An e-sports low-latency earphone audio processing method and system

Abstract

The application relates to the technical field of audio processing, and provides an e-sports low-delay earphone audio processing method and system.The method comprises the following steps: performing abnormality identification on a sound source according to current audio data and scene restriction information to obtain an abnormality identification result; updating the current audio confidence of the sound source according to the abnormality identification result to obtain updated current audio confidence; when the abnormality identification result indicates that the sound source has an abnormality and a target abnormality type of the sound source, judging whether the updated current audio confidence is greater than a target confidence threshold; and when the updated current audio confidence is greater than the target confidence threshold, correcting the current audio data according to the target abnormality type of the sound source to obtain target audio data. Through the scheme, the accuracy of e-sports earphone audio processing can be improved.

Description

Technical Field

[0001] This application relates to the field of audio processing technology, and in particular to an audio processing method and system for low-latency gaming headsets. Background Technology

[0002] In professional esports training facilities or high-end digital entertainment environments, users typically wear specially designed gaming headsets. These headsets are designed to provide an immersive audio experience, which is crucial for competitive gaming because sound cues (such as footsteps, gunshots, or skill activation sounds) provide key tactical information. The main computing device generates complex multi-channel audio data, which needs to be processed and rendered into a three-dimensional sound field. Traditionally, audio processing tasks, including decoding and spatialization, are handled by the main computing device's central processing unit (CPU) or dedicated sound card. However, with the increasing demands of game graphics processing, the CPU often becomes a bottleneck, leading to increased audio processing latency. This latency directly impacts a player's immediate reaction time and overall gaming performance. To mitigate this issue, modern low-latency gaming headsets typically integrate dedicated hardware processing units within the headset. These units offload audio decoding and spatial rendering tasks from the main computing device, ensuring that audio signals can be processed independently and delivered with extremely low latency, thereby improving the user's responsiveness and immersion.

[0003] However, in such a complex, multi-source, networked, and dynamically fused environment, the integrity of the "spatial information" in audio data becomes crucial, introducing a key element of "unexpectedness." Imagine that, due to a transient network anomaly, a subtle software flaw in the simulation engine, or even a momentary malfunction in the source processing, an incoming audio packet for a critical sound cue (e.g., the sound of enemy footsteps or a distant vehicle) may appear "complete" in terms of data integrity, but it could contain subtle, incorrect spatial metadata. For example, the perceived location might deviate from the actual location. This isn't a complete audio interruption or obvious stutter; the sound is still heard, but its spatial origin or characteristics are subtly misleading.

[0004] Headphone processing units, designed to faithfully render received spatial information, process and present subtle errors in the data to the user. Existing low-latency processing workflows, focused on speed and rendering fidelity based on "assumed correct" input, lack the ability to detect and correct subtle semantic errors in these spatial cues in real-time, especially without external real-world references. This problem is particularly insidious because the user receives "some" information, but this information is subtly "incorrect," leading to incorrect judgments without the user being aware of the source of the error. Therefore, headphones not only need to process audio quickly and smoothly, but also need to ensure the accuracy of the spatial information they convey, even when the source data may contain subtle flaws. Summary of the Invention

[0005] This application provides an audio processing method and system for low-latency gaming headsets, aiming to solve the technical problem that existing low-latency audio processing flows cannot effectively detect and correct subtle spatial information errors in audio data in real-time, leading to incorrect judgments by users.

[0006] The technical solution of this application is as follows: In a first aspect, this application discloses an audio processing method for low-latency gaming headsets, comprising: acquiring current audio data of a sound source in a virtual scene, scene constraint information in the virtual scene, and current audio confidence of the sound source; the scene constraint information includes physical constraint information of the sound source and scene terrain information of the virtual scene; performing anomaly identification on the sound source based on the current audio data and scene constraint information to obtain anomaly identification results; the anomaly identification results are used to indicate that the sound source does not have anomalies, or to indicate that the sound source has anomalies, and the target anomaly type of the sound source; the target anomaly type includes sound source velocity anomaly and / or sound source position anomaly; updating the current audio confidence of the sound source based on the anomaly identification results to obtain updated current audio confidence; when the anomaly identification results indicate that the sound source has anomalies and the target anomaly type of the sound source, determining whether the updated current audio confidence is greater than a target confidence threshold; when the updated current audio confidence is greater than the target confidence threshold, correcting the current audio data according to the target anomaly type of the sound source to obtain target audio data.

[0007] Furthermore, based on the current audio data and scene constraint information, anomaly identification is performed on the sound source to obtain anomaly identification results, including: determining whether the sound source has a velocity anomaly based on the current audio data and the physical constraint information of the sound source; if the sound source has a velocity anomaly, determining that the anomaly identification result indicates that the sound source has an anomaly, and that the target anomaly type of the sound source includes a velocity anomaly; determining whether the sound source has a positional anomaly based on the current audio data and the scene terrain information of the virtual scene; if the sound source has a positional anomaly, determining that the anomaly identification result indicates that the sound source has an anomaly, and that the target anomaly type of the sound source includes a positional anomaly; if the sound source does not have a positional anomaly and the sound source does not have a velocity anomaly, determining that the anomaly identification result indicates that the sound source does not have an anomaly.

[0008] More specifically, in some implementation schemes, the physical constraint information of the sound source includes the maximum moving speed of the sound source, and the current audio data includes the current position information of the sound source and the time when the audio data was generated. Based on the current audio data and the physical constraint information of the sound source, it is determined whether the sound source has a speed anomaly, including: obtaining first position information; the first position information is the position information of the sound source in the first audio data, and the first audio data is the audio data of the previous frame of the current audio data; determining a first displacement distance between the first position information and the current position information; using the ratio of the first displacement distance to a first duration as the calculation speed of the sound source; the first duration is the duration between the time when the current audio data was generated and the time when the first audio data was generated; determining that the sound source has a speed anomaly when the calculation speed of the sound source is greater than the maximum moving speed of the sound source; and determining that the sound source does not have a speed anomaly when the calculation speed of the sound source is less than or equal to the maximum moving speed of the sound source.

[0009] Based on this, according to the current audio data and the scene terrain information of the virtual scene, it is determined whether there is a location anomaly of the sound source, including: according to the scene terrain information of the virtual scene, it is determined whether there is scene terrain between the first location information of the sound source and the current location information of the sound source; if there is scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that there is a location anomaly of the sound source; if there is no scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that there is no location anomaly of the sound source.

[0010] As a technical improvement, the current audio confidence of the sound source is updated based on the anomaly identification result to obtain the updated current audio confidence. This includes: obtaining a first correspondence; the first correspondence includes a one-to-one correspondence between multiple current audio confidence ranges and multiple first coefficients; using the first coefficient corresponding to the current audio confidence range in which the sound source's current audio confidence is located in the first correspondence as a second coefficient; and updating the current audio confidence of the sound source based on the second coefficient, the anomaly identification result, and a preset confidence adjustment step size to obtain the updated current audio confidence.

[0011] Based on the above, the current audio confidence of the sound source is updated according to the second coefficient, the anomaly identification result, and the preset confidence adjustment step size, to obtain the updated current audio confidence. This includes: when the anomaly identification result indicates that the sound source has an anomaly and the target anomaly type of the sound source is specified, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size; the product of the initial confidence adjustment step size and the number of target anomaly types is used as the target confidence adjustment step size; and the difference between the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

[0012] As an optional approach, the current audio confidence of the sound source is updated based on the second coefficient, the anomaly identification result, and the preset confidence adjustment step size to obtain the updated current audio confidence. This includes: when the anomaly identification result indicates that there is no anomaly in the sound source, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size; the product of the initial confidence adjustment step size and 2 is used as the target confidence adjustment step size; and the sum of the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

[0013] To further address the issue, the method determines whether the updated current audio confidence level is greater than the target confidence level threshold. This involves: obtaining the preset game influence index of the sound source; using the product of the preset game influence index and the preset threshold as the target confidence level threshold; and determining whether the updated audio confidence level of the sound source is greater than the target confidence level threshold.

