A wireless audio transmission method and system based on sound

Abstract

The present application discloses a wireless audio transmission method and system based on audio, which belongs to the field of data processing technology. The present application realizes precise adjustment of the sound quality optimization in the subsequent audio transmission stage by deeply analyzing and utilizing the behavior and characteristics of the target wireless audio device in the preceding audio transmission stage. In this way, not only the personalization level and accuracy of the sound quality optimization are significantly improved, but also the real-time and smoothness of the audio transmission are effectively guaranteed, bringing users a more excellent and personalized audio experience.

Description

Technical Field

[0001] The present application belongs to the field of data processing technology, and specifically relates to a wireless audio transmission method and system based on audio. Background Art

[0002] In the current field of audio transmission technology, although there are many sound quality optimization methods for wireless audio devices, most of these methods are based on fixed optimization strategies or simple user feedback mechanisms, which makes it difficult to achieve truly personalized sound quality optimization. At the same time, due to the lack of comprehensive consideration of the device behavior and characteristics in the previous audio transmission stage, these methods often cannot accurately predict and optimize the time consumption of subsequent audio transmission when optimizing sound quality, resulting in the real-time and smoothness of audio transmission being affected. Summary of the invention

[0003] The present application provides a speaker-based wireless audio transmission method and system, which can solve or partially solve the technical problems involved in the above-mentioned background technology.

[0004] The embodiment of the present application provides a wireless audio transmission method based on audio, which is applied to an audio transmission processing system. The method comprises: obtaining audio signal transmission information corresponding to a target wireless audio device in a same interactive session period in a plurality of preceding audio transmission stages; the audio signal transmission information comprises a preceding sound quality status label to which preceding voice data output by the target wireless audio device in the interactive session period belongs, and output node information for outputting the preceding voice data; extracting a preceding sound quality characterization vector according to the preceding sound quality status label in the audio signal transmission information, and identifying a sound quality preference label corresponding to the target wireless audio device according to the preceding sound quality characterization vector; the sound quality preference label characterizes the speech quality of the speech data to be transmitted in the interactive session period in the subsequent audio transmission stage. The invention relates to a method for transmitting a sound quality index of audio data; extracting a preceding context audio vector according to the output node information in the audio signal transmission information, and identifying an output prediction window corresponding to the target wireless audio device according to the preceding context audio vector; the output prediction window represents the time consumption information of transmitting the speech data set during the interactive conversation period in the subsequent audio transmission stage; obtaining a plurality of sound quality optimization strategy instructions of the sound quality preference label through the output prediction window; generating a sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions, wherein the global time consumption information of the sound quality optimization strategy instruction set is matched with the output prediction window; the sound quality optimization strategy instruction set is used to realize the sound quality optimization adjustment of the target wireless audio device during the interactive conversation period in the subsequent audio transmission stage.

[0005] Preferably, the step of identifying the sound quality preference label corresponding to the target wireless audio device according to the preceding sound quality characterization vector includes:

[0006] Obtaining initial sound quality representation vectors extracted from a plurality of initial sound quality labels;

[0007] For each of the initial sound quality labels, based on the initial sound quality characterization vector of the initial sound quality label and the preceding sound quality characterization vector, identify the sound quality feature weight corresponding to the initial sound quality label; the sound quality feature weight characterizes the possibility that the speech data set to be transmitted includes the data of the initial sound quality label;

[0008] A sound quality preference label is selected from the multiple initial sound quality labels, and the sound quality feature weight of the sound quality preference label meets the sound quality preference decision requirement.

[0009] Preferably, the obtaining of several sound quality optimization strategy instructions of the sound quality preference label through the output prediction window includes:

[0010] Acquire an initial voice data set belonging to the sound quality preference label, and determine a time series interval feature corresponding to each initial voice data in the initial voice data set;

[0011] Performing at least two types of integration on the initial speech data in the initial speech data set to form at least two speech data clusters, each speech data cluster including at least two of the initial speech data;

[0012] Determine the initial global time series interval feature corresponding to each of the voice data clusters according to the time series interval feature of the initial voice data in each of the voice data clusters;

[0013] A speech data cluster corresponding to an initial global time series interval feature paired with the output prediction window is determined, and a plurality of sound quality optimization strategy instructions of the sound quality preference label are determined based on the initial speech data included in the determined speech data cluster.

[0014] Preferably, the generating of the sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions comprises:

[0015] Determining the default sound quality element features corresponding to the plurality of sound quality optimization strategy instructions respectively;

[0016] Determine target sound quality element characteristics of the sound quality optimization strategy instruction set to be generated;

[0017] Based on the default sound quality element characteristics and the target sound quality element characteristics respectively corresponding to the several sound quality optimization strategy instructions, the several sound quality optimization strategy instructions are subjected to sound quality element fusion to obtain a sound quality optimization strategy instruction set.

[0018] Preferably, the sound quality element fusion of the plurality of sound quality optimization strategy instructions based on the default sound quality element features and the target sound quality element features respectively corresponding to the plurality of sound quality optimization strategy instructions to obtain a sound quality optimization strategy instruction set includes:

[0019] Determine a first sound quality optimization strategy instruction and a second sound quality optimization strategy instruction from the plurality of sound quality optimization strategy instructions, wherein the default sound quality element feature of the first sound quality optimization strategy instruction has a sound quality element fusion relationship with the target sound quality element feature, and the default sound quality element feature of the second sound quality optimization strategy instruction does not have a sound quality element fusion relationship with the target sound quality element feature;

[0020] Converting the second sound quality optimization strategy instruction into the target sound quality element feature according to a sound quality element mapping vector from the default sound quality element feature of the second sound quality optimization strategy instruction to the target sound quality element feature;

[0021] The first sound quality optimization strategy instruction and the second sound quality optimization strategy instruction of the target sound quality element feature are subjected to sound quality element fusion to obtain a sound quality optimization strategy instruction set.

[0022] Preferably, the default sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features or spatial stereoscopic element features; the target sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features or spatial stereoscopic element features; the sound quality optimization strategy instruction set includes at least one of clarity optimization instruction set, fullness optimization instruction set, balance optimization instruction set, timbre optimization instruction set or spatial stereoscopic optimization instruction set.

[0023] Preferably, the identifying the output prediction window corresponding to the target wireless audio device according to the preceding context audio vector includes:

[0024] Get some initial time-consuming information;

[0025] For each of the initial time consumption information, based on the initial time consumption information and the preceding context audio vector, identifying a time consumption prediction weight corresponding to the initial time consumption information;

[0026] An output prediction window is selected from the plurality of initial time-consuming information, and the time-consuming prediction weight corresponding to the output prediction window meets the set time-consuming discrimination requirement.

[0027] Preferably, the generating of the sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions comprises:

[0028] Obtaining target sound quality element features indicated by the user APP corresponding to the target wireless audio device;

[0029] Based on the several sound quality optimization strategy instructions, a sound quality optimization strategy instruction set of the target sound quality element characteristics is generated, and the sound quality optimization strategy instruction set includes the several sound quality optimization strategy instructions; the sound quality optimization strategy instruction set.

[0030] Preferably, the method further comprises:

[0031] Sending the global time-consuming information to the user APP corresponding to the target wireless audio device;

[0032] Obtaining the global time-consuming information after the global time-consuming information is prioritized;

[0033] The sound quality optimization strategy instructions included in the sound quality optimization strategy instruction set are changed through the global time-consuming information after the priority is updated, so as to obtain a sound quality optimization strategy instruction set that has been changed; the sound quality optimization strategy instruction set that has been changed is used to realize the sound quality optimization adjustment of the target wireless audio device during the interactive session period in the subsequent audio transmission stage.

[0034] Preferably, the method further comprises:

[0035] Acquire audio signal transmission information samples corresponding to the wireless audio device sample in the same interactive session period samples in a plurality of preceding audio transmission phase samples; the audio signal transmission information samples include preceding sound quality status label samples belonging to preceding voice data samples output by the wireless audio device sample in the interactive session period samples, and output node information samples outputting the preceding voice data samples;

[0036] Extracting a sound quality characterization vector sample according to the preceding sound quality status label sample in the audio signal transmission information sample, and identifying a sound quality preference label sample corresponding to the wireless audio device sample according to the sound quality characterization vector sample; the sound quality preference label sample characterizes the sound quality index of the voice data in the voice data set sample to be transmitted in the interactive session period sample in the subsequent audio transmission phase sample;

[0037] Extracting a preceding context audio vector sample according to the output node information in the audio signal transmission information sample, and identifying an output prediction window sample corresponding to the wireless audio device sample according to the preceding context audio vector sample; the output prediction window sample represents the time consumption information of the speech data set sample required to be transmitted for the interactive session period sample in the subsequent audio transmission phase sample;

[0038] Obtaining several sound quality optimization strategy instruction samples of the sound quality preference label samples through the output prediction window samples;

[0039] Generate a sound quality optimization strategy instruction sample set including the plurality of sound quality optimization strategy instruction samples, wherein the global time consumption information of the sound quality optimization strategy instruction sample set is paired with the output prediction window sample;

[0040] Obtain a priori sound quality optimization strategy instruction set corresponding to the wireless audio device sample in the subsequent audio transmission stage sample, perform algorithm debugging based on the sound quality optimization strategy instruction sample set and the priori sound quality optimization strategy instruction set, and obtain a sound quality optimization adjustment algorithm.

[0041] In some independent embodiments, the algorithm debugging is performed according to the sound quality optimization strategy instruction sample set and the priori sound quality optimization strategy instruction set to obtain the sound quality optimization adjustment algorithm, including:

[0042] Determining a strategy instruction training error between the sound quality optimization strategy instruction sample set and the prior sound quality optimization strategy instruction set;

[0043] Obtain a priori sound quality preference label corresponding to the wireless audio device sample in the subsequent audio transmission phase, and determine a sound quality preference training error between the sound quality preference label sample and the priori sound quality preference label;

[0044] Algorithm debugging is performed according to the strategy instruction training error and the sound quality preference training error to obtain a sound quality optimization adjustment algorithm.

[0045] In some independent embodiments, the algorithm debugging is performed according to the strategy instruction training error and the sound quality preference training error to obtain a sound quality optimization adjustment algorithm, including:

[0046] Acquire a priori time-consuming information corresponding to the sample of the wireless audio device in the subsequent audio transmission phase, and determine a time-consuming training error between the output prediction window sample and the a priori time-consuming information;

[0047] Algorithm debugging is performed according to the strategy instruction training error, the sound quality preference training error and the time-consuming training error to obtain a sound quality optimization adjustment algorithm.

[0048] An embodiment of the present application provides an audio transmission processing system, comprising at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor performs the above method.

[0049] An embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the above method are implemented.

[0050] The embodiment of the present application achieves precise adjustment of the sound quality optimization in the subsequent audio transmission stage by deeply analyzing and utilizing the behavior and characteristics of the target wireless audio device in the previous audio transmission stage. In this way, not only the personalization level and accuracy of the sound quality optimization are significantly improved, but also the real-time and smoothness of the audio transmission are effectively guaranteed, bringing users a more excellent and personalized audio experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A flowchart of a speaker-based wireless audio transmission method provided in an embodiment of the present application.

