantage, the generated real-time broadcast content 30 is "This ambush in the bottom lane bush is quite deadly. The equipment gap between the two sides is obvious. The first party's lane hero will probably not be able to escape unscathed!" This is only an example and is not intended to limit the application scope of the present invention.
[0043] It is worth noting that the computer device can also adjust the real-time broadcast content 30 based on the sentiment indicators output by the deep context model. For example, when the tension and excitement levels in the sentiment indicators are high, exclamation words or intensifying words can be added (e.g., "Critical moment!", "This is too dangerous!"); when the sentiment indicators are low, more stable narrative sentences are generated, and the computer device can set a sentence diversity control mechanism to avoid reusing the same template. This mechanism can maintain linguistic diversity and natural fluency in continuous broadcast content through random sampling strategies, semantic similarity detection, or sentence repetition rate control parameters.
[0044] The computer device integrates a virtual esports commentator 41 into the match streaming screen 40. The computer device controls the virtual esports commentator 41 to perform voice broadcasts and facial expression rendering of the real-time broadcast content 30 based on the emotions corresponding to the emotion indicators. Specifically, the tone, speed, and volume of the virtual esports commentator 41's voice broadcasts of the real-time broadcast content 30 are adjusted based on the emotions corresponding to the emotion indicators, thus presenting different emotional rendering effects to provide esports event broadcasts. Please refer to the diagram illustrating the integration of the virtual esports commentator 41 into the match streaming screen 40. Figure 4 As shown, Figure 4 The illustration shows a schematic diagram of the integrated virtual e-sports anchor broadcasting of the present invention, which is based on real-time game data analysis and context awareness.
[0045] Embodiments of this application also provide a computer device, the computer device comprising: The storage device stores multiple computer-readable instructions; and one or more hardware processors are electrically connected to the storage device to execute the multiple computer-readable instructions, enabling the computer device to implement the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described above.
[0046] Embodiments of this application also provide a computer-readable recording medium storing a computer program that, when executed by one or more hardware processors of a computer device, causes the computer device to perform a virtual e-sports broadcasting method based on real-time game data analysis and context awareness. This computer-readable recording medium may be included in the computer device described in the above embodiments, or it may exist independently and not be assembled into the computer device.
[0047] Please refer to Figure 5 As shown, Figure 5 The diagram illustrates the computer equipment system architecture for virtual esports broadcasting based on real-time game data analysis and context awareness, as per the present invention. It should be noted that... Figure 5 The computer system 500 of the computer device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.
[0048] like Figure 5 As shown, the computer system 500 includes a Central Processing Unit (CPU) 501, which can perform various appropriate actions and processes, such as executing the methods described in the above embodiments, based on programs stored in Read-Only Memory (ROM) 502 or programs loaded from storage portion 508 into Random Access Memory (RAM) 503. The RAM 503 also stores various programs and data required for system operation. The CPU 501, ROM 502, and RAM 503 are interconnected via a bus 504. An Input / Output (I / O) interface 505 is also connected to the bus 504.
[0049] The following components are connected to the I / O interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), and speakers, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as needed. Removable media 511, such as magnetic disks, optical disks, magneto-optical disks, semiconductor memory, etc., are installed on the drive 510 as needed so that computer programs read from them can be installed into the storage section 508 as needed.
[0050] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 509, and / or installed from removable medium 511. When the computer program is executed by central processing unit (CPU) 501, it performs various functions defined in the system of the present invention.
[0051] It should be noted that the computer-readable medium shown in the embodiments of the present invention can be a computer-readable signal medium or a computer-readable recording medium, or any combination thereof. A computer-readable recording medium can be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable recording medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), electronically erasable programmable read-only memory (EPROM), flash memory, optical fiber, compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable signal medium may include a data signal propagated in a baseband frequency or as part of a carrier wave, wherein a computer-readable computer program is carried. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable recording media, which can send, propagate, or transmit programs for use by or in connection with an instruction execution system, apparatus, or device. Computer programs contained on a computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination thereof.