[0014] Based on the above, the audio data also includes the sound source's movement speed. When the updated current audio confidence score is greater than the target confidence score threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain target audio data. This includes: when the target anomaly type of the sound source is a positional anomaly, the endpoint of the intersection line segment closest to the first position information is used as the corrected position information in the current audio data to obtain the target audio data; the intersection line segment is the line segment in the target line segment that overlaps with the scene terrain, and the target line segment is the line connecting the first position information and the current position information; when the target anomaly type of the sound source is a speed anomaly, the intersection position of the target circle and the target line segment is used as... The target audio data is obtained by using the corrected position information from the current audio data. The center of the target circle is the first position information, and the radius of the target circle is the product of the sound source's movement speed and the first duration in the first audio data. When the target anomaly type of the sound source includes position anomaly and speed anomaly, it is determined whether there is an intersecting line segment within the target circle. If there is an intersecting line segment within the target circle, the endpoint of the intersecting line segment closest to the first position information is used as the corrected position information in the current audio data to obtain the target audio data. If there is no intersecting line segment within the target circle, the intersection position of the target circle and the target line segment is used as the corrected position information in the current audio data to obtain the target audio data.

[0015] Secondly, this application also discloses an audio processing system for low-latency gaming headsets, comprising: an acquisition device and a processing device; the acquisition device is used to acquire current audio data of a sound source in a virtual scene, scene constraint information in the virtual scene, and current audio confidence of the sound source; the scene constraint information includes physical constraint information of the sound source and scene terrain information of the virtual scene; the processing device is used to perform anomaly identification on the sound source based on the current audio data and scene constraint information, and obtain anomaly identification results; the anomaly identification results are used to indicate that the sound source does not have anomalies, or to indicate that the sound source has anomalies, and the target of the sound source. Anomaly type; target anomaly types include sound source velocity anomaly and / or sound source location anomaly; the processing device is further configured to update the current audio confidence of the sound source based on the anomaly identification result to obtain the updated current audio confidence; the processing device is further configured to determine whether the updated current audio confidence is greater than the target confidence threshold when the anomaly identification result indicates that the sound source has an anomaly and the target anomaly type of the sound source; the processing device is further configured to correct the current audio data according to the target anomaly type of the sound source when the updated current audio confidence is greater than the target confidence threshold to obtain the target audio data.

[0016] Beneficial Effects: The low-latency audio processing method for e-sports headsets disclosed in this application acquires the current audio data of a sound source in a virtual scene, scene constraint information, and the current audio confidence level of the sound source. Based on this information, it performs anomaly identification on the sound source to obtain anomaly identification results. When the anomaly identification result indicates that the sound source is abnormal and the updated current audio confidence level is greater than the target confidence level threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain the target audio data. This method effectively solves the problem in existing low-latency audio processing procedures that cannot detect and correct subtle spatial information errors in audio data in real time.

[0017] By introducing anomaly detection and confidence update mechanisms, this application can proactively identify and correct potential speed or positional anomalies in audio data, preventing erroneous spatial metadata caused by transient network anomalies, software defects, or source-end processing failures from being directly rendered to users. This allows esports players to receive more accurate and reliable audio cues, enabling them to make more precise judgments and reactions, significantly improving the gaming experience and competitive performance. Compared to existing technologies that focus solely on speed and rendering fidelity while neglecting data semantic accuracy, this application, while ensuring low latency, greatly enhances the accuracy of audio spatial information, overcoming the hidden problem of users receiving "subtle errors" that could lead to misjudgments, demonstrating significant and superior technical effects. Attached Figure Description

[0018] Figure 1A flowchart illustrating an audio processing method for a low-latency gaming headset provided in this application; Figure 2 A flowchart illustrating another method for audio processing of low-latency gaming headsets provided in this application; Figure 3 This application provides an architectural diagram of a low-latency audio processing system for e-sports headsets. Detailed Implementation

[0019] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0020] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] In professional esports training facilities or high-end digital entertainment environments, users typically wear specially designed gaming headsets. These headsets are designed to provide an immersive audio experience, which is crucial for competitive gaming because sound cues (such as footsteps, gunshots, or skill activation sounds) provide key tactical information. The main computing device generates complex multi-channel audio data, which needs to be processed and rendered into a three-dimensional sound field. Traditionally, audio processing tasks, including decoding and spatialization, are handled by the main computing device's central processing unit (CPU) or dedicated sound card. However, with the increasing demands of game graphics processing, the CPU often becomes a bottleneck, leading to increased audio processing latency. This latency directly impacts a player's immediate reaction time and overall gaming performance. To mitigate this issue, modern low-latency gaming headsets typically integrate dedicated hardware processing units within the headset. These units offload audio decoding and spatial rendering tasks from the main computing device, ensuring that audio signals can be processed independently and delivered with extremely low latency, thereby improving the user's responsiveness and immersion.

[0022] However, in such a complex, multi-source, networked, and dynamically fused environment, the integrity of the "spatial information" in audio data becomes crucial, introducing a key element of "unexpectedness." Imagine that, due to a transient network anomaly, a subtle software flaw in the simulation engine, or even a momentary malfunction in the source processing, an incoming audio packet for a critical sound cue (e.g., the sound of enemy footsteps or a distant vehicle) may appear "complete" in terms of data integrity, but it could contain subtle, incorrect spatial metadata. For example, the perceived location might deviate from the actual location. This isn't a complete audio interruption or obvious stutter; the sound is still heard, but its spatial origin or characteristics are subtly misleading.

[0023] Headphone processing units, designed to faithfully render received spatial information, process and present subtle errors in the data to the user. Existing low-latency processing workflows, focused on speed and rendering fidelity based on "assumed correct" input, lack the ability to detect and correct subtle semantic errors in these spatial cues in real-time, especially without external real-world references. This problem is particularly insidious because the user receives "some" information, but this information is subtly "incorrect," leading to incorrect judgments without the user being aware of the source of the error. Therefore, headphones not only need to process audio quickly and smoothly, but also need to ensure the accuracy of the spatial information they convey, even when the source data may contain subtle flaws.

[0024] In this regard, such as Figure 1 As shown, this application proposes an audio processing method for low-latency gaming headsets, including: S101. Obtain the current audio data of the sound source in the virtual scene, the scene constraint information in the virtual scene, and the current audio confidence of the sound source; the scene constraint information includes the physical constraint information of the sound source and the scene terrain information of the virtual scene.

[0025] S102. Based on the current audio data and scene limitation information, perform anomaly identification on the sound source and obtain the anomaly identification result.

[0026] The anomaly identification result is used to indicate whether the sound source is not abnormal, or to indicate whether the sound source is abnormal, and the target anomaly type of the sound source; the target anomaly type includes sound source velocity anomaly and / or sound source location anomaly.

[0027] S103. Update the current audio confidence of the sound source based on the anomaly identification result to obtain the updated current audio confidence.

[0028] S104. When the anomaly identification result indicates that there is an anomaly in the sound source and the target anomaly type of the sound source, determine whether the updated current audio confidence is greater than the target confidence threshold.

[0029] S105. When the updated current audio confidence score is greater than the target confidence score threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain the target audio data.