[0052] Figure 2 A schematic diagram of the structure of an audio transmission processing system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] The terms "first", "second", etc. in this application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are usually a class, and the number of objects is not limited. For example, the first object can be one or more. In addition, in this application, "and / or" represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0055] Figure 1 A wireless audio transmission method based on a speaker is shown, which is applied to an audio transmission processing system. The method includes the following steps 101 to 104.

[0056] Step 101: obtaining audio signal transmission information corresponding to the same interactive session period in several preceding audio transmission stages of the target wireless audio device; the audio signal transmission information includes a preceding sound quality status label to which preceding voice data output by the target wireless audio device during the interactive session period belongs, and output node information for outputting the preceding voice data.

[0057] Step 102: extract a preceding sound quality characterization vector based on the preceding sound quality status label in the audio signal transmission information, and identify a sound quality preference label corresponding to the target wireless audio device based on the preceding sound quality characterization vector; the sound quality preference label characterizes a sound quality indicator of voice data in a voice data set that needs to be transmitted during the interactive session period in a subsequent audio transmission phase.

[0058] Step 103: extract a preceding context audio vector based on the output node information in the audio signal transmission information, and identify an output prediction window corresponding to the target wireless audio device based on the preceding context audio vector; the output prediction window represents time-consuming information for transmitting a voice data set during the interactive session period in the subsequent audio transmission stage.

[0059] Step 104: obtaining, through the output prediction window, a plurality of sound quality optimization strategy instructions of the sound quality preference tag; generating a sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions, wherein the global time consumption information of the sound quality optimization strategy instruction set is paired with the output prediction window; the sound quality optimization strategy instruction set is used to implement sound quality optimization adjustment of the target wireless audio device during the interactive session period in the subsequent audio transmission stage.

[0060] In the field of modern audio transmission and processing, it is crucial to ensure that wireless audio devices can provide a high-quality audio experience. To achieve this goal, the audio transmission processing system needs to implement a series of complex strategies to ensure that excellent sound quality is maintained under various transmission conditions. The following is a specific application scenario example, which describes in detail how the audio transmission processing system optimizes the sound quality of the target wireless audio device through steps 101 to 104.

[0061] Take a large conference room as an example. The conference room is equipped with a high-end wireless audio system for daily meetings and presentations. In order to ensure that each meeting can provide a clear, high-quality audio experience, the administrator of the conference room decides to use the audio transmission processing system of the embodiment of the present application to optimize the sound quality of the audio equipment.

[0062] 1. The audio transmission processing system first starts to collect the audio signal transmission information of the target wireless audio device during the same interactive session period in several previous audio transmission stages. This step involves a detailed analysis of the voice data output by the audio device during the previous session period. Specifically, the system records the sound quality status label to which the voice data belongs in each previous session period. These labels may include "clear", "muddy", "echoic", "moderate volume", etc., which are used to describe the specific state of the audio at that time. At the same time, the system also records the output node information of the output of these voice data, including the transmission path, node delay, signal strength, etc., for subsequent analysis.

[0063] 2. Next, the audio transmission processing system extracts the preceding sound quality characterization vector based on the preceding sound quality status label in the collected audio signal transmission information. These vectors are obtained by numerically encoding the sound quality status label and can reflect various characteristics of the sound quality. The system further uses these preceding sound quality characterization vectors to identify the sound quality preference label corresponding to the target wireless audio device through a machine learning algorithm. The sound quality preference label represents the sound quality index of the voice data in the voice data set that needs to be transmitted during the interactive session in the subsequent audio transmission stage. For example, if most of the sound quality status labels in the preceding session are "clear", the sound quality preference label may be identified as "prefer clear sound quality".

[0064] 3. After obtaining the sound quality preference label, the audio transmission processing system continues to extract the previous context audio vector based on the output node information in the audio signal transmission information. These vectors contain various context information about the audio transmission in the previous session, such as transmission delay, packet loss rate, signal fluctuation, etc. The system uses these previous context audio vectors to identify the output prediction window corresponding to the target wireless audio device through time series analysis technology. The output prediction window represents the time-consuming information of the voice data set that needs to be transmitted during the interactive session period in the subsequent audio transmission stage. This step is crucial to determining the implementation time window of the sound quality optimization strategy.

[0065] 4. Finally, the audio transmission processing system obtains several sound quality optimization strategy instructions corresponding to the sound quality preference label through the output prediction window. These instructions may include adjusting audio encoding parameters, optimizing transmission paths, enhancing signal strength, etc., aiming to improve the sound quality in the subsequent audio transmission stage. The system generates a sound quality optimization strategy instruction set including these sound quality optimization strategy instructions, and ensures that the global timing information of the instruction set is paired with the output prediction window. This means that all sound quality optimization strategies will be executed within the predetermined time window to ensure timely and effective adjustment of the sound quality.

[0066] Through the implementation of the above steps, the audio transmission processing system can significantly optimize the sound quality of the target wireless audio device. Specifically, the system can intelligently identify the sound quality preference and formulate corresponding optimization strategies based on the sound quality status and transmission context of the previous conversation. This not only improves the clarity and intelligibility of the audio, but also reduces transmission delays and signal fluctuations, thereby providing a more stable and high-quality audio experience. In addition, the embodiment of the present application can also dynamically adjust the sound quality optimization strategy according to different application scenarios and transmission conditions to ensure that the best sound quality effects can be provided in various environments.

[0067] In summary, the audio transmission processing system achieves intelligent optimization of the sound quality of the target wireless audio device and improves the audio transmission quality and user experience through a series of steps such as acquiring audio signal transmission information, extracting sound quality characterization vectors and context audio vectors, identifying sound quality preference labels and output prediction windows, and acquiring and generating sound quality optimization strategy instruction sets.

[0068] In an embodiment of the present application, step 102 involves extracting key features from audio signal transmission information and identifying the sound quality preference of the target wireless audio device based on these features. First, the audio transmission processing system obtains the audio signal transmission information of the target wireless audio device in the same interactive session period in several preceding audio transmission stages. These information include preceding sound quality status labels, which are a description of the sound quality, such as "clear", "echo", etc., which reflect the specific state of the audio signal in the preceding transmission stage. Secondly, the system extracts the preceding sound quality characterization vector based on these preceding sound quality status labels. This step is achieved by numerically encoding the sound quality status labels, and each label is converted into one or more numerical values, which constitute the sound quality characterization vector. These vectors can capture various characteristics of the sound quality and provide a basis for subsequent analysis and processing. Then, the system uses these preceding sound quality characterization vectors to identify the sound quality preference label corresponding to the target wireless audio device. This step is achieved by a machine learning algorithm, and the algorithm infers the sound quality preference label based on the characteristics of the preceding sound quality characterization vector. This label characterizes the sound quality index of the voice data in the voice data set that needs to be transmitted during the interactive conversation period in the subsequent audio transmission stage. In other words, it tells the system how to adjust the sound quality in the subsequent transmission to meet the preferences of the target wireless audio device. Therefore, step 102 realizes the intelligent identification of the sound quality preference of the target wireless audio device by extracting and analyzing the previous sound quality status label and using a machine learning algorithm to identify the sound quality preference label. This technical solution provides an important basis for the formulation of subsequent sound quality optimization strategies, ensuring that the sound quality optimization can be carried out according to the specific preferences of the target wireless audio device, thereby providing a more personalized sound quality experience.

[0069] Furthermore, the preceding sound quality characterization vector, as one of the core invention points of the present application, plays the role of a bridge connecting the sound quality status label and the sound quality preference identification.

[0070] First, the preamble sound quality characterization vector is the result of numerically encoding the sound quality status label. Each sound quality status label, whether "clear", "muddy" or "echoic", will be converted into one or more specific values, which can comprehensively and accurately reflect the various characteristics of the sound quality.

[0071] Taking the sound quality status label "clear" as an example, it may be converted into a numerical feature vector containing multiple dimensions. In this vector, each dimension represents a specific aspect of the sound quality, such as the spectral balance, dynamic range, noise level, etc. of the audio. Specifically, this vector may include the values ​​of the following dimensions:

[0072] Spectral balance: This value reflects whether the energy distribution of the audio signal in different frequency bands is uniform. For "clear" sound quality, this value should be relatively high, indicating that the spectrum distribution of the audio signal is relatively balanced;

[0073] Dynamic range: This value indicates the difference between the strongest and weakest parts of the audio signal. For "clear" sound quality, the dynamic range should be moderate, neither too large nor too small, to ensure that the details of the audio signal can be fully displayed;

[0074] Noise level: This value measures the degree of background noise in the audio signal. For "clear" sound quality, the noise level should be low to reduce interference with the main audio content.

[0075] In addition to the above dimensions, the preamble sound quality representation vector may also contain other numerical features related to sound quality, such as distortion, stereo separation, etc. These numerical features together constitute the preamble sound quality representation vector, providing a rich information basis for subsequent sound quality preference identification.

[0076] Secondly, the numerical features of the preceding sound quality characterization vector are not fixed. They are dynamically adjusted and optimized according to the actual sound quality status label and audio signal characteristics. This means that for different sound quality status labels and audio signals, the numerical features of the preceding sound quality characterization vector will also be different.

[0077] Finally, by using these information-rich preamble sound quality characterization vectors, the audio transmission processing system can more accurately identify the sound quality preferences of the target wireless audio device. This step is crucial for developing a personalized sound quality optimization strategy, because it ensures that the sound quality optimization can be performed for the specific sound quality preferences of the target wireless audio device, thereby providing a more excellent sound quality experience.

[0078] For example, the sound quality has been simplified and divided into four main dimensions for numerical representation, namely "spectral balance", "dynamic range", "noise level" and "distortion". Each dimension is assigned a specific value to quantify the sound quality characteristics under that dimension.

[0079] Preamble sound quality characterization vector = [spectral balance, dynamic range, noise level, distortion]

[0080] For a specific preceding sound quality state label (such as "clear"), the corresponding sound quality representation vector may be as follows: [0.9, 15dB, -60dBFS, 0.01%].

[0081] Spectral balance: 0.9 means that the spectrum distribution of the audio signal is very balanced and close to the ideal state. This value may be obtained by comparing the energy distribution of the signal in different frequency bands.

[0082] Dynamic range: 15dB means the difference between the strongest and weakest parts of the audio signal is 15 decibels. This value is moderate, ensuring that the details of the audio signal are fully displayed while avoiding the sound being too sharp or dull.

[0083] Noise level: -60dBFS (decibel full scale) means that the background noise is very low and will hardly interfere with the main audio content. This value is relative to the maximum possible amplitude of the audio signal.

[0084] Distortion: 0.01% means that the distortion introduced by the audio signal during transmission or processing is very small and almost negligible. This value is one of the important indicators to measure the purity of sound quality.

[0085] In practical applications, the dimensions and values ​​of the sound quality characterization vector can be adjusted and optimized according to specific application scenarios and requirements.

[0086] In addition, the extraction of the preamble sound quality characterization vector usually involves complex audio signal processing techniques, including but not limited to spectrum analysis, dynamic range measurement, noise suppression, and distortion detection, etc. These techniques can help the system accurately capture the sound quality features in the audio signal and convert them into numerical feature vectors that can be used for subsequent processing.

[0087] Finally, by utilizing these information-rich preceding sound quality representation vectors, the audio transmission processing system can more intelligently identify the sound quality preferences of the target wireless audio device and formulate a personalized sound quality optimization strategy accordingly, thereby improving the user's audio experience.