[0052] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0053] The units described in the embodiments of the present invention can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the specific unit itself. Therefore, the technical solutions according to the embodiments of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, portable hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, touch terminal, or network device, etc.) to execute the method according to the embodiments of the present invention.
[0054] In summary, by integrating multiple layers of technologies, including real-time game data acquisition, character behavior feature analysis, action event prediction, tactical intent inference, deep context model emotion calculation, and natural language generation, it is possible not only to identify the behavior of a single character but also to understand the collaborative tactical relationships of multiple characters. Furthermore, by combining deep context models to generate temporally continuous emotion indicators, the voice broadcasts and facial expressions of virtual e-sports anchors can be dynamically adjusted according to the event context, enhancing the realism and immersiveness of the broadcasts.
[0055] This technology can solve the problems existing in current technologies, such as the lack of ability to automatically determine the viewer's perspective position based on real-time game data, accurately analyze tactical intentions and generate professional broadcast content by combining emotional indicators, and achieve virtual anchor emotional rendering effects with temporal continuity and tactical relevance. In this way, it can improve the ability to capture key tactical events, optimize the continuity of virtual anchor emotional expression, and enhance the professionalism of real-time broadcast semantics.
[0056] While the embodiments disclosed in this invention are as described above, the content is not intended to directly limit the scope of patent protection of this invention. Anyone skilled in the art to which this invention pertains may make minor modifications in form and detail without departing from the spirit and scope disclosed herein. The scope of patent protection of this invention shall still be determined by the scope defined in the appended claims.
Claims
1. A method for virtual esports broadcasting based on real-time game data analysis and context awareness, comprising the following steps performed via computer equipment: The game application interface obtains real-time game data and the observer's game perspective from the competitive game server. Based on the real-time game data of the corresponding character selected from the observer's game perspective screen, action events and tactical intentions are analyzed from the character's status, position, movement route, and action instructions in the real-time game data. The real-time game data of the selected characters, the parsed action events, and the tactical intentions are used to obtain emotion indicators using a deep contextual model. Natural language processing technology is used to combine the tactical intentions and the real-time game data of the selected characters with a competitive terminology database to generate real-time broadcast content. Integrate virtual esports anchors into the match streaming footage; and The virtual esports anchor is controlled to broadcast the real-time news content with voice and facial expressions based on the emotions corresponding to the emotion indicators, so as to provide esports event broadcasts.
2. The virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in claim 1, wherein the character state, position, and movement route of each character in the real-time game data are determined using an action event prediction model to determine the viewpoint display position of the observer's game view.
3. The virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in claim 1, wherein when the position of each character in the real-time game data is close to a pre-set important resource position or when the position of more than or equal to a preset number of game characters in the real-time game data is close to a pre-set important resource position, the viewpoint position of the observer's game view screen is set to the important resource position.
4. The virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in claim 1, wherein the deep context model is trained using the real-time game data pre-labeled with corresponding emotion scores and the action events pre-labeled with corresponding emotion scores.
5. The virtual esports anchor broadcasting method based on real-time game data analysis and context awareness as described in claim 1, wherein the emotion index is calculated cumulatively over time within the time range of the action event, so that the emotion of the virtual esports anchor's esports event broadcast has the effect of continuous change with time correlation.
6. The virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in claim 1, wherein the tone, speed and volume of the real-time broadcast content are adjusted based on the emotion corresponding to the emotion index to present different emotion rendering effects.
7. A computer device, the computer device comprising, wherein the following steps are included: Storage device, storing multiple computer-readable instructions; and One or more hardware processors, electrically connected to the storage device, execute the plurality of computer-readable instructions to enable the virtual e-sports anchor broadcasting device based on real-time game data analysis and context awareness to implement the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in any one of claims 1 to 6.
8. A computer-readable recording medium having a computer program stored thereon, which, when executed by one or more hardware processors of a computer device, causes the computer device to perform the virtual e-sports anchor broadcasting method based on real-time game data analysis and context awareness as described in any one of claims 1 to 6.