[0030] This application introduces an anomaly detection mechanism and a confidence update strategy, which can detect and correct subtle errors in audio data, thereby ensuring the accuracy of spatial information in the audio output of gaming headsets and significantly improving the player's gaming experience and competitive performance.

[0031] To better understand the audio processing method for low-latency gaming headsets proposed in this application, some key terms involved will be explained first.

[0032] "Current audio data of a sound source" refers to the audio information generated by a specific sound source at a certain moment in a virtual scene. It usually includes spatial and attribute data such as the location, speed, volume, and timbre of the sound source.

[0033] "Scene constraint information" refers to the constraints in a virtual scene that affect the behavior or propagation of a sound source. It includes "physical constraint information of the sound source" and "scene terrain information of the virtual scene." "Physical constraint information of the sound source" defines the kinematic or dynamic limitations of the sound source in the virtual world, such as maximum movement speed and acceleration limits. "Scene terrain information of the virtual scene" describes the geometric structure and physical properties of the virtual environment, such as walls, obstacles, and elevation differences. This information affects the visibility, accessibility, and sound propagation path of the sound source.

[0034] "Current audio confidence of the sound source" indicates the degree of reliability of the currently received sound source audio data in terms of spatial information. The higher the value, the more reliable the data.

[0035] "Anomaly identification result" is a judgment made by the system after analyzing the audio data of the sound source. It is used to indicate whether the sound source has "abnormal sound source velocity" or "abnormal sound source location" or other "target anomaly types".

[0036] The "target confidence threshold" is a preset reference value used to determine whether the updated current audio confidence is high enough to decide whether to correct abnormal audio data.

[0037] The core of the low-latency audio processing method for e-sports headphones proposed in this application lies in the intelligent anomaly identification and correction of audio data of sound sources in virtual scenes, so as to ensure the accuracy of audio spatial information.

[0038] First, this method requires acquiring the current audio data of sound sources in the virtual scene, scene constraint information, and the current audio confidence level of the sound sources. For example, an esports game engine generates audio data for various sound sources in the game (such as enemy characters, explosions, and environmental sound effects) in real time. This data may include information such as the coordinates, movement speed, and volume of the sound sources in three-dimensional space. Simultaneously, the system acquires the constraint information of the game scene, including physical constraints such as the maximum movement speed of the sound sources, as well as scene terrain information such as walls, obstacles, and terrain undulations. Furthermore, the system maintains the current audio confidence level of a sound source, which reflects the reliability of the current audio data in terms of spatial information. This information can be acquired in various ways, such as directly reading it through the game engine's API interface or parsing it by listening to the game data stream.

[0039] Secondly, based on the acquired current audio data and scene constraint information, anomaly identification is performed on the sound source to obtain anomaly identification results. These results indicate whether the sound source is not abnormal, or whether an anomaly exists, and specify the target anomaly type. Target anomaly types can include sound source velocity anomalies and / or sound source position anomalies. For example, the system can calculate the actual movement velocity of the sound source based on its current position and the position in the previous frame, and compare it with the maximum movement velocity defined in the sound source's physical constraint information. If the calculated velocity exceeds the maximum movement velocity, the system can determine that the sound source has a velocity anomaly. As another example, the system can determine whether the sound source has penetrated an impassable obstacle or terrain based on its current position and scene terrain information. If penetration occurs, the system can determine that the sound source has a positional anomaly. Anomaly identification can be implemented in various ways; for example, rule-based judgment or machine learning models can be used for anomaly pattern recognition.

[0040] Next, the current audio confidence level of the sound source is updated based on the anomaly identification results, resulting in the updated current audio confidence level. When the anomaly identification results indicate that the sound source is abnormal, its confidence level is usually lowered; conversely, if the sound source is not abnormal, its confidence level can be appropriately increased. For example, a confidence level adjustment step size can be preset. When an abnormal sound source velocity is detected, the current audio confidence level is subtracted by a preset value; when an abnormal sound source position is detected, another preset value is subtracted. If both types of anomalies exist simultaneously, the sum of the two preset values ​​is subtracted. If the sound source is not abnormal, the current audio confidence level is added to a preset value. This confidence level update mechanism can dynamically reflect the reliability of the audio data, providing a basis for subsequent correction decisions.

[0041] When the anomaly identification result indicates an anomaly in the sound source and identifies the target anomaly type, the system determines whether the updated current audio confidence score is greater than a target confidence score threshold. This target confidence score threshold can be set according to the needs of the actual application scenario. For example, a fixed threshold such as 0.7 or 0.8 can be set. If the updated current audio confidence score is lower than this threshold, the anomalous audio data is considered unreliable, and it is discarded or processed using default settings without correction. If it is higher than this threshold, the anomalous audio data is considered to have a certain degree of credibility, although it is abnormal, and correction can be attempted.

[0042] Finally, when the updated current audio confidence score exceeds the target confidence score threshold, the current audio data is corrected based on the target anomaly type of the sound source to obtain the target audio data. The purpose of this correction is to restore the accuracy of the spatial information as much as possible while maintaining the reasonableness of the audio data. For example, if the target anomaly type is a sound source velocity anomaly, the system can correct the sound source's movement speed to within the maximum allowable speed based on the sound source's maximum movement speed and recalculate the sound source's position. If the target anomaly type is a sound source position anomaly, the system can correct the sound source's position to the nearest reachable area based on scene terrain information to avoid penetrating obstacles. The corrected audio data is the target audio data, which will be used for subsequent audio rendering and output.

[0043] The low-latency audio processing method for esports headsets proposed in this application works by establishing a closed-loop audio data quality assurance mechanism. Traditional esports headset audio processing primarily focuses on the rapid transmission and rendering of audio data, but lacks effective detection and correction capabilities for subtle spatial information errors that may exist within the audio data. This can lead to players receiving seemingly complete but actually misleading audio cues in competitive games, resulting in incorrect judgments and impacting their gaming experience and competitive performance.

[0044] This application introduces an anomaly detection module, which can monitor the current audio data of a sound source in real time and, in conjunction with scene constraint information of the virtual scene, actively identify possible speed or positional anomalies of the sound source. For example, when a sound source moves an excessive distance in a very short time (speed anomaly), or when the sound source appears inside an impassable obstacle (positional anomaly), the system can promptly detect these phenomena that do not conform to physical or logical laws.

[0045] Once an anomaly is identified, the system does not immediately correct it. Instead, it first updates the current audio confidence level of the sound source based on the anomaly identification result. This confidence level mechanism is one of the key innovations of this application. It allows the system to quantitatively assess the severity of the anomaly and the reliability of the data. When an anomaly is detected, the confidence level decreases accordingly; conversely, when the data appears normal, the confidence level increases. This dynamically adjusted confidence level avoids a one-size-fits-all approach to all anomalous data, instead classifying and processing data according to its reliability.

[0046] The system then determines whether the updated audio confidence level exceeds the target confidence threshold. This step is crucial, as it decides whether to correct the anomalous data. If the confidence level is too low, even if anomalies exist, it may indicate that the original data is severely flawed, posing a high risk of correction. In this case, the system may choose to discard the data or adopt other conservative strategies. Conversely, when the confidence level is above the threshold, it indicates that although anomalies exist, the data still has some reference value and is worth correcting.

[0047] Finally, provided the confidence level is met, the system will make targeted corrections to the current audio data based on the type of anomaly in the sound source. For example, for speed anomalies, the system can reasonably limit the speed and calculate the location based on the maximum moving speed of the sound source; for location anomalies, the system can adjust the sound source location to the nearest legal area based on scene terrain information. In this way, this application can effectively correct subtle spatial information errors in audio data and generate more accurate target audio data.