[0088] On the basis of the above, step 103 focuses on extracting key data from the audio signal transmission information, and based on this, predicting the time-consuming requirements of the target wireless audio device in the subsequent audio transmission stage. First, the system obtains the output node information in the audio signal transmission information. These output node information are data about each node that the audio signal passes through during the transmission process, and they contain key information such as the path, speed, and possible delay of the audio signal transmission. This information is crucial for understanding the behavior pattern of the audio signal during the transmission process. Then, the system extracts the previous context audio vector based on these output node information. This step is achieved by analyzing the parameters such as the timestamp, data packet size, and transmission speed in the output node information and converting these parameters into numerical vectors. These vectors capture the context information of the audio signal in the previous transmission stage and provide a data basis for subsequent predictions. Then, the system uses these previous context audio vectors to identify the output prediction window corresponding to the target wireless audio device. This step is implemented by a machine learning algorithm, which predicts the time-consuming information of the speech data set to be transmitted during the interactive conversation period in the subsequent audio transmission stage based on the characteristics of the previous context audio vector. This time-consuming information is the so-called output prediction window, which characterizes the time range required for the audio signal in the subsequent transmission. It can be seen that step 103 realizes the intelligent prediction of the time consumption requirements of the target wireless audio device in the subsequent audio transmission stage by extracting and analyzing the output node information and using the machine learning algorithm to predict the output prediction window. This technical solution provides an important basis for the formulation of subsequent audio transmission strategies, ensuring that audio transmission can be optimized according to the specific needs of the target wireless audio device, thereby providing a more efficient and stable audio transmission experience.

[0089] In the embodiment of the present application, the preceding context audio vector is a numerical feature vector extracted based on the preceding (i.e., previously occurring) audio signal and its transmission context information during the audio signal transmission process. It captures the key features of the audio signal before transmission, including but not limited to the signal strength, frequency distribution, transmission path characteristics, node delay, etc. These features together constitute a vector representation with predictive value for subsequent audio transmission.

[0090] For example, through a series of complex audio signal processing and analysis techniques, the following key features are extracted from the previous audio signal and encoded into a previous context audio vector:

[0091] Preamble context audio vector = [signal average power, spectrum center frequency, maximum delay node ID, average transmission speed]

[0092] Example vector = [0.5W, 1kHz, Node3, 2Mbps]

[0093] Signal average power (0.5W): Indicates the average energy level of the preceding audio signal and is an important indicator for evaluating signal strength.

[0094] Spectrum center frequency (1kHz): reflects the central trend of the audio signal spectrum distribution, which is helpful for understanding the timbre characteristics of the signal.

[0095] Maximum delay node ID (Node3): identifies the node with the largest delay in the audio signal transmission path, which is crucial for predicting the delay of subsequent transmissions.

[0096] Average transmission speed (2Mbps): Indicates the average speed of audio signals during transmission, and is a key parameter for measuring transmission efficiency.

[0097] In the embodiment of the present application, the output prediction window refers to the time range of the voice data set required to transmit during a specific interactive session in the subsequent audio transmission phase, which is predicted based on the previous context information (including the previous context audio vector, etc.) during the audio signal transmission process. It represents the system's expectation of the time consumption of audio transmission and provides an important basis for formulating transmission strategies and scheduling resources.

[0098] For example, in a video conference, the system predicts that the time required to complete voice data transmission in the next interactive session (such as the user speaking period) is between 3 seconds and 5 seconds based on the previous context audio vector and other relevant information. This time range is the output prediction window. The system can allocate network resources and adjust encoding parameters in advance based on this prediction window to ensure the smoothness and real-time performance of audio transmission. In actual applications, the size and accuracy of the output prediction window will be affected by many factors, including but not limited to network conditions, device performance, audio signal characteristics, etc. Therefore, the system needs to continuously update the prediction model based on real-time data to improve the accuracy and reliability of the prediction.

[0099] Finally, step 104 focuses on using the output prediction window to intelligently obtain and generate a sound quality optimization strategy instruction set to achieve sound quality optimization and adjustment of the target wireless audio device in the subsequent audio transmission stage. First, the system will obtain several sound quality optimization strategy instructions corresponding to the sound quality preference label through the output prediction window that has been identified. These sound quality optimization strategy instructions are formulated based on the in-depth understanding and analysis of the sound quality preference label, and they contain specific operations and adjustment parameters required to achieve specific sound quality optimization. Secondly, the system will generate a sound quality optimization strategy instruction set containing these sound quality optimization strategy instructions. This instruction set is an ordered set, in which the instructions are arranged according to a specific logic and execution order to ensure the smooth progress of the sound quality optimization process. At the same time, the global time-consuming information of this instruction set is paired with the output prediction window, which means that the execution time of the instruction set will not exceed the time range set by the output prediction window, thereby ensuring the real-time and smoothness of audio transmission. Finally, this sound quality optimization strategy instruction set will be used to achieve sound quality optimization and adjustment of the target wireless audio device during the interactive session period in the subsequent audio transmission stage. By executing the instructions in the instruction set, the system can process and adjust the audio signal in real time to meet the specific sound quality preferences of the target wireless audio device, thereby providing a better quality and personalized audio experience. It can be seen that step 104 uses the output prediction window to obtain and generate the sound quality optimization strategy instruction set, thereby realizing the intelligent adjustment of the sound quality optimization of the target wireless audio device in the subsequent audio transmission stage. This technical solution not only improves the accuracy and efficiency of sound quality optimization, but also ensures the real-time and smoothness of audio transmission, bringing users a more excellent audio experience.

[0100] In summary, the technical solution of the present application realizes precise adjustment of the sound quality optimization in the subsequent audio transmission stage by comprehensively considering the behavior and characteristics of the target wireless audio device in the previous audio transmission stage. Specifically, the present application first obtains the audio signal transmission information of the target device in the same interactive session period in several previous audio transmission stages, which includes the previous sound quality status label and output node information, providing a rich data basis for subsequent sound quality preference identification and transmission time prediction. Then, the present application extracts the previous sound quality characterization vector based on the previous sound quality status label, and identifies the sound quality preference label of the target device based on this. This step ensures that the sound quality optimization can closely fit the personalized needs of the user. At the same time, through the extraction of the previous context audio vector, the present application also successfully identifies the output prediction window of the target device, which provides strict time constraints for the formulation of the sound quality optimization strategy and ensures the real-time and smoothness of audio transmission. Finally, the application obtains several sound quality optimization strategy instructions according to the output prediction window, and generates a sound quality optimization strategy instruction set matched with it, realizing efficient and accurate optimization and adjustment of the sound quality of the target wireless audio device during the interactive session period. In summary, the application significantly improves the personalization level, accuracy and efficiency of sound quality optimization, and brings users a better quality and smoother audio experience.

[0101] In some optional embodiments, identifying the sound quality preference label corresponding to the target wireless audio device based on the previous sound quality characterization vector includes: obtaining initial sound quality characterization vectors extracted from multiple initial sound quality labels; for each of the initial sound quality labels, identifying the sound quality feature weight corresponding to the initial sound quality label based on the initial sound quality characterization vector of the initial sound quality label and the previous sound quality characterization vector; the sound quality feature weight characterizes the possibility that the speech data set to be transmitted includes data of the initial sound quality label; and selecting a sound quality preference label from the multiple initial sound quality labels, wherein the sound quality feature weight of the sound quality preference label meets the sound quality preference decision requirements.

[0102] In this embodiment, the process of identifying the sound quality preference label corresponding to the target wireless audio device based on the preceding sound quality characterization vector is a complex process involving multi-dimensional vector analysis, weight calculation and decision making.

[0103] First, the system needs to obtain the initial sound quality characterization vectors extracted from a variety of initial sound quality labels. These initial sound quality labels are predefined and represent various sound quality characteristics that audio signals may have, such as clarity, timbre, dynamic range, etc. Each sound quality label corresponds to a unique sound quality characterization vector, which is extracted through machine learning or signal processing algorithms based on a large amount of audio data and can accurately reflect the characteristics of the sound quality label.

[0104] Secondly, for each initial sound quality label, the system needs to identify the sound quality feature weight corresponding to the initial sound quality label based on the initial sound quality characterization vector of the initial sound quality label and the previous sound quality characterization vector. This weight is a quantitative indicator that represents the possibility that the speech data set to be transmitted includes the data of the initial sound quality label. In other words, this weight reflects the preference of the target wireless audio device for a specific sound quality feature in the subsequent audio transmission stage.

[0105] When calculating the weight of the sound quality feature, the system can use various machine learning algorithms or similarity calculation methods. For example, you can use cosine similarity, Euclidean distance, or Manhattan distance and other measurement methods to calculate the similarity or distance between the previous sound quality characterization vector and various initial sound quality characterization vectors. Then, according to the size of the similarity or distance, a corresponding weight is assigned to each initial sound quality label. The higher the similarity or the smaller the distance, the closer the previous sound quality characterization vector is to a certain initial sound quality characterization vector, and therefore the corresponding sound quality feature weight is greater.

[0106] Then, the system needs to select a sound quality preference label from a variety of initial sound quality labels. This selection process is based on the sound quality feature weights, and the system will filter labels according to the preset sound quality preference decision requirements. The sound quality preference decision requirements can be based on weighted threshold judgments or selection strategies based on weight rankings. For example, the system can set a threshold, and only when the sound quality feature weight of a certain initial sound quality label exceeds this threshold, it will be considered a sound quality preference label. Alternatively, the system can sort the initial sound quality labels according to the size of the sound quality feature weights, and then select the first few labels with the largest weights as sound quality preference labels.

[0107] The selected sound quality preference tags represent the sound quality preferences of the target wireless audio device in the subsequent audio transmission stage. These tags will be used to guide the subsequent formulation and execution of sound quality optimization strategies. For example, if the selected sound quality preference tags include "clarity" and "timbre", the system will give priority to the sound quality optimization of these two aspects in the subsequent audio transmission stage to ensure that the transmitted voice data set can meet the specific requirements of the target wireless audio device for clarity and timbre.

[0108] In this way, by deeply analyzing the relationship between the previous sound quality characterization vector and various initial sound quality characterization vectors, the sound quality preference labels of the target wireless audio device are accurately identified. These labels provide clear directions and goals for subsequent sound quality optimization, so that the sound quality optimization strategy can better meet the actual needs of users, thereby improving the overall effect of audio transmission and user experience.

[0109] Furthermore, the implementation of this embodiment also relies on the support of some key technologies. For example, when extracting the preamble sound quality characterization vector and the initial sound quality characterization vector, it is necessary to use advanced audio signal processing algorithms and feature extraction techniques to ensure that the extracted vectors can accurately reflect the sound quality characteristics. When calculating the sound quality feature weights, it is necessary to select a suitable machine learning algorithm or similarity calculation method, and adjust and optimize the parameters according to the actual situation to improve the accuracy and stability of the weight calculation. When selecting sound quality preference labels, it is necessary to formulate reasonable sound quality preference decision requirements, and flexibly adjust them according to the actual application scenarios to ensure that the selected labels can truly reflect the user's sound quality preferences.

[0110] In addition, in practical applications, the initial sound quality label can be added or deleted as needed to adapt to different sound quality optimization needs and scenarios. At the same time, this embodiment can also be combined with other sound quality optimization technologies to form a more complete sound quality optimization solution. For example, this embodiment can be combined with sound quality optimization technology based on user feedback, and the sound quality preference label and sound quality optimization strategy can be further adjusted and optimized by collecting actual user feedback to achieve a more personalized and accurate sound quality optimization effect.