[0048] Therefore, the overall technical solution of this application forms an intelligent closed loop for audio data processing: from acquiring raw data, to anomaly identification, to confidence assessment, and finally to conditional correction. This process ensures that the audio output from the gaming headset not only has low latency but also high accuracy in spatial information, thereby significantly improving the player's sense of direction, distance, and overall immersion in competitive games, and solving the problem of misjudgment caused by errors in the spatial information of audio data in existing technologies.

[0049] The low-latency audio processing method for esports headsets proposed in this application demonstrates significant progress and innovation in solving the problem of spatial information accuracy in audio data compared to existing technologies. Traditional low-latency audio processing workflows primarily focus on rapid decoding and spatial rendering of audio signals, with the core assumption being that the input audio data is completely accurate in spatial information. However, as described in the background section, in complex esports environments, due to various transient factors, subtle and misleading spatial metadata may exist in the audio data, and existing technologies lack effective detection and correction mechanisms for this.

[0050] The core innovation of this application lies in the introduction of mechanisms for "anomaly recognition" and "confidence updating and correction." Existing technologies, when receiving audio data containing erroneous spatial information, faithfully render it according to these errors, causing players to receive incorrect sound cues. For example, an enemy's footsteps might be rendered in the wrong direction or at the wrong distance, misleading the player's judgment. This application, however, proactively detects anomalies that do not conform to the physical laws or scene limitations of the virtual world by performing real-time anomaly recognition of the speed and location of sound sources. For example, if a sound source moves an impossible distance in a short period of time, or penetrates a wall in the virtual scene, this application can immediately mark it as an anomaly.

[0051] Furthermore, this application introduces the concept of "current audio confidence of the sound source" and dynamically updates it based on anomaly identification results. This enables the system to quantitatively assess the reliability of audio data. In existing technologies, all received audio data is considered equally reliable, lacking an inherent assessment of data quality. The confidence mechanism of this application allows the system to lower the confidence level when an anomaly is detected and raise the confidence level when the data is normal, thus providing a more refined basis for subsequent correction decisions. This mechanism avoids blindly correcting all anomalous data, instead performing conditional corrections based on its confidence level, improving the accuracy and robustness of the correction.

[0052] Furthermore, this application corrects the current audio data based on the target anomaly type of the sound source when the updated current audio confidence score exceeds the target confidence score threshold. This conditional and targeted correction strategy is not available in existing technologies. Existing technologies either completely trust the original data or directly discard the data when a serious error is detected, lacking an intermediate intelligent correction step. The correction mechanism of this application can use different correction algorithms according to the specific anomaly type (such as speed anomaly or position anomaly) to adjust the anomalous data to a reasonable range that conforms to the logic of the virtual scene. For example, for speed anomalies, the movement speed of the sound source can be limited; for position anomalies, its position can be adjusted to the nearest legal area.

[0053] In summary, this application effectively addresses the limitations of existing technologies in processing audio data containing subtle spatial information errors by constructing an audio processing method for low-latency e-sports headsets that integrates anomaly identification, confidence assessment, and intelligent correction. It not only ensures low-latency audio transmission but, more importantly, significantly improves the accuracy of audio spatial information, thereby providing players with more realistic and reliable sound cues. This greatly enhances the fairness and immersion of competitive games, demonstrating significant technological progress and innovative value.

[0054] Specifically, the steps described above for identifying sound sources based on current audio data and scene constraints to obtain anomaly identification results include: Based on the current audio data and the physical constraints of the sound source, determine whether the sound source has a velocity anomaly; if the sound source has a velocity anomaly, determine that the anomaly identification result indicates that the sound source has an anomaly, and that the target anomaly type of the sound source includes a velocity anomaly; based on the current audio data and the scene terrain information of the virtual scene, determine whether the sound source has a position anomaly; if the sound source has a position anomaly, determine that the anomaly identification result indicates that the sound source has an anomaly, and that the target anomaly type of the sound source includes a position anomaly; if the sound source does not have a position anomaly and does not have a velocity anomaly, determine that the anomaly identification result indicates that the sound source does not have an anomaly.

[0055] The physical constraints of a sound source can be understood as limitations on its movement within the virtual scene, such as its maximum speed. By comparing the sound source's motion state (e.g., speed) reflected in the current audio data with these physical constraints, it can be determined whether the sound source's movement exceeds a reasonable range, thus identifying speed anomalies. Furthermore, the scene terrain information refers to various obstacles, boundaries, or impassable areas existing in the virtual environment. By analyzing the sound source's location information and the scene terrain information in the current audio data, it can be determined whether the sound source is in an unreasonable position, such as passing through walls or clipping through objects, thus identifying location anomalies. When a speed anomaly is detected, the anomaly identification result will be determined as indicating an anomaly in the sound source, and its target anomaly type will be clearly defined as a sound source speed anomaly. Similarly, when a location anomaly is detected, the anomaly identification result will be determined as indicating an anomaly in the sound source, and its target anomaly type will be clearly defined as a sound source location anomaly. If the sound source has neither speed nor location anomalies, the anomaly identification result will indicate that the sound source is not abnormal.

[0056] The proposed solution refines the anomaly identification process into independent judgments of sound source velocity anomalies and location anomalies, enabling more precise localization of potential abnormal sound source behavior. By separately utilizing the physical constraint information of the sound source and the scene terrain information of the virtual scene to verify the sound source's motion velocity and spatial location, the comprehensiveness and accuracy of anomaly identification are ensured. This step-by-step judgment approach allows the system to address different types of anomalies specifically, improving the efficiency and reliability of anomaly identification.

[0057] The aforementioned technical solution enables refined identification of sound source anomalies in virtual scenes, distinguishing between abnormal sound source speed, abnormal sound source location, or neither. This clear anomaly type identification facilitates targeted correction of audio data, avoiding miscorrection or insufficient correction that might result from general anomaly assessments. This improves the accuracy of audio processing and user experience, especially in esports scenarios, effectively reducing auditory misjudgments caused by abnormal sound sources and enhancing game fairness and immersion.

[0058] This application further proposes to refine and optimize the method for judging abnormal sound source velocity, so as to improve the accuracy and reliability of abnormality identification.

[0059] The physical constraint information of the sound source includes the maximum moving speed of the sound source. The current audio data includes the current position information of the sound source and the time when the audio data was generated. Based on the current audio data and the physical constraint information of the sound source, it is determined whether there is an abnormal speed of the sound source, including: Obtain first position information; the first position information is the position information of the sound source in the first audio data, and the first audio data is the audio data of the previous frame of the current audio data; determine the first displacement distance between the first position information and the current position information; use the ratio of the first displacement distance to the first duration as the calculation speed of the sound source; the first duration is the duration between the generation time of the current audio data and the generation time of the first audio data; when the calculation speed of the sound source is greater than the maximum moving speed of the sound source, determine that the sound source has a speed anomaly; when the calculation speed of the sound source is less than or equal to the maximum moving speed of the sound source, determine that the sound source does not have a speed anomaly.

[0060] Specifically, the physical constraints of the sound source are further defined as including its maximum movement speed. This maximum movement speed refers to the highest movement speed a specific sound source can achieve under normal circumstances in a virtual scene, and its setting aims to provide a baseline upper limit for judging speed anomalies. The current audio data is explicitly defined as including the current location information of the sound source and the time when the audio data was generated, which provides the necessary time and space coordinates for subsequent speed calculations.