[0111] In summary, by deeply analyzing the relationship between the preceding sound quality characterization vector and various initial sound quality characterization vectors, the sound quality preference label of the target wireless audio device is accurately identified, and a clear direction and goal are provided for subsequent sound quality optimization. This embodiment has good feasibility and practicality, and can significantly improve the overall effect of audio transmission and user experience.

[0112] In some other optional embodiments, the method of obtaining several sound quality optimization strategy instructions of the sound quality preference label through the output prediction window includes: obtaining an initial speech data set belonging to the sound quality preference label, and determining the time interval characteristics corresponding to each initial speech data in the initial speech data set; performing at least two types of integration on the initial speech data in the initial speech data set to form at least two speech data clusters, each speech data cluster including at least two of the initial speech data; determining the initial global time interval characteristics corresponding to each speech data cluster based on the time interval characteristics of the initial speech data in each of the speech data clusters; determining the speech data cluster corresponding to the initial global time interval characteristics paired with the output prediction window, and determining several sound quality optimization strategy instructions of the sound quality preference label based on the initial speech data included in the determined speech data cluster.

[0113] In this embodiment, the process of obtaining several sound quality optimization strategy instructions of the sound quality preference label by outputting the prediction window is a complex process involving speech data set analysis, data integration, time series feature matching and strategy instruction generation.

[0114] First, the system needs to obtain an initial voice data set that belongs to a sound quality preference label. This set contains various voice data corresponding to the sound quality preference label, and these voice data have the sound quality characteristics represented by the sound quality preference label. For example, if the sound quality preference label is "clarity", then the voice data in the initial voice data set should have higher clarity.

[0115] Secondly, the system needs to determine the time series interval features corresponding to each initial speech data in the initial speech data set. The time series interval features refer to the characteristics of the distribution and duration of speech data on the time axis, which reflects the time series structure and dynamic changes of speech data. By extracting the time series interval features of each initial speech data, the system can conduct in-depth analysis and understanding of the time series characteristics of speech data.

[0116] Next, the system needs to perform at least two types of integration on the initial speech data in the initial speech data set to form at least two speech data clusters. Each speech data cluster includes at least two initial speech data, and these speech data have similarities or associations in certain features or attributes. By integrating the initial speech data to form speech data clusters, the system can further explore the internal connections and rules between the speech data.

[0117] Then, the system needs to determine the initial global temporal interval features corresponding to each voice data cluster based on the temporal interval features of the initial voice data in each voice data cluster. The initial global temporal interval features are obtained by synthesizing and abstracting the temporal interval features of all the initial voice data in the voice data cluster, which reflects the temporal characteristics and structure of the voice data cluster as a whole.

[0118] Next, the system needs to determine the speech data clusters that correspond to the initial global temporal interval features that are paired with the output prediction window. This step is based on temporal feature matching. The system will search for clusters whose initial global temporal interval features match the temporal features of the output prediction window among all speech data clusters. This matching is based on the similarity or consistency of the temporal features. For example, the distribution and duration of two temporal features on the time axis may be very close or exactly the same.

[0119] Finally, the system needs to determine several sound quality optimization strategy instructions for the sound quality preference tag based on the initial voice data included in the determined voice data cluster. These instructions are formulated based on the sound quality characteristics and timing characteristics of the initial voice data in the voice data cluster, and they can guide the subsequent sound quality optimization process to ensure that the transmitted voice data set can meet the specific requirements of the target wireless audio device for the sound quality preference tag. For example, if the sound quality preference tag is "clarity", then the sound quality optimization strategy instructions may include increasing the high-frequency components of the audio, reducing noise and interference, etc.

[0120] This design deeply analyzes the time series interval characteristics of the initial voice data and integrates them into voice data clusters. Then, based on the time series feature matching, the voice data clusters that match the output prediction window are determined, and finally the sound quality optimization strategy instructions are generated. This process not only takes into account the requirements of the sound quality preference label, but also fully considers the time series characteristics and dynamic changes of the voice data, so that a more accurate and effective sound quality optimization strategy can be formulated.

[0121] In a replaceable embodiment, the generation of a sound quality optimization strategy instruction set including the several sound quality optimization strategy instructions includes: determining default sound quality element characteristics corresponding to the several sound quality optimization strategy instructions respectively; determining target sound quality element characteristics of the sound quality optimization strategy instruction set to be generated; based on the default sound quality element characteristics corresponding to the several sound quality optimization strategy instructions respectively and the target sound quality element characteristics, the several sound quality optimization strategy instructions are subjected to sound quality element fusion to obtain a sound quality optimization strategy instruction set.

[0122] This embodiment can be understood as a technical solution for generating a sound quality optimization strategy instruction set including a plurality of sound quality optimization strategy instructions. The core lies in how to effectively integrate and deploy these strategy instructions to achieve the expected sound quality optimization effect. In order to achieve this goal, this embodiment adopts a series of orderly and logically rigorous steps, which together constitute a complete technical process.

[0123] First, it is necessary to clarify the default sound quality factor characteristics corresponding to each sound quality optimization strategy instruction. In short, sound quality factor characteristics refer to the key factors that can determine or affect the sound quality, which may include but are not limited to the frequency response, dynamic range, signal-to-noise ratio, distortion, etc. of the audio. In this step, through in-depth analysis of each strategy instruction, their respective default and inherent sound quality factor characteristics can be revealed and recorded. This step is fundamental, and it provides the necessary reference and basis for the subsequent fusion of sound quality elements.

[0124] Next, it is necessary to determine the target sound quality element characteristics of the sound quality optimization strategy instruction set to be generated. This step is goal-oriented, and it requires a clear understanding of what sound quality effect is expected to be achieved by integrating these strategy instructions. The setting of target sound quality element characteristics should be based on the needs of actual application scenarios, such as pursuing higher audio clarity, emphasizing lower noise levels, or hoping to reduce data transmission as much as possible while maintaining sound quality. After clarifying the target sound quality element characteristics, there is a standard for measuring whether the sound quality optimization strategy instruction set is effective.

[0125] With the preparations in the above two steps, we can enter the core link of the fusion of sound quality elements. This step is based on the default sound quality element features corresponding to several sound quality optimization strategy instructions, as well as the target sound quality element features that have been set. The process of fusion of sound quality elements is actually a process of finding the optimal solution. It requires that while retaining the advantages of each strategy instruction, the conflicts and adverse effects between them should be eliminated or reduced as much as possible to achieve the optimization of the overall sound quality.

[0126] When integrating sound quality elements, a variety of different technologies and methods can be used. For example, mathematical models or algorithms can be used to quantitatively analyze and compare the sound quality element characteristics of each strategy instruction to find the best integration solution. Alternatively, machine learning or artificial intelligence technology can be used to allow the computer to automatically learn and adjust the combination of these strategy instructions to achieve the best sound quality effect. Regardless of which method is used, the core goal is the same, which is to achieve effective integration of sound quality elements and generate a sound quality optimization strategy instruction set that meets specific needs.

[0127] It is worth mentioning that the process of fusion of sound quality elements is often not a one-time process. It may require multiple iterations and optimizations to achieve the best results. In each iteration, it is necessary to evaluate the current fusion result, analyze the gap between it and the characteristics of the target sound quality elements, and then adjust the fusion strategy and method accordingly, in order to achieve better results in the next iteration. Although this iterative optimization process may require more time and computing resources, it is the key to ensuring the effectiveness of the sound quality optimization strategy instruction set.

[0128] Finally, when the process of fusion of sound quality elements reaches a satisfactory convergence state, the final sound quality optimization strategy instruction set can be obtained. This instruction set is an organic combination of several sound quality optimization strategy instructions. It not only retains the advantages of each instruction, but also eliminates the adverse effects between them through effective fusion, thereby achieving the optimization of the overall sound quality. In practical applications, this instruction set can be used to guide audio processing and encoding to achieve multiple effects such as improving sound quality, reducing noise, and reducing data transmission volume.

[0129] It can be seen that through a series of orderly and logically rigorous steps such as determining the default sound quality element characteristics of the sound quality optimization strategy instructions, setting the target sound quality element characteristics, and performing sound quality element fusion, the effective integration and deployment of several sound quality optimization strategy instructions are achieved. This technical solution is not only clearly goal-oriented, but also ensures the effectiveness of the sound quality optimization strategy instruction set through iterative optimization. In practical applications, it can significantly improve the audio processing and encoding effects and meet diverse sound quality optimization needs.

[0130] In another replaceable embodiment, the default sound quality element characteristics corresponding to the several sound quality optimization strategy instructions and the target sound quality element characteristics are respectively used to fuse the several sound quality optimization strategy instructions to obtain a sound quality optimization strategy instruction set, including: determining a first sound quality optimization strategy instruction and a second sound quality optimization strategy instruction from the several sound quality optimization strategy instructions, the default sound quality element characteristics of the first sound quality optimization strategy instruction having a sound quality element fusion relationship with the target sound quality element characteristics, and the default sound quality element characteristics of the second sound quality optimization strategy instruction not having a sound quality element fusion relationship with the target sound quality element characteristics; converting the second sound quality optimization strategy instruction into the target sound quality element characteristics based on a sound quality element mapping vector from the default sound quality element characteristics of the second sound quality optimization strategy instruction to the target sound quality element characteristics; fusing the first sound quality optimization strategy instruction and the second sound quality optimization strategy instruction of the target sound quality element characteristics to obtain a sound quality optimization strategy instruction set.

[0131] Based on this embodiment, the audio transmission processing system is faced with a technical solution on how to fuse the sound quality elements based on the default sound quality element features and target sound quality element features of several sound quality optimization strategy instructions to obtain a sound quality optimization strategy instruction set. The core of this technical solution is how to effectively fuse the sound quality element features of these strategy instructions to achieve the expected sound quality optimization effect.

[0132] First, the audio transmission processing system needs to determine the first sound quality optimization strategy instruction and the second sound quality optimization strategy instruction from several sound quality optimization strategy instructions. The selection of these two instructions is based on their relationship with the target sound quality element characteristics. Specifically, the default sound quality element characteristics of the first sound quality optimization strategy instruction have a sound quality element fusion relationship with the target sound quality element characteristics, which means that their sound quality element characteristics are similar or complementary to some extent, so it is possible to further improve the sound quality through fusion. The default sound quality element characteristics of the second sound quality optimization strategy instruction do not have a sound quality element fusion relationship with the target sound quality element characteristics, which means that there may be differences or conflicts in their sound quality element characteristics, so some conversion or adjustment is required to merge with the target sound quality element characteristics.

[0133] After determining these two sound quality optimization strategy instructions, the audio transmission processing system needs to deal with the mismatch between the second instruction and the target sound quality element characteristics. Here, the audio transmission processing system converts the sound quality element mapping vector from the default sound quality element characteristics of the second sound quality optimization strategy instruction to the target sound quality element characteristics. The sound quality element mapping vector is a mathematical model or algorithm that describes the conversion relationship between sound quality element characteristics, which can help the audio transmission processing system convert the default sound quality element characteristics of the second instruction into a form that matches the target sound quality element characteristics. Through this conversion process, the audio transmission processing system can eliminate the mismatch between the second instruction and the target sound quality element characteristics, so that it can be better integrated with the first instruction.

[0134] After completing the conversion of the second sound quality optimization strategy instruction, the audio transmission processing system can proceed to the last step of sound quality element fusion. This step is to fuse the sound quality elements of the first sound quality optimization strategy instruction and the second sound quality optimization strategy instruction that has been converted into the target sound quality element characteristics. The fusion process may involve a variety of technologies and methods, such as mathematical models, algorithms, machine learning, etc. The specific method selected depends on the specific needs and scenarios of sound quality optimization. Through fusion, the audio transmission processing system can effectively combine the sound quality element characteristics of the two instructions to form a new instruction set with better sound quality optimization effect.

[0135] It is worth noting that the process of fusion of sound quality elements is often not a one-time process. It may require multiple iterations and optimizations to achieve the best results. In each iteration, the audio transmission processing system needs to evaluate the current fusion result, analyze the gap between it and the characteristics of the target sound quality elements, and then adjust the fusion strategy and method accordingly, in order to achieve better results in the next iteration. Although this iterative optimization process may require more time and computing resources, it is the key to ensuring the effectiveness of the sound quality optimization strategy instruction set.

[0136] In addition, in the process of integrating sound quality elements, the audio transmission processing system needs to consider some other factors, such as the mutual influence between sound quality optimization strategy instructions, the influence of the integrated instruction set on audio processing and encoding, etc. These factors may have an important impact on the final sound quality optimization effect, so they need to be fully considered and weighed in the integration process.

[0137] Finally, when the process of fusion of sound quality elements reaches a satisfactory convergence state, the audio transmission processing system can obtain the final sound quality optimization strategy instruction set. This instruction set is an organic combination of several sound quality optimization strategy instructions. It not only retains the advantages of each instruction, but also eliminates the adverse effects between them through effective fusion, thereby achieving the optimization of the overall sound quality. In practical applications, this instruction set can be used to guide audio processing and encoding to achieve multiple effects such as improving sound quality, reducing noise, and reducing data transmission. At the same time, because it fully considers the fusion relationship and mapping vectors between sound quality elements, it has better adaptability and flexibility, and can adapt to different sound quality optimization needs and scenarios.

[0138] Therefore, the technical solution recorded in this embodiment realizes the effective integration and deployment of several sound quality optimization strategy instructions by determining the first and second sound quality optimization strategy instructions, converting according to the sound quality element mapping vector, and performing sound quality element fusion, etc. A series of orderly and logically rigorous steps. This technical solution not only has a clear goal orientation, but also ensures the effectiveness of the sound quality optimization strategy instruction set through iterative optimization. In practical applications, it can significantly improve the processing and encoding effects of audio and meet diverse sound quality optimization needs. At the same time, because it fully considers the fusion relationship and mapping vector between sound quality elements, it has better adaptability and flexibility, and can bring more innovations and breakthroughs to the field of audio processing and coding.

[0139] In some examples, the default sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features, or spatial stereoscopic element features; the target sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features, or spatial stereoscopic element features; the sound quality optimization strategy instruction set includes at least one of clarity optimization instruction set, fullness optimization instruction set, balance optimization instruction set, timbre optimization instruction set, or spatial stereoscopic optimization instruction set.

[0140] It can be understood that this example is about how to fuse sound quality elements based on the default sound quality element features and target sound quality element features of several sound quality optimization strategy instructions, so as to obtain a sound quality optimization strategy instruction set. In order to deeply understand this technical solution, it is first necessary to clarify the specific contents of the default sound quality element features and the target sound quality element features, as well as the composition of the sound quality optimization strategy instruction set.

[0141] First of all, the default sound quality element features refer to the sound quality element characteristics that are inherent and unchanged in the sound quality optimization strategy instructions themselves. These characteristics are important indicators for measuring the quality of sound, and they directly determine the auditory performance of the audio signal. In the example, the default sound quality element features include at least one or more of the clarity element features, the fullness element features, the balance element features, the timbre element features, or the spatial three-dimensional element features. The clarity element features focus on the clarity of the audio signal, that is, the ability to resolve and layer the sound details; the fullness element features emphasize the fullness of the audio, that is, the thickness and fullness of the sound; the balance element features involve the balanced distribution of energy in each frequency band in the audio signal, ensuring that the sounds of different frequency bands can be properly displayed; the timbre element features focus on the timbre performance of the audio, that is, the color and charm of the sound; and the spatial three-dimensional element features emphasize the spatial sense and stereoscopic sense of the audio, making the listener feel as if they are in a real sound field environment.

[0142] Secondly, the target sound quality element characteristics are the sound quality effects that the audio transmission processing system hopes to achieve through the sound quality optimization strategy instruction set. It represents the expectations and goals of the audio transmission processing system for sound quality optimization and is an important basis for guiding the fusion of sound quality elements. In the example, the target sound quality element characteristics also include one or more of clarity element characteristics, fullness element characteristics, balance element characteristics, timbre element characteristics, or spatial stereoscopic element characteristics. This means that the audio transmission processing system may want to improve the clarity of the audio, enhance the fullness of the sound, adjust the balance of the frequency band, change the timbre performance, or enhance the spatial stereoscopic sense to meet specific sound quality requirements.

[0143] Then, the sound quality optimization strategy instruction set is the specific means and methods to achieve sound quality optimization. It contains an instruction set for optimizing different sound quality element features. By executing these instructions, the audio transmission processing system can process the audio signal in a targeted manner to achieve the expected sound quality effect. In the example, the sound quality optimization strategy instruction set includes at least one or more of a clarity optimization instruction set, a fullness optimization instruction set, a balance optimization instruction set, a timbre optimization instruction set, or a spatial stereoscopic optimization instruction set. These instruction sets correspond to different sound quality element features, and by executing the corresponding instruction sets, the audio transmission processing system can process the audio signal to improve clarity, enhance fullness, adjust balance, change timbre, or enhance spatial stereoscopic effect.

[0144] After clarifying the default sound quality element characteristics, target sound quality element characteristics and sound quality optimization strategy instruction set, the audio transmission processing system can perform specific operations on sound quality element fusion. The core of this process is how to effectively integrate and deploy several sound quality optimization strategy instructions to achieve the expected sound quality optimization effect.

[0145] First, the audio transmission processing system needs to determine the instructions that have a sound quality element fusion relationship with the target sound quality element characteristics from a number of sound quality optimization strategy instructions. The default sound quality element characteristics of these instructions are similar or complementary to the target sound quality element characteristics to some extent, so it is possible to further improve the sound quality through fusion. For example, if the target sound quality element characteristic is to improve clarity, then the audio transmission processing system needs to select those sound quality optimization strategy instructions whose default sound quality element characteristics contain clarity element characteristics.

[0146] Secondly, for instructions that do not have a sound quality element fusion relationship with the target sound quality element characteristics, the audio transmission processing system needs to convert or adjust. This can be achieved by mapping the sound quality element characteristics from the default sound quality element characteristics of these instructions to the target sound quality element characteristics. The sound quality element mapping vector is a mathematical model or algorithm that describes the conversion relationship between sound quality element characteristics, which can help the audio transmission processing system convert the default sound quality element characteristics of these instructions into a form that matches the target sound quality element characteristics. Through this conversion process, the audio transmission processing system can eliminate the mismatch between these instructions and the target sound quality element characteristics, so that it can be better integrated with other instructions.

[0147] Then, the audio transmission processing system fuses the instructions with the sound quality element fusion relationship and the instructions that have been converted into the target sound quality element features. The fusion process may involve a variety of technologies and methods, such as mathematical models, algorithms, machine learning, etc. The specific method to be selected depends on the specific needs and scenarios of sound quality optimization. Through fusion, the audio transmission processing system can effectively combine the sound quality element features of these instructions to form a new instruction set with better sound quality optimization effect.

[0148] It is worth noting that the process of fusion of sound quality elements is often not a one-time process. It may require multiple iterations and optimizations to achieve the best results. In each iteration, the audio transmission processing system needs to evaluate the current fusion result, analyze the gap between it and the characteristics of the target sound quality elements, and then adjust the fusion strategy and method accordingly, in order to achieve better results in the next iteration. Although this iterative optimization process may require more time and computing resources, it is the key to ensuring the effectiveness of the sound quality optimization strategy instruction set.

[0149] In addition, in the process of integrating sound quality elements, the audio transmission processing system also needs to consider some other factors, such as the mutual influence between the sound quality optimization strategy instructions, the impact of the integrated instruction set on audio processing and encoding, etc. These factors may have an important impact on the final sound quality optimization effect, so they need to be fully considered and weighed in the integration process. For example, the audio transmission processing system needs to ensure that the integrated instruction set does not cause excessive processing or distortion of the audio signal, and also needs to consider its impact on audio coding efficiency and compatibility.

[0150] Finally, when the process of fusion of sound quality elements reaches a satisfactory convergence state, the audio transmission processing system can obtain the final sound quality optimization strategy instruction set. This instruction set is an organic combination of several sound quality optimization strategy instructions. It not only retains the advantages of each instruction, but also eliminates the adverse effects between them through effective fusion, thereby achieving the optimization of the overall sound quality. In practical applications, this instruction set can be used to guide audio processing and encoding to achieve multiple effects such as improving sound quality, reducing noise, and reducing data transmission. At the same time, because it fully considers the fusion relationship and mapping vectors between sound quality elements, it has better adaptability and flexibility, and can adapt to different sound quality optimization needs and scenarios.

[0151] This design, by determining the specific content of the default sound quality element characteristics and the target sound quality element characteristics, as well as the composition of the sound quality optimization strategy instruction set, realizes the effective integration and deployment of several sound quality optimization strategy instructions. This technical solution not only has a clear goal orientation, but also ensures the effectiveness of the sound quality optimization strategy instruction set through iterative optimization. In practical applications, it can significantly improve the audio processing and encoding effects and meet diverse sound quality optimization needs. At the same time, because it fully considers the fusion relationship and mapping vectors between sound quality elements, it has better adaptability and flexibility, and can bring more innovations and breakthroughs to the field of audio processing and coding.

[0152] In an optional embodiment, the method of identifying the output prediction window corresponding to the target wireless audio device based on the previous context audio vector includes: obtaining a plurality of initial time consumption information; for each of the initial time consumption information, identifying the time consumption prediction weight corresponding to the initial time consumption information based on the initial time consumption information and the previous context audio vector; and selecting an output prediction window from the plurality of initial time consumption information, wherein the time consumption prediction weight corresponding to the output prediction window meets the set time consumption judgment requirement.

[0153] It can be understood that the core of this embodiment is to identify the output prediction window corresponding to the target wireless audio device based on the previous context audio vector. The implementation of this process not only involves in-depth analysis and processing of the audio vector, but also requires accurate prediction and discrimination of the time-consuming information. The following is a detailed description of the technical solution.

[0154] First, the audio transmission processing system needs to obtain some initial time information. This initial time information represents the time required for the target wireless audio device to process the audio signal under different conditions. These conditions may include the complexity of the audio signal, the processing power of the audio device, the current network status, etc. Obtaining this initial time information is the basis for identifying the output prediction window, which provides the audio transmission processing system with a preliminary estimate of the time required for the audio device to process the audio.