[0061] In this context, obtaining the first position information refers to the system extracting the sound source's position information from the previous frame of audio data. This previous frame of audio data is called the first audio data, and the sound source position information it contains is the first position information. By obtaining the position information from the previous moment, it can be compared with the position information at the current moment, thereby calculating the displacement of the sound source between the two moments.

[0062] Furthermore, a first displacement distance between the first location information and the current location information is determined. This distance can be obtained by calculating the Euclidean distance between the two location points or other suitable distance metrics, representing the actual distance the sound source moves within the first time duration.

[0063] Subsequently, the ratio of the first displacement distance to the first duration is used as the calculated speed of the sound source. The first duration refers to the time interval between the generation time of the current audio data and the generation time of the first audio data. In this way, the average moving speed of the sound source between two consecutive frames can be accurately calculated.

[0064] Finally, the presence of speed anomalies in the sound source is determined by comparing its calculation speed with its maximum movement speed. Specifically, if the sound source's calculation speed exceeds the preset maximum movement speed, a speed anomaly is identified, indicating that the sound source's movement speed exceeds its reasonable range within the virtual scene. Conversely, if the sound source's calculation speed is less than or equal to its maximum movement speed, no speed anomaly is identified, and the sound source's movement speed is considered to be within the normal range.

[0065] This application's solution introduces the maximum moving speed of the sound source as physical constraint information, and combines this with the positional information and generation time from two consecutive frames of audio data to accurately calculate the actual moving speed of the sound source within a specific time interval. It is precisely this calculation method based on historical data and physical constraints that enables the system to quantitatively evaluate the moving speed of the sound source. By comparing the calculated speed with the preset maximum moving speed, sound sources that exceed normal motion patterns can be effectively identified, thereby avoiding audio processing errors or delays caused by abnormal sound source data. This mechanism ensures the objectivity and accuracy of anomaly identification, providing a reliable basis for subsequent audio data correction.

[0066] Through the above technical solution, this application provides a more accurate and reliable method for identifying sound source velocity anomalies. By introducing the maximum moving speed of the sound source as the judgment criterion and calculating the actual moving speed of the sound source using audio data from consecutive frames, misjudgments caused by sudden changes in sound source data or transmission errors can be effectively avoided. This solution significantly improves the accuracy of anomaly identification, reduces unnecessary audio correction operations, thereby ensuring the smoothness of audio processing and user experience of low-latency gaming headsets, while also reducing system resource consumption.

[0067] Specifically, such as Figure 2 As shown, according to the above-mentioned low-latency audio processing method for e-sports headphones, based on the current audio data and the scene terrain information of the virtual scene, it is determined whether there is an abnormal location of the sound source, including: S201. Based on the scene terrain information of the virtual scene, determine whether there is scene terrain between the first position information of the sound source and the current position information of the sound source.

[0068] S202. If there is a scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that there is a location anomaly of the sound source.

[0069] S203. If there is no scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that the sound source does not have a location anomaly.

[0070] In this context, scene terrain information refers to pre-defined, impassable or difficult-to-cross obstacles, terrain undulations, walls, and other data within the virtual environment. This information is typically stored as a 3D model, mesh data, or a set of colliders, used to simulate the physical limitations of sound sources in the virtual world. The first position information of the sound source refers to its position in the previous frame of audio data, while the current position information refers to its position in the current frame of audio data. Determining whether scene terrain exists between the first and current positions of the sound source can be achieved by calculating whether the line segment connecting these two positions collides or overlaps with any obstacles in the scene terrain information. For example, raycasting or line segment-polygon / mesh intersection detection algorithms can be used.

[0071] This application's solution obtains the location information of a sound source at continuous time points and combines it with preset terrain data of the virtual scene to determine whether the sound source's movement trajectory conflicts with the physical constraints of the scene. Specifically, when the sound source moves from the first location information to the current location information, if scene terrain exists along its path, it indicates that the sound source's movement trajectory does not conform to the physical rules of the virtual world, such as passing through walls or obstacles. This movement that does not conform to physical rules is identified as a location anomaly. In this way, unreasonable sound source positions caused by data transmission errors, cheating, or system failures can be effectively detected.

[0072] The aforementioned technical solution enables precise determination of sound source location anomalies based on the terrain information of a virtual scene. This method directly utilizes the physical constraints of the virtual environment, making anomaly identification more accurate and reliable, avoiding potential misjudgments that might occur when relying solely on speed or distance. Consequently, it provides a more accurate basis for anomaly type identification for subsequent audio data correction, thereby enhancing the realism and fairness of audio processing in low-latency gaming headsets.

[0073] This application proposes a more refined confidence update mechanism, which introduces a dynamic adjustment coefficient to enable the confidence update process to adaptively adjust according to the current confidence level.

[0074] Specifically, the current audio confidence of the sound source is updated based on the anomaly identification results to obtain the updated current audio confidence, including: Obtain the first correspondence; the first correspondence includes a one-to-one correspondence between multiple current audio confidence ranges and multiple first coefficients; take the first coefficient corresponding to the current audio confidence range in which the current audio confidence of the sound source is located in the first correspondence as the second coefficient; update the current audio confidence of the sound source according to the second coefficient, the anomaly identification result and the preset confidence adjustment step size, and obtain the updated current audio confidence.

[0075] Specifically, the first correspondence refers to a pre-established mapping relationship used to guide confidence adjustment, which associates different current audio confidence ranges with different first coefficients. For example, the current audio confidence can be divided into multiple intervals, each interval corresponding to a specific first coefficient, which is used to adjust the sensitivity of confidence adjustment. The first coefficient can be understood as a weighting factor for confidence adjustment, its purpose being to dynamically adjust the magnitude of confidence changes based on the current confidence level. In practical applications, the first correspondence can be stored in a lookup table or defined through a function relationship to ensure that the system can select the most appropriate adjustment strategy at different confidence levels.

[0076] Furthermore, the second coefficient is obtained by querying the aforementioned first correspondence. When the current audio confidence of the sound source is obtained, the system determines which current audio confidence range the confidence value falls into, and determines the first coefficient corresponding to that range as the second coefficient. Thus, the second coefficient is dynamically determined based on the real-time current audio confidence of the sound source, thereby ensuring the adaptability of confidence adjustment.

[0077] Based on this, the current audio confidence of the sound source is updated according to the second coefficient, the anomaly identification result, and the preset confidence adjustment step size. The preset confidence adjustment step size is a basic adjustment amount, while the second coefficient weights it to form the actual adjustment step size. The anomaly identification result indicates whether the confidence should be increased (e.g., when there are no anomalies) or decreased (e.g., when anomalies exist), and by how much (e.g., based on the number of anomaly types). In this way, the confidence update considers not only the presence of anomalies but also the current confidence level, making the adjustment process more accurate and reasonable.

[0078] This application's solution addresses the potential instability in confidence updates inherent in basic schemes by introducing a first correspondence and a mechanism for dynamically selecting a second coefficient. When the current audio confidence of a sound source changes, the system adaptively selects an appropriate second coefficient based on its current confidence range. This second coefficient serves as a weight for the adjustment step size, allowing the confidence to be updated with varying sensitivities in different contexts. For example, when the confidence is high, the second coefficient may be smaller, resulting in a smoother confidence adjustment and avoiding excessive fluctuations; conversely, when the confidence is low or an anomaly is detected, the second coefficient may be larger, causing the confidence to drop or rise rapidly to quickly reflect the new situation. This dynamic adjustment mechanism ensures the flexibility and accuracy of confidence updates, thereby enhancing the robustness of the entire audio processing method.