[0155] Secondly, for each initial time consumption information, the audio transmission processing system needs to identify the time consumption prediction weight corresponding to the initial time consumption information based on the initial time consumption information and the preceding context audio vector. This step is the key to the technical solution. The preceding context audio vector contains the preceding features of the audio signal, which are of great reference value for predicting the time required for the audio equipment to process the audio. By combining these features with the initial time consumption information, the audio transmission processing system can more accurately predict the time consumption of the audio equipment under different conditions. The time consumption prediction weight reflects the importance of the initial time consumption information in predicting the time consumption of the audio equipment. The higher the weight, the greater the role of the time consumption information in the prediction.

[0156] In the process of identifying the time consumption prediction weight, the audio transmission processing system may need to use some complex algorithms or models, such as machine learning algorithms, deep learning models, etc. These algorithms or models can automatically learn the mapping relationship between audio vectors and time consumption information, so as to achieve accurate prediction of time consumption. At the same time, the audio transmission processing system also needs to train and optimize the algorithm or model to improve the accuracy and stability of the prediction.

[0157] Then, the audio transmission processing system needs to select an output prediction window from several initial time consumption information. The output prediction window is a time range determined by the audio transmission processing system based on the time consumption prediction weight and the set time consumption discrimination requirements. It represents the most likely time required for the audio equipment to process the audio signal. The process of selecting the output prediction window is a comprehensive consideration process. The audio transmission processing system needs to comprehensively consider the prediction weights of each initial time consumption information and their matching degree with the set time consumption discrimination requirements.

[0158] The set time consumption discrimination requirement is a threshold or range set by the audio transmission processing system according to actual needs. It is used to determine which initial time consumption information is most likely to become the actual time consumption. When selecting the output prediction window, the audio transmission processing system will select the time range corresponding to the initial time consumption information with a high time consumption prediction weight and meeting the set time consumption discrimination requirement as the output prediction window. In this way, the audio transmission processing system can more accurately predict the time required for the audio equipment to process the audio signal, and provide strong support for subsequent audio processing and playback.

[0159] Finally, it is worth mentioning that the implementation of this technical solution also needs to consider some other factors and challenges. For example, the extraction and representation of audio vectors is a complex issue. The audio transmission processing system needs to select appropriate feature extraction methods and vector representation methods to ensure the accuracy and effectiveness of the audio vectors. At the same time, the acquisition and processing of time-consuming information is also a challenging task. The audio transmission processing system needs to design reasonable experiments and test plans to obtain accurate and reliable time-consuming data. In addition, the audio transmission processing system also needs to continuously optimize and improve the technical solution to adapt to different application scenarios and changes in demand.

[0160] It can be seen that by obtaining several pieces of initial time-consuming information, and identifying the time-consuming prediction weight for each piece of initial time-consuming information based on the previous context audio vector, and finally selecting the output prediction window that meets the set time-consuming discrimination requirements from the several pieces of initial time-consuming information, this technical solution not only realizes the accurate prediction of the time required for the audio equipment to process the audio, but also provides strong support for the subsequent audio processing and playback.

[0161] In some possible technical solutions, the generation of a sound quality optimization strategy instruction set including the several sound quality optimization strategy instructions includes: obtaining the target sound quality element characteristics indicated by the user APP corresponding to the target wireless audio device; based on the several sound quality optimization strategy instructions, generating a sound quality optimization strategy instruction set of the target sound quality element characteristics, the sound quality optimization strategy instruction set including the several sound quality optimization strategy instructions; the sound quality optimization strategy instruction set.

[0162] Based on this technical solution, generating a sound quality optimization strategy instruction set containing several sound quality optimization strategy instructions is a complex and delicate process. The core of this process is how to effectively generate a strategy instruction set that can improve the sound quality based on the target sound quality element characteristics indicated by the user APP corresponding to the target wireless audio device.

[0163] First, the audio transmission processing system needs to obtain the target sound quality element characteristics indicated by the user APP corresponding to the target wireless audio device. This step is the basis for generating a sound quality optimization strategy instruction set, because the audio transmission processing system must know clearly what sound quality effect the user expects to achieve. The target sound quality element characteristics may include clarity, fullness, balance, timbre, and spatial stereoscopic sense. These characteristics together constitute the user's overall perception of sound quality. Through the user APP, the audio transmission processing system can intuitively obtain the user's specific requirements and preferences for these sound quality elements.

[0164] Secondly, based on several sound quality optimization strategy instructions, the audio transmission processing system needs to generate a sound quality optimization strategy instruction set corresponding to the target sound quality element characteristics. This step is the key to the technical solution. It requires the audio transmission processing system to convert the user's target sound quality element characteristics into specific strategy instructions in order to optimize the sound quality of the audio equipment. The sound quality optimization strategy instruction set is a set of multiple instructions, each of which corresponds to a specific sound quality optimization strategy. These strategies may involve multiple links such as audio signal processing, encoding, decoding, and playback. By precisely controlling these links, the audio transmission processing system can achieve fine adjustment of the sound quality.

[0165] In the process of generating the sound quality optimization strategy instruction set, the audio transmission processing system needs to fully consider the mutual influence between the sound quality optimization strategy instructions and their comprehensive effect on the characteristics of the sound quality elements. Because different sound quality optimization strategies may have different effects on the same sound quality element, and sometimes such effects may be contradictory. Therefore, the audio transmission processing system needs to determine the specific impact of each strategy on the sound quality elements through a large number of experiments and data analysis, and how to effectively combine these strategies to achieve the best sound quality optimization effect.

[0166] At the same time, the audio transmission processing system also needs to pay attention to the impact of the sound quality optimization strategy instruction set on the performance of the audio device. Because sound quality optimization often requires a trade-off between the processing power, storage space, and battery life of the audio device. If the optimization strategy is too complex or the amount of calculation is too large, it may have a negative impact on the performance of the audio device. Therefore, when generating the sound quality optimization strategy instruction set, the audio transmission processing system needs to comprehensively consider the balance between sound quality optimization and audio device performance to ensure the practical feasibility and sustainability of the optimization strategy.

[0167] Then, when the audio transmission processing system has completed the generation of the sound quality optimization strategy instruction set, it can be applied to the target wireless audio device. By executing these instructions, the audio device can process and optimize the audio signal accordingly, thereby improving the sound quality. Users can perceive the changes in sound quality in real time through the user APP, and further adjust the sound quality optimization strategy instruction set according to their preferences and needs.

[0168] In addition, it is worth mentioning that the generation of the sound quality optimization strategy instruction set is not a static process. As users' demands for sound quality continue to change and the performance of audio equipment continues to improve, the audio transmission processing system needs to continuously update and optimize the sound quality optimization strategy instruction set. This may require the audio transmission processing system to conduct continuous research and development and innovation to explore new sound quality optimization strategies and technical means. At the same time, the audio transmission processing system also needs to maintain close communication and feedback mechanisms with users in order to promptly understand user needs and feedback, and make corresponding adjustments and improvements to the sound quality optimization strategy instruction set.

[0169] Finally, it should be emphasized that the generation and application of the sound quality optimization strategy instruction set is a systematic project, which involves knowledge and technology in multiple fields such as audio signal processing, machine learning, and user experience. In order to achieve the best sound quality optimization effect, the audio transmission processing system requires interdisciplinary cooperation and innovation to fully utilize the technical advantages and latest research results in various fields. At the same time, the audio transmission processing system also needs to pay attention to industry dynamics and technological development trends, so as to timely adjust and optimize the technical solutions of the audio transmission processing system to adapt to the ever-changing market and user needs.

[0170] Therefore, generating a sound quality optimization strategy instruction set containing several sound quality optimization strategy instructions is a complex and delicate process. It requires the audio transmission processing system to fully understand the needs and preferences of users, convert the characteristics of the user's target sound quality elements into specific strategy instructions, and achieve the best sound quality optimization effect under the premise of considering the performance of the audio equipment. At the same time, the audio transmission processing system also needs to continuously update and optimize the sound quality optimization strategy instruction set to adapt to the ever-changing market and user needs. Through interdisciplinary cooperation and innovation, the audio transmission processing system can provide users with a better and more personalized sound quality experience.

[0171] In other possible embodiments, the method also includes: sending the global time-consuming information to the user APP corresponding to the target wireless audio device; obtaining the global time-consuming information after the priority of the global time-consuming information is updated; changing the sound quality optimization strategy instructions included in the sound quality optimization strategy instruction set through the global time-consuming information after the priority update, and obtaining the sound quality optimization strategy instruction set that has been changed; the sound quality optimization strategy instruction set that has been changed is used to realize the sound quality optimization adjustment of the target wireless audio device in the interactive session period within the subsequent audio transmission stage.

[0172] In other possible embodiments, the method is not limited to the initial generation and application of the sound quality optimization strategy instruction set, but further involves real-time interaction with users and dynamic adjustment of the sound quality optimization strategy. Specifically, the method also includes sending global time-consuming information to the user APP corresponding to the target wireless audio device, and updating the priority based on the user's feedback on this information, and then changing the sound quality optimization strategy instruction set to achieve sound quality optimization adjustment of the target wireless audio device during the interactive session period in the subsequent audio transmission stage.

[0173] First, an important part of this embodiment is to send global time information to the user APP corresponding to the target wireless audio device. The purpose of this step is to enable users to understand the time consumption in the audio processing process in real time, thereby enhancing the user's perception and control of the audio processing progress. The global time information covers the time consumption of the entire process from audio signal input to sound quality optimization processing to final output, which provides users with a comprehensive overview of the time consumption.

[0174] Secondly, obtaining the global time-consuming information after the global time-consuming information is prioritized is a key step. Users may prioritize certain parts of the global time-consuming information according to their needs and preferences. For example, users may be more concerned about the real-time nature of audio processing and therefore want to reduce the time consumption of certain processing steps; or users may pay more attention to the sound quality and are willing to sacrifice some processing time in exchange for better sound quality. By prioritizing the global time-consuming information, the audio transmission processing system can better meet the personalized needs of users.

[0175] Then, based on the global time-consuming information after the priority update, the audio transmission processing system needs to change the sound quality optimization strategy instructions included in the sound quality optimization strategy instruction set. This step is the core of realizing dynamic adjustment of sound quality optimization. According to the user's priority update of the global time-consuming information, the audio transmission processing system needs to re-evaluate and adjust the various instructions in the sound quality optimization strategy instruction set. For example, if the user increases the priority of real-time performance, the audio transmission processing system may need to reduce some sound quality optimization strategy instructions with higher computational complexity; if the user pays more attention to the sound quality effect, the audio transmission processing system may need to add some strategy instructions that can improve the sound quality but may take longer. Through such adjustments, the audio transmission processing system can ensure that the sound quality optimization strategy instruction set is consistent with the user's needs and preferences.

[0176] The changed sound quality optimization strategy instruction set is the final output of the method, which will be used to optimize the sound quality of the target wireless audio device during the interactive session period in the subsequent audio transmission phase. In the subsequent audio transmission phase, the audio signal will be processed and optimized according to the changed sound quality optimization strategy instruction set to ensure that the final sound quality effect meets the user's expectations and needs. At the same time, since the sound quality optimization strategy instruction set is dynamically adjusted based on the user's real-time feedback and priority updates, it can better adapt to the changes in sound quality requirements of different users in different scenarios.

[0177] In addition, it is worth mentioning that this embodiment also emphasizes the real-time interaction and feedback mechanism with users. Through the user APP, users can understand the time consumption of audio processing anytime and anywhere, and update the priority according to their needs. This real-time interaction and feedback mechanism not only enhances the user's perception and control of the audio processing process, but also enables the sound quality optimization strategy instruction set to more accurately reflect the user's needs and preferences.