[0079] Through the aforementioned technical solution, the confidence update process becomes more intelligent and adaptive. By dynamically selecting adjustment coefficients based on the current range of audio confidence, the system can achieve more refined management of confidence. This not only helps avoid unnecessary drastic fluctuations when confidence is high, thus maintaining system stability, but also enables rapid and effective adjustment of confidence when confidence is low or anomalies are detected, allowing for quick responses to environmental changes. Therefore, this solution significantly improves the adaptability and accuracy of low-latency gaming headset audio processing methods in complex virtual scenarios, providing users with a more stable and reliable audio experience.

[0080] When the anomaly identification result indicates that the sound source has an anomaly, and the target anomaly type of the sound source is specified, the current audio confidence of the sound source is updated based on the second coefficient, the anomaly identification result, and the preset confidence level adjustment step size, resulting in the updated current audio confidence, including: When the anomaly identification result indicates that there is an anomaly in the sound source and the target anomaly type of the sound source, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size; the product of the initial confidence adjustment step size and the number of target anomaly types is used as the target confidence adjustment step size; the difference between the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

[0081] Specifically, the second coefficient is a coefficient obtained from the preset first correspondence based on the current audio confidence level of the sound source. Its value can be dynamically adjusted according to the current confidence level. For example, when the current confidence level is high, the second coefficient may be small to avoid over-adjustment; when the current confidence level is low, the second coefficient may be large to accelerate the decrease in confidence. The preset confidence adjustment step size is a pre-set basic adjustment amount used to provide a benchmark adjustment range during the confidence update process. The initial confidence adjustment step size is obtained by multiplying the second coefficient by the preset confidence adjustment step size. Its purpose is to initially weight the basic adjustment step size according to the current confidence level (reflected by the second coefficient).

[0082] The target anomaly type refers to the specific type of anomaly indicated by the anomaly identification results, such as anomaly in sound source velocity and / or anomaly in sound source location. The number of target anomaly types refers to the total number of anomaly types currently detected; for example, if both velocity and location anomalies exist simultaneously, the number is 2. The target confidence adjustment step size is obtained by multiplying the initial confidence adjustment step size by the number of target anomaly types. Its purpose is to further amplify the adjustment step size according to the complexity or diversity of the anomalies, ensuring that the decrease in confidence fully reflects the severity of the anomaly. The updated current audio confidence is obtained by subtracting the target confidence adjustment step size from the current audio confidence, representing the value after the reliability of the sound source audio data has been reassessed following the detection of anomalies.

[0083] This application's solution addresses the aforementioned limitations by dynamically adjusting the confidence level through the introduction of a second coefficient and the number of target anomaly types. Specifically, when the anomaly identification result indicates an anomaly in the sound source, firstly, the second coefficient is used to weight a preset confidence adjustment step size based on the current audio confidence level, thus obtaining an initial confidence adjustment step size. This weighting mechanism ensures that the magnitude of the confidence adjustment matches the current confidence level, avoiding the inaccuracies that may arise from a fixed step size. Secondly, the initial confidence adjustment step size is further multiplied by the number of target anomaly types to generate a target confidence adjustment step size. This step is crucial, as it allows the magnitude of the confidence decrease to directly reflect the complexity or multiplicity of the detected anomaly. For example, if the sound source exhibits both velocity and location anomalies, the target confidence adjustment step size will be larger, resulting in a greater decrease in the current audio confidence level, which more accurately reflects the fact that the reliability of the sound source's audio data has significantly decreased. Finally, by subtracting the dynamically calculated target confidence adjustment step size from the current audio confidence, a more accurate and responsive updated current audio confidence can be obtained.

[0084] Through the above technical solution, when anomalies are identified in the audio data of a sound source, the current audio confidence level of the sound source can be adjusted more finely and dynamically based on the severity and variety of the anomalies. This adjustment mechanism allows the updated current audio confidence level to more accurately reflect the actual reliability of the sound source audio data, avoiding the problem of inaccurate confidence level updates caused by using a fixed adjustment step size. Therefore, it can provide a more reliable basis for subsequent audio data correction, thereby improving the overall accuracy of audio processing and the user experience of low-latency gaming headsets.

[0085] When the anomaly identification result indicates that the sound source is not abnormal, the current audio confidence of the sound source is updated based on the second coefficient, the anomaly identification result, and the preset confidence level, by adjusting the step size. The updated current audio confidence level includes: When the anomaly identification result indicates that there is no anomaly in the sound source, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size; the product of the initial confidence adjustment step size and 2 is used as the target confidence adjustment step size; the sum of the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

[0086] Specifically, when the anomaly identification result clearly indicates that the sound source has no anomalies, firstly, the second coefficient determined based on the current audio confidence of the sound source is multiplied by a preset confidence adjustment step size to obtain an initial confidence adjustment step size. This initial confidence adjustment step size reflects the basic confidence adjustment amount under the condition of no anomalies. Further, to more effectively improve the confidence, this initial confidence adjustment step size is multiplied by a fixed coefficient of 2 to obtain a target confidence adjustment step size. This target confidence adjustment step size is an amplified adjustment amount designed to accelerate confidence recovery. Finally, the current audio confidence of the sound source is summed with the target confidence adjustment step size, and the result is the updated current audio confidence.

[0087] The proposed solution significantly increases the confidence level by multiplying the initial confidence adjustment step size by 2 when the anomaly identification result indicates that the sound source is not abnormal. This non-linear, accelerated confidence growth mechanism effectively solves the problem of slow audio confidence recovery after a sound source has maintained normal behavior. Due to this amplification effect, the system can more quickly confirm the normal state of the sound source, thus avoiding audio processing delays or inaccuracies that may result from slow confidence recovery, ensuring low latency and high accuracy of audio in esports scenarios.

[0088] Through the above technical solution, when a sound source is identified as normal, its current audio confidence level can be improved more quickly and significantly. This enables the system to respond rapidly to the normal behavior of the sound source, accelerates the confidence level recovery process, and thus ensures that in scenarios with extremely high real-time requirements, such as e-sports, the audio data of normal sound sources can be processed in a timely and accurate manner, effectively improving user experience and competitive performance.

[0089] This application further proposes a step for determining whether the updated current audio confidence score is greater than the target confidence threshold, including: Obtain the preset game influence index of the sound source; use the product of the preset game influence index and the preset threshold as the target confidence threshold; determine whether the updated audio confidence of the sound source is greater than the target confidence threshold.

[0090] Specifically, the preset game impact index of a sound source can be understood as a quantitative indicator related to the importance of a specific sound source in the game, its impact on the player experience, or the potential harm of its abnormal behavior. For example, the preset game impact index for key enemy sound sources or mission objective sound sources can be set higher to ensure that their abnormalities can be more rigorously judged and corrected. The preset threshold, on the other hand, can be a system-preset constant used to adjust the target confidence threshold benchmark. Its purpose is to provide an adjustable multiplier factor, allowing the target confidence threshold to be dynamically adjusted based on the preset game impact index. By using the product of the preset game impact index and the preset threshold as the target confidence threshold, personalized and dynamic settings of the target confidence threshold can be achieved.

[0091] This application's solution dynamically determines the target confidence threshold by introducing a preset game influence index for the sound source and multiplying it by a preset threshold, thus solving the problem of fixed or unadaptable target confidence thresholds in traditional solutions. Specifically, the preset game influence index reflects the importance of different sound sources in the game or their impact on the player experience; for example, the influence index can be set higher for crucial enemy footsteps or teammate voices. When this index is multiplied by the preset threshold, a target confidence threshold closely related to the sound source's characteristics is generated. Therefore, for high-influence sound sources, the target confidence threshold is correspondingly increased, ensuring that correction is only considered acceptable when the updated current audio confidence is sufficiently high, avoiding misjudgment or over-correction of important sound sources. Conversely, for low-influence sound sources, the target confidence threshold is relatively low, allowing correction at lower confidence levels, improving system flexibility and responsiveness. This dynamic adjustment mechanism ensures that correction decisions for abnormal audio can better adapt to the game environment and player needs.

[0092] Through the above technical solution, this application can dynamically adjust the target confidence threshold according to the actual importance or influence of the sound source in the game. This avoids the problems of inaccurate or untimely correction that may be caused by using a single fixed threshold, and significantly improves the intelligence and adaptability of abnormal audio correction. Specifically, for sound sources that have a significant impact on the gaming experience, a higher target confidence threshold can be set, thereby ensuring that correction is only performed when there is a high degree of certainty that the sound is abnormal, reducing the risk of miscorrection; while for sound sources with a smaller impact, a lower threshold can be set to improve the timeliness of correction. This refined threshold management mechanism enables the audio processing method for low-latency e-sports headsets to more accurately serve players, optimize the gaming experience, and effectively reduce the negative impact caused by improper correction of audio anomalies.

[0093] This application further proposes a more refined audio data correction method, the specific implementation of which is as follows: The audio data also includes the sound source's movement speed. When the updated current audio confidence score is greater than the target confidence score threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain the target audio data, including: When the target anomaly type of the sound source is a positional anomaly, the endpoint of the intersection line segment closest to the first position information is used as the corrected position information in the current audio data to obtain the target audio data; the intersection line segment is the line segment in the target line segment that overlaps with the scene terrain, and the target line segment is the line connecting the first position information and the current position information; when the target anomaly type of the sound source is a velocity anomaly, the intersection position of the target circle and the target line segment is used as the corrected position information in the current audio data to obtain the target audio data; the center of the target circle is the first position information, and the radius of the target circle is the product of the sound source's movement velocity and the first duration in the first audio data; when the target anomaly type of the sound source includes both positional anomaly and velocity anomaly, it is determined whether there is an intersection line segment within the target circle; when there is an intersection line segment within the target circle, the endpoint of the intersection line segment closest to the first position information is used as the corrected position information in the current audio data to obtain the target audio data; when there is no intersection line segment within the target circle, the intersection position of the target circle and the target line segment is used as the corrected position information in the current audio data to obtain the target audio data.

[0094] Specifically, to more accurately correct the audio data of the sound source, the audio data is configured to include the sound source's movement speed. This movement speed can be real-time or predicted speed information of the sound source obtained from the game engine or other physics simulation modules, which is used to provide necessary kinematic parameters when speed anomalies are corrected. The target line segment can be understood as a straight line segment connecting the sound source's first position information and its current position information. This line segment represents the theoretical movement trajectory of the sound source between two consecutive audio frames. The intersection line segment refers to the portion of the target line segment that overlaps with the scene terrain of the virtual scene. When the sound source moves between the first position information and the current position information, if its path crosses the scene terrain (e.g., walls, obstacles), the overlapping portion is the intersection line segment. The target circle is a circular area with the sound source's first position information as its center and the maximum distance the sound source can theoretically move within a first duration as its radius. This maximum distance is calculated by multiplying the sound source's movement speed in the first audio data by the first duration. The target circle is designed to define the maximum range of movement that the sound source can achieve at normal speeds.

[0095] In practical applications, when the target anomaly type of a sound source is only a positional anomaly, it indicates that the sound source's current position information may have penetrated the scene terrain. In this case, the corrected position information is determined to be the endpoint of the intersecting line segment closest to the first position information. This means that the sound source is "traced back" to its last valid position in front of the scene terrain, thus avoiding clipping. When the target anomaly type of a sound source is only a velocity anomaly, it indicates that the sound source's current position information exceeds the range allowed by its maximum movement speed. In this case, the corrected position information is determined to be the intersection of the target circle and the target line segment. This restricts the sound source's position to its reasonable speed range, ensuring the physical realism of its movement trajectory. When the target anomaly type of a sound source includes both positional and velocity anomalies, both anomalies need to be considered comprehensively.

[0096] First, determine if there are intersecting line segments within the target circle. If so, prioritize the correction logic for positional anomalies, using the endpoint of the intersecting line segment closest to the first position information as the corrected position information to resolve the clipping issue. If there are no intersecting line segments within the target circle, it indicates that even under the speed anomaly limit, the sound source has not penetrated the terrain. In this case, prioritize the correction logic for speed anomalies, using the intersection of the target circle and the target line segment as the corrected position information to resolve the speeding issue. This case-by-case approach ensures the priority and rationality of the correction.

[0097] This application's solution addresses the lack of fine-grained processing in audio data correction by introducing the sound source's movement speed as a correction criterion and designing specific correction strategies for different anomaly types (positional anomaly, speed anomaly, or both). Specifically, when a sound source experiences a positional anomaly, the solution identifies the intersection of the target line segment and the scene terrain, correcting the sound source's position to the terrain boundary, effectively preventing the sound source from penetrating obstacles and producing unrealistic audio effects. When a sound source experiences a speed anomaly, a target circle is constructed to limit the sound source's reasonable movement range, correcting the sound source's position to the boundary allowed by its maximum movement speed, thus ensuring the physical rationality of the sound source's movement. Furthermore, when both positional and speed anomalies exist simultaneously, this solution prioritizes handling clipping issues by determining whether there are intersecting line segments within the target circle; if no clipping occurs, it handles overspeed issues. This priority processing mechanism ensures the logic and effectiveness of the correction, enabling the corrected audio data to more accurately reflect the sound source's true state in the virtual scene.

[0098] Through the above technical solution, this application can provide a more refined and accurate audio data correction method based on the specific anomaly type of the sound source. Especially in esports scenarios, the rapid movement of sound sources and interactions with complex terrain place higher demands on the real-time performance and realism of audio. This solution, by introducing the sound source's movement speed and combining it with geometric calculations (such as intersecting line segments and target circles), effectively solves problems such as sound source clipping and excessive speed, ensuring that the corrected sound source position conforms to physical constraints while avoiding visual inconsistencies. As a result, esports players can hear more realistic and accurate sound source positioning information, significantly improving the immersion of the gaming experience and the fairness of competition.

[0099] This application proposes an audio processing system for low-latency gaming headsets, comprising: an acquisition device and a processing device; the acquisition device is used to acquire current audio data of a sound source in a virtual scene, scene constraint information in the virtual scene, and current audio confidence of the sound source; the scene constraint information includes physical constraint information of the sound source and scene terrain information of the virtual scene; the processing device is used to perform anomaly identification on the sound source based on the current audio data and scene constraint information, and obtain anomaly identification results; the anomaly identification results are used to indicate that the sound source does not have anomalies, or to indicate that the sound source has anomalies, and the target anomaly type of the sound source; the target anomaly type includes sound source velocity anomaly and / or sound source position anomaly; the processing device is further used to update the current audio confidence of the sound source based on the anomaly identification results, and obtain updated current audio confidence; the processing device is further used to determine whether the updated current audio confidence is greater than a target confidence threshold when the anomaly identification results indicate that the sound source has anomalies and the target anomaly type of the sound source; the processing device is further used to correct the current audio data according to the target anomaly type of the sound source when the updated current audio confidence is greater than the target confidence threshold, and obtain target audio data.

[0100] In a preferred embodiment, the acquisition device can be a communication module integrated inside the gaming headset, which receives audio data streams and scene metadata output by the game engine in real time by communicating with the main computing device.

[0101] The processing device may be one or more microcontrollers, digital signal processors (DSPs), field-programmable gate arrays (FPGAs), or general-purpose processors, which have integrated memory for storing program code and data to execute the processing device of this application.