[0178] At the same time, in order to achieve the effectiveness and efficiency of this embodiment, the audio transmission processing system also needs to consider some key technical challenges and solutions. First, how to accurately obtain and calculate global time-consuming information is an important technical challenge. The audio transmission processing system needs to design accurate timing mechanisms and algorithms to measure the time consumption of each link in the audio processing process, and integrate them into global time-consuming information. Secondly, how to effectively change the sound quality optimization strategy instruction set according to the user's priority update is also a problem that needs to be solved. The audio transmission processing system needs to develop intelligent algorithms and models to analyze and understand the user's priority update requirements, and convert them into specific sound quality optimization strategy instruction adjustment plans.

[0179] Finally, it should be emphasized that this embodiment is not only a technical implementation method, but also a user-centered design concept and practice. By providing users with real-time global time-consuming information and priority update mechanism, the audio transmission processing system can allow users to participate more deeply in the sound quality optimization process, so that the sound quality optimization strategy instruction set is more in line with the user's personalized needs and expectations. At the same time, by constantly interacting with users and feedback loops, the audio transmission processing system can also continuously optimize and improve the sound quality optimization strategy instruction set to adapt to the ever-changing market demand and user needs.

[0180] It can be seen that by sending global time-consuming information to the user APP corresponding to the target wireless audio device, and updating the priority and changing the sound quality optimization strategy instruction set based on the user's feedback on this information, the sound quality optimization adjustment of the target wireless audio device is achieved during the interactive session period in the subsequent audio transmission stage. This method not only enhances the user's perception and control of the audio processing process, but also improves the personalization and accuracy of the sound quality optimization strategy instruction set. Through continuous interaction and feedback loops with users, the audio transmission processing system can continuously optimize and improve the sound quality optimization strategy instruction set to provide users with a better quality and personalized sound quality experience.

[0181] In some optional embodiments, the method further includes: obtaining audio signal transmission information samples corresponding to the wireless audio device samples in the same interactive session period samples in several preceding audio transmission stage samples; the audio signal transmission information samples include preceding sound quality status label samples to which preceding voice data samples output by the wireless audio device samples in the interactive session period samples belong, and output node information samples for outputting the preceding voice data samples; extracting sound quality characterization vector samples based on the preceding sound quality status label samples in the audio signal transmission information samples, and identifying sound quality preference label samples corresponding to the wireless audio device samples based on the sound quality characterization vector samples; the sound quality preference label samples characterize the sound quality indicators of voice data in voice data set samples to be transmitted by the interactive session period samples in the subsequent audio transmission stage samples; and extracting sound quality information based on the preceding sound quality status label samples in the audio signal transmission information samples. The output node information extracts the preceding context audio vector sample, and identifies the output prediction window sample corresponding to the wireless audio device sample based on the preceding context audio vector sample; the output prediction window sample characterizes the time consumption information of the speech data set sample required to be transmitted in the interactive session period sample in the subsequent audio transmission stage sample; through the output prediction window sample, obtains several sound quality optimization strategy instruction samples of the sound quality preference label sample; generates a sound quality optimization strategy instruction sample set including the several sound quality optimization strategy instruction samples, and the global time consumption information of the sound quality optimization strategy instruction sample set is paired with the output prediction window sample; obtains the prior sound quality optimization strategy instruction set corresponding to the wireless audio device sample in the subsequent audio transmission stage sample, and performs algorithm debugging based on the sound quality optimization strategy instruction sample set and the prior sound quality optimization strategy instruction set to obtain a sound quality optimization adjustment algorithm.

[0182] It can be understood that in the actual implementation process, the method is not only limited to the basic operation and control of the wireless audio equipment, but also goes further into the intelligent formulation and implementation of the sound quality optimization strategy. The core of this method is to extract the sound quality characterization vector and the preceding context audio vector by acquiring and analyzing the audio signal transmission information of the wireless audio equipment during the interactive conversation period in the preceding audio transmission stage, and then identify the sound quality preference label and output prediction window of the device. These two key parameters provide an important reference for the formulation of the sound quality optimization strategy in the subsequent audio transmission stage.

[0183] First, the method needs to obtain the audio signal transmission information samples corresponding to the wireless audio device samples in the same interactive session period samples in several preceding audio transmission stage samples. This step is the basis for subsequent analysis, which requires the audio transmission processing system to collect comprehensive data on the performance of the wireless audio device in the preceding audio transmission stage. The audio signal transmission information samples include the preceding sound quality status labels of the preceding voice data output by the wireless audio device during the interactive session period, as well as the output node information of the preceding voice data. This information provides the audio transmission processing system with detailed data on the sound quality performance and transmission behavior of the wireless audio device.

[0184] Secondly, based on the preceding sound quality status label sample in the audio signal transmission information sample, the audio transmission processing system can extract a sound quality characterization vector sample. The sound quality characterization vector is a quantitative description of the sound quality status of the wireless audio device, which includes various key factors that affect the sound quality. Through machine learning and other algorithms, the audio transmission processing system can extract these key factors from the sound quality status label and form a sound quality characterization vector. Furthermore, the audio transmission processing system can identify the sound quality preference label sample corresponding to the wireless audio device sample based on the sound quality characterization vector sample. The sound quality preference label characterizes the sound quality index of the voice data set that needs to be transmitted during the interactive conversation period in the subsequent audio transmission stage. The implementation of this step enables the audio transmission processing system to predict the preference and demand for sound quality in the subsequent stage based on the performance of the wireless audio device in the preceding audio transmission stage.

[0185] Then, the audio transmission processing system also needs to extract the preceding context audio vector sample based on the output node information in the audio signal transmission information sample. The preceding context audio vector is a quantitative description of the transmission behavior of the wireless audio device in the preceding audio transmission stage. It contains the context information of the audio transmission, such as the transmission speed, transmission volume, etc. By extracting the preceding context audio vector, the audio transmission processing system can further identify the output prediction window sample corresponding to the wireless audio device sample. The output prediction window characterizes the time-consuming information of the voice data set that needs to be transmitted during the interactive conversation period in the subsequent audio transmission stage. The implementation of this step enables the audio transmission processing system to predict the transmission behavior of the wireless audio device in the subsequent audio transmission stage, thereby providing a time reference for the formulation of sound quality optimization strategies.

[0186] Next, by outputting the prediction window sample, the audio transmission processing system can obtain several sound quality optimization strategy instruction samples of the sound quality preference label sample. The sound quality optimization strategy instruction is a specific operation instruction for optimizing the sound quality of the wireless audio device, which includes various parameters and methods for adjusting the sound quality. According to the sound quality preference label and the output prediction window, the audio transmission processing system can formulate a sound quality optimization strategy instruction that meets the sound quality requirements and transmission time requirements of the wireless audio device. Furthermore, the audio transmission processing system can generate a sound quality optimization strategy instruction sample set including several sound quality optimization strategy instruction samples. This instruction set is a specific implementation plan for the sound quality optimization strategy in the subsequent audio transmission stage.

[0187] In addition, in order to ensure the effectiveness and feasibility of the sound quality optimization strategy instruction sample set, the audio transmission processing system also needs to perform algorithm debugging on it and the prior sound quality optimization strategy instruction set corresponding to the wireless audio equipment sample in the subsequent audio transmission stage. The prior sound quality optimization strategy instruction set is a collection of sound quality optimization strategy instructions summarized by the audio transmission processing system based on past experience and data. It contains various possible sound quality optimization schemes and parameter settings. Through algorithm debugging, the audio transmission processing system can compare and analyze the sound quality optimization strategy instruction sample set with the prior instruction set, thereby optimizing and adjusting the parameters and methods in the instruction set to obtain the final sound quality optimization adjustment algorithm.

[0188] It is worth noting that by comprehensively collecting and analyzing the performance data of wireless audio equipment in the previous audio transmission stage, the audio transmission processing system can more accurately predict its sound quality requirements and transmission behavior in the subsequent stage. Furthermore, the audio transmission processing system can formulate a more personalized and accurate sound quality optimization strategy to meet the sound quality requirements of different devices in different scenarios. At the same time, through algorithm debugging and continuous optimization, the audio transmission processing system can also continuously improve the effectiveness and feasibility of the sound quality optimization strategy, providing users with a better quality and more efficient sound quality experience.

[0189] In this way, by acquiring and analyzing the audio signal transmission information of the wireless audio device during the interactive conversation period in the previous audio transmission stage, the sound quality characterization vector and the previous context audio vector are extracted, and then the sound quality preference label and the output prediction window are identified. These two key parameters provide an important reference for the formulation of the sound quality optimization strategy in the subsequent audio transmission stage. By formulating the sound quality optimization strategy instruction set and debugging the algorithm, the audio transmission processing system can obtain the final sound quality optimization adjustment algorithm to provide a personalized and accurate sound quality optimization solution for wireless audio equipment. This method not only improves the accuracy and effectiveness of sound quality optimization, but also brings users a better quality and more efficient sound quality experience.

[0190] In some independent embodiments, the algorithm is debugged based on the sound quality optimization strategy instruction sample set and the prior sound quality optimization strategy instruction set to obtain a sound quality optimization adjustment algorithm, including: determining a strategy instruction training error between the sound quality optimization strategy instruction sample set and the prior sound quality optimization strategy instruction set; obtaining a priori sound quality preference label corresponding to the wireless audio device sample in the subsequent audio transmission stage sample, and determining a sound quality preference training error between the sound quality preference label sample and the prior sound quality preference label; and algorithm debugging is performed based on the strategy instruction training error and the sound quality preference training error to obtain a sound quality optimization adjustment algorithm.

[0191] In some other independent embodiments, the algorithm is debugged according to the strategy instruction training error and the sound quality preference training error to obtain a sound quality optimization adjustment algorithm, including: obtaining the prior time consumption information corresponding to the sample of the wireless audio device in the subsequent audio transmission stage, and determining the time consumption training error between the output prediction window sample and the prior time consumption information; the algorithm is debugged according to the strategy instruction training error, the sound quality preference training error and the time consumption training error to obtain a sound quality optimization adjustment algorithm.

[0192] When discussing the above two core inventions, we first need to understand the core concept behind them: that is, to obtain a more accurate and efficient sound quality optimization and adjustment algorithm through continuous algorithm debugging and optimization. The realization of this goal depends on the in-depth analysis and comprehensive consideration of multiple key training errors.

[0193] First, in the first independent embodiment, the key to the technology is to determine the policy instruction training error between the sound quality optimization policy instruction sample set and the prior sound quality optimization policy instruction set, and simultaneously obtain the prior sound quality preference label corresponding to the wireless audio device sample in the subsequent audio transmission stage to determine the sound quality preference training error between the sound quality preference label sample and the prior sound quality preference label. The determination of these two training errors provides a comprehensive evaluation framework for the audio transmission processing system to measure the accuracy and effectiveness of the sound quality optimization policy instruction sample set and the fit of the sound quality preference label sample.

[0194] With these two key training errors as references, the audio transmission processing system can further debug the algorithm. Specifically, the audio transmission processing system will continuously adjust and optimize the sound quality optimization strategy instruction sample set based on the comprehensive consideration of the strategy instruction training error and the sound quality preference training error until an adjustment algorithm that can meet the expected sound quality optimization effect is obtained. This process not only ensures the accurate formulation of the sound quality optimization strategy, but also greatly improves its feasibility and effectiveness in practical applications.

[0195] Next, the audio transmission processing system turns to a second independent embodiment. In this embodiment, the focus of the technology is further expanded. In addition to considering the strategy instruction training error and the sound quality preference training error, a new dimension of time-consuming training error is introduced. Specifically, the audio transmission processing system obtains the prior time-consuming information corresponding to the wireless audio device sample in the subsequent audio transmission stage to determine the time-consuming training error between the output prediction window sample and the prior time-consuming information. The addition of this step enables the audio transmission processing system to more comprehensively evaluate the time-consuming performance of the sound quality optimization strategy in actual application, thereby ensuring that it not only has a significant sound quality optimization effect, but also meets strict time-consuming requirements.

[0196] With the three key evaluation indicators of strategy instruction training error, sound quality preference training error, and time-consuming training error, the audio transmission processing system can perform more complex and comprehensive algorithm debugging. Specifically, the audio transmission processing system will continuously adjust and optimize the sound quality optimization strategy instruction sample set based on the comprehensive consideration of these three training errors until an adjustment algorithm that can meet the expected sound quality optimization effect and strictly control the time consumption is obtained. This process not only greatly improves the accuracy and feasibility of the sound quality optimization strategy, but also lays a solid foundation for its widespread promotion in practical applications.

[0197] In summary, the above two core invention points are centered around the core concept of how to obtain a more accurate and efficient sound quality optimization and adjustment algorithm through algorithm debugging and optimization. The first embodiment debugs the algorithm by determining the strategy instruction training error and the sound quality preference training error, while the second embodiment further introduces a new evaluation dimension of time-consuming training error on this basis. The organic combination of these two embodiments not only provides a comprehensive and in-depth sound quality optimization strategy formulation framework for the audio transmission processing system, but also lays a solid foundation for its wide promotion and application in practical applications.

[0198] Further, Figure 2 FIG. 2 is a schematic diagram of the structure of an audio transmission processing system 200 provided in an embodiment of the present application. Figure 2 The audio transmission processing system 200 shown includes a processor 210, and the processor 210 can call and run a computer program from a memory to implement the method in the embodiment of the present application.

[0199] Alternatively, if Figure 2 As shown, the audio transmission processing system 200 may further include a memory 230. The processor 210 may call and run a computer program from the memory 230 to implement the method in the embodiment of the present application.

[0200] The memory 230 may be a separate device independent of the processor 210 , or may be integrated into the processor 210 .

[0201] Alternatively, if Figure 2 As shown, the audio transmission processing system 200 may further include a transceiver 220, and the processor 210 may control the transceiver 220 to interact with other devices, specifically, may send information or data to other devices, or receive information or data sent by other devices.

[0202] Optionally, the audio transmission processing system 200 may implement the corresponding processes corresponding to the storage engine or components in the storage engine (such as a processing module) or a device deployed with a storage engine in each method of the embodiments of the present application, which will not be described here for the sake of brevity.

[0203] It should be understood that the processor of the embodiment of the present application may be an integrated circuit chip with signal processing capabilities.

[0204] It is understood that the memory in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory. It should be noted that the memory of the systems and methods described herein is intended to include but is not limited to suitable types of memory.

[0205] Based on the above, a readable storage medium is provided, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the above method are implemented.

[0206] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0207] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a computer software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0208] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the present application, all of which are within the protection of the present application.

Claims

1. A wireless audio transmission method based on sound, characterized in that: The method is applied to an audio transmission processing system, and the method comprises: Acquire audio signal transmission information corresponding to the target wireless audio device in the same interactive session period in a plurality of preceding audio transmission stages; the audio signal transmission information includes a preceding sound quality status label to which preceding voice data output by the target wireless audio device in the interactive session period belongs, and output node information for outputting the preceding voice data; Extracting a preceding sound quality characterization vector according to the preceding sound quality status label in the audio signal transmission information, and identifying a sound quality preference label corresponding to the target wireless audio device according to the preceding sound quality characterization vector; the sound quality preference label characterizes a sound quality index of the voice data in the voice data set to be transmitted during the interactive session period in the subsequent audio transmission stage; Extracting a preceding context audio vector according to the output node information in the audio signal transmission information, and identifying an output prediction window corresponding to the target wireless audio device according to the preceding context audio vector; the output prediction window represents time-consuming information for transmitting a voice data set during the interactive session period in the subsequent audio transmission phase; Obtaining, through the output prediction window, a plurality of sound quality optimization strategy instructions of the sound quality preference label; Generate a sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions, wherein the global time consumption information of the sound quality optimization strategy instruction set is paired with the output prediction window; the sound quality optimization strategy instruction set is used to implement sound quality optimization adjustment of the target wireless audio device during the interactive session period within the subsequent audio transmission stage.

2. The method according to claim 1, characterized in that The step of identifying the sound quality preference label corresponding to the target wireless audio device according to the preceding sound quality characterization vector includes: Obtaining initial sound quality representation vectors extracted from a plurality of initial sound quality labels; For each of the initial sound quality labels, based on the initial sound quality characterization vector of the initial sound quality label and the preceding sound quality characterization vector, identify the sound quality feature weight corresponding to the initial sound quality label; the sound quality feature weight characterizes the possibility that the speech data set to be transmitted includes the data of the initial sound quality label; A sound quality preference label is selected from the multiple initial sound quality labels, and the sound quality feature weight of the sound quality preference label meets the sound quality preference decision requirement.

3. The method according to claim 1, characterized in that The step of obtaining a plurality of sound quality optimization strategy instructions of the sound quality preference label through the output prediction window includes: Acquire an initial voice data set belonging to the sound quality preference label, and determine a time series interval feature corresponding to each initial voice data in the initial voice data set; Performing at least two types of integration on the initial speech data in the initial speech data set to form at least two speech data clusters, each speech data cluster including at least two of the initial speech data; Determine the initial global time series interval feature corresponding to each of the voice data clusters according to the time series interval feature of the initial voice data in each of the voice data clusters; A speech data cluster corresponding to an initial global time series interval feature paired with the output prediction window is determined, and a plurality of sound quality optimization strategy instructions of the sound quality preference label are determined based on the initial speech data included in the determined speech data cluster.

4. The method according to claim 1, characterized in that: The generating of the sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions comprises: Determining the default sound quality element features corresponding to the plurality of sound quality optimization strategy instructions respectively; Determine target sound quality element characteristics of the sound quality optimization strategy instruction set to be generated; Based on the default sound quality element features and the target sound quality element features respectively corresponding to the plurality of sound quality optimization strategy instructions, the plurality of sound quality optimization strategy instructions are subjected to sound quality element fusion to obtain a sound quality optimization strategy instruction set; The method of fusing the sound quality elements of the plurality of sound quality optimization strategy instructions based on the default sound quality element features and the target sound quality element features respectively corresponding to the plurality of sound quality optimization strategy instructions to obtain a sound quality optimization strategy instruction set includes: Determine a first sound quality optimization strategy instruction and a second sound quality optimization strategy instruction from the plurality of sound quality optimization strategy instructions, wherein the default sound quality element feature of the first sound quality optimization strategy instruction has a sound quality element fusion relationship with the target sound quality element feature, and the default sound quality element feature of the second sound quality optimization strategy instruction does not have a sound quality element fusion relationship with the target sound quality element feature; Converting the second sound quality optimization strategy instruction into the target sound quality element feature according to a sound quality element mapping vector from the default sound quality element feature of the second sound quality optimization strategy instruction to the target sound quality element feature; The first sound quality optimization strategy instruction and the second sound quality optimization strategy instruction of the target sound quality element feature are integrated to obtain a sound quality optimization strategy instruction set; The default sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features or spatial three-dimensional element features; the target sound quality element features include at least one of clarity element features, fullness element features, balance element features, timbre element features or spatial three-dimensional element features; the sound quality optimization strategy instruction set includes at least one of clarity optimization instruction set, fullness optimization instruction set, balance optimization instruction set, timbre optimization instruction set or spatial three-dimensional optimization instruction set.

5. The method according to claim 1, characterized in that The step of identifying the output prediction window corresponding to the target wireless audio device according to the preceding context audio vector includes: Get some initial time-consuming information; For each of the initial time consumption information, based on the initial time consumption information and the preceding context audio vector, identifying a time consumption prediction weight corresponding to the initial time consumption information; An output prediction window is selected from the plurality of initial time-consuming information, and the time-consuming prediction weight corresponding to the output prediction window meets the set time-consuming discrimination requirement.

6. The method according to any one of claims 1 to 5, characterized in that: The generating of the sound quality optimization strategy instruction set including the plurality of sound quality optimization strategy instructions comprises: Obtaining target sound quality element features indicated by the user APP corresponding to the target wireless audio device; According to the plurality of sound quality optimization strategy instructions, a sound quality optimization strategy instruction set of the target sound quality element characteristics is generated, and the sound quality optimization strategy instruction set includes the plurality of sound quality optimization strategy instructions.

7. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: Sending the global time-consuming information to the user APP corresponding to the target wireless audio device; Obtaining the global time-consuming information after the global time-consuming information is prioritized; The sound quality optimization strategy instructions included in the sound quality optimization strategy instruction set are changed through the global time-consuming information after the priority is updated, so as to obtain a sound quality optimization strategy instruction set that has been changed; the sound quality optimization strategy instruction set that has been changed is used to realize the sound quality optimization adjustment of the target wireless audio device during the interactive session period in the subsequent audio transmission stage.

8. The method according to claim 1, characterized in that The method further comprises: Acquire audio signal transmission information samples corresponding to the wireless audio device sample in the same interactive session period samples in a plurality of preceding audio transmission phase samples; the audio signal transmission information samples include preceding sound quality status label samples belonging to preceding voice data samples output by the wireless audio device sample in the interactive session period samples, and output node information samples outputting the preceding voice data samples; Extracting a sound quality characterization vector sample according to the preceding sound quality status label sample in the audio signal transmission information sample, and identifying a sound quality preference label sample corresponding to the wireless audio device sample according to the sound quality characterization vector sample; the sound quality preference label sample characterizes the sound quality index of the voice data in the voice data set sample to be transmitted in the interactive session period sample in the subsequent audio transmission phase sample; Extracting a preceding context audio vector sample according to the output node information in the audio signal transmission information sample, and identifying an output prediction window sample corresponding to the wireless audio device sample according to the preceding context audio vector sample; the output prediction window sample represents the time consumption information of the speech data set sample required to be transmitted for the interactive session period sample in the subsequent audio transmission phase sample; Obtaining several sound quality optimization strategy instruction samples of the sound quality preference label samples through the output prediction window samples; Generate a sound quality optimization strategy instruction sample set including the plurality of sound quality optimization strategy instruction samples, wherein the global time consumption information of the sound quality optimization strategy instruction sample set is paired with the output prediction window sample; Obtain a priori sound quality optimization strategy instruction set corresponding to the wireless audio device sample in the subsequent audio transmission stage sample, perform algorithm debugging based on the sound quality optimization strategy instruction sample set and the priori sound quality optimization strategy instruction set, and obtain a sound quality optimization adjustment algorithm.

9. An audio transmission processing system, characterized in that: The method comprises at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the method according to any one of claims 1 to 8.

Images