[0102] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for audio processing of low-latency gaming headphones, characterized in that, include: Obtain the current audio data of the sound source in the virtual scene, the scene limitation information in the virtual scene, and the current audio confidence of the sound source; Scene constraint information includes the physical constraints of the sound source and the scene terrain information of the virtual scene; Based on the current audio data and scene constraints, anomaly identification is performed on the sound source to obtain anomaly identification results. The anomaly identification results are used to indicate whether the sound source does not have anomalies, or to indicate whether the sound source has anomalies, as well as the target anomaly type of the sound source. The target anomaly type includes abnormal sound source velocity and / or abnormal sound source location. The current audio confidence of the sound source is updated based on the anomaly identification results to obtain the updated current audio confidence. When the anomaly identification result indicates that there is an anomaly in the sound source and the target anomaly type of the sound source, determine whether the updated current audio confidence is greater than the target confidence threshold. When the updated current audio confidence score is greater than the target confidence score threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain the target audio data.

2. The audio processing method for low-latency gaming headphones according to claim 1, characterized in that, Based on the current audio data and scene constraints, anomaly identification is performed on the sound source to obtain the anomaly identification results, including: Based on the current audio data and the physical constraints of the sound source, determine whether there is an abnormal speed of the sound source; If the sound source has an abnormal velocity, the abnormality identification result indicates that the sound source has an abnormality, and the target abnormality type of the sound source includes abnormal sound source velocity. Based on the current audio data and the scene terrain information of the virtual scene, determine whether there is an abnormal location of the sound source; If the sound source has an abnormal location, the abnormality identification result indicates that the sound source has an abnormality, and the target abnormality type of the sound source includes abnormal sound source location; If the sound source does not have any abnormal location or speed, the anomaly identification result indicates that the sound source does not have any abnormality.

3. The audio processing method for low-latency gaming headphones according to claim 2, characterized in that, The physical constraint information of the sound source includes the maximum moving speed of the sound source. The current audio data includes the current position information of the sound source and the time when the audio data was generated. Based on the current audio data and the physical constraint information of the sound source, it is determined whether there is an abnormal speed of the sound source, including: Obtain the first location information; the first location information is the location information of the sound source in the first audio data, and the first audio data is the audio data of the previous frame of the current audio data; Determine the first displacement distance between the first position information and the current position information; The ratio of the first displacement distance to the first duration is used as the calculation speed of the sound source; the first duration is the duration between the time when the current audio data is generated and the time when the first audio data is generated. When the calculated speed of the sound source is greater than the maximum moving speed of the sound source, it is determined that there is a speed anomaly in the sound source; If the calculated speed of the sound source is less than or equal to the maximum moving speed of the sound source, it is determined that there is no speed anomaly in the sound source.

4. The audio processing method for low-latency gaming headphones according to claim 3, characterized in that, Based on the current audio data and the virtual scene's terrain information, determine whether the sound source has an abnormal location, including: Based on the scene terrain information of the virtual scene, determine whether there is scene terrain between the first position information of the sound source and the current position information of the sound source; If there is a scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that the sound source has a location anomaly; If there is no scene terrain between the first location information of the sound source and the current location information of the sound source, it is determined that the sound source does not have a location anomaly.

5. The audio processing method for low-latency gaming headphones according to claim 1, characterized in that, The current audio confidence score of the sound source is updated based on the anomaly identification results, resulting in the updated current audio confidence score, including: Obtain the first correspondence; the first correspondence includes a one-to-one correspondence between multiple current audio confidence ranges and multiple first coefficients; The first coefficient corresponding to the current audio confidence range of the sound source in the first correspondence is used as the second coefficient; Based on the second coefficient, the anomaly identification results, and the preset confidence adjustment step size, the current audio confidence of the sound source is updated to obtain the updated current audio confidence.

6. The audio processing method for a low-latency gaming headset according to claim 5, characterized in that, Based on the second coefficient, the anomaly identification results, and the preset confidence adjustment step size, the current audio confidence of the sound source is updated to obtain the updated current audio confidence, including: When the anomaly identification result indicates that there is an anomaly in the sound source and the target anomaly type of the sound source, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size. The product of the initial confidence adjustment step size and the number of target anomaly types is used as the target confidence adjustment step size; The difference between the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

7. The audio processing method for a low-latency gaming headset according to claim 5, characterized in that, Based on the second coefficient, the anomaly identification results, and the preset confidence adjustment step size, the current audio confidence of the sound source is updated to obtain the updated current audio confidence, including: When the anomaly identification result indicates that there is no anomaly in the sound source, the product of the second coefficient and the preset confidence adjustment step size is used as the initial confidence adjustment step size; The product of the initial confidence adjustment step size and 2 is used as the target confidence adjustment step size; The sum of the current audio confidence and the target confidence adjustment step size is used as the updated current audio confidence.

8. The audio processing method for low-latency gaming headphones according to claim 1, characterized in that, Determine whether the updated current audio confidence score is greater than the target confidence score threshold, including: Preset game impact index for acquiring sound sources; The product of the preset game influence index and the preset threshold is used as the target confidence threshold; Determine whether the updated audio confidence of the sound source is greater than the target confidence threshold.

9. The audio processing method for a low-latency gaming headset according to claim 4, characterized in that, The audio data also includes the sound source's movement speed. When the updated current audio confidence score is greater than the target confidence score threshold, the current audio data is corrected according to the target anomaly type of the sound source to obtain the target audio data, including: When the target anomaly type of the sound source is a location anomaly, the endpoint of the intersection line segment closest to the first location information is taken as the corrected location information in the current audio data to obtain the target audio data; the intersection line segment is the line segment in the target line segment that overlaps with the scene terrain, and the target line segment is the line connecting the first location information and the current location information; When the target anomaly type of the sound source is velocity anomaly, the intersection of the target circle and the target line segment is used as the corrected position information in the current audio data to obtain the target audio data; the center of the target circle is the first position information, and the radius of the target circle is the product of the sound source's moving velocity and the first duration in the first audio data. When the target anomaly types of the sound source include positional anomalies and velocity anomalies, determine whether there are intersecting line segments within the target circle; When there is an intersecting line segment within the target circle, the endpoint of the intersecting line segment closest to the first position information is used as the corrected position information in the current audio data to obtain the target audio data; When there is no intersecting line segment within the target circle, the intersection position of the target circle and the target line segment is used as the corrected position information in the current audio data to obtain the target audio data.

10. An audio processing system for low-latency gaming headsets, characterized in that, include: Acquisition device and processing device; The acquisition device is used to acquire the current audio data of a sound source in a virtual scene, scene constraint information in the virtual scene, and the current audio confidence of the sound source; the scene constraint information includes the physical constraint information of the sound source and the scene terrain information of the virtual scene; The processing device is used to identify anomalies in the sound source based on the current audio data and scene constraint information, and obtain anomaly identification results; the anomaly identification results are used to indicate that there is no anomaly in the sound source, or to indicate that there is an anomaly in the sound source, and the target anomaly type of the sound source; the target anomaly type includes sound source velocity anomaly and / or sound source position anomaly. The processing device is also used to update the current audio confidence of the sound source based on the anomaly identification result, so as to obtain the updated current audio confidence. The processing device is also used to determine whether the updated current audio confidence level is greater than the target confidence level threshold when the anomaly identification result indicates that the sound source has an anomaly and the target anomaly type of the sound source. The processing device is also used to correct the current audio data according to the target anomaly type of the sound source when the updated current audio confidence is greater than the target confidence threshold, so as to obtain the target audio data.

Images