Universal audio quality analysis method and system

Abstract

The invention provides a universal audio quality analysis method and system, and belongs to the technical field of audio, and the method comprises the steps: determining a standard audio quality index corresponding to a target scene; feature extraction is carried out on the target audio, multiple pieces of audio feature data are obtained, and the target audio is the audio in the target scene; and determining a target repair mode corresponding to the target scene based on the standard audio quality index and the multiple pieces of audio feature data, and repairing the to-be-transmitted audio based on the target repair mode. According to the universal audio quality analysis method and system provided by the invention, the accuracy of audio quality analysis can be improved.

Description

Technical Field

[0001] The present disclosure belongs to the field of audio technology, and more specifically, to a universal audio quality analysis method and system. Background Art

[0002] Audio quality analysis has a wide range of application needs in entertainment, communications, education, security and monitoring, industry, etc. With the continuous advancement of technology, audio quality analysis will play an important role in more fields, bringing more convenience and value to people's lives and work.

[0003] In real-time communication, live broadcast and other scenarios, the demand for real-time analysis and adaptive processing of audio quality is increasing. Especially in voice calls, audio quality will directly affect the clarity and intelligibility of the call. However, the existing quality analysis technology still has the problem of low accuracy, and the technology that repairs and adjusts the audio based on the analysis results is difficult to respond flexibly to different scenarios and needs. Summary of the invention

[0004] The purpose of the present disclosure is to provide a universal audio quality analysis method and system to improve the accuracy of audio quality analysis.

[0005] A first aspect of an embodiment of the present disclosure provides a universal audio quality analysis method, including: Determine the standard audio quality indicators corresponding to the target scenario.

[0006] Feature extraction is performed on the target audio to obtain multiple audio feature data, where the target audio is the audio in the target scene.

[0007] A target repair mode corresponding to a target scene is determined based on a standard audio quality index and a plurality of audio feature data, and the audio to be transmitted is repaired based on the target repair mode.

[0008] A second aspect of the embodiments of the present disclosure provides a universal audio quality analysis system, including: The first quality analysis module is used to determine a standard audio quality index corresponding to a target scene.

[0009] The second quality analysis module is used to extract features of the target audio to obtain multiple audio feature data, where the target audio is the audio in the target scene.

[0010] The audio repair module is used to determine a target repair mode corresponding to a target scene based on a standard audio quality index and multiple audio feature data, and repair the audio to be transmitted based on the target repair mode.

[0011] According to a third aspect of an embodiment of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of the above-mentioned universal audio quality analysis method when executing the computer program.

[0012] According to a fourth aspect of the embodiments of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the universal audio quality analysis method are implemented.

[0013] The beneficial effect of the universal audio quality analysis method and system provided by the embodiments of the present disclosure is that by determining the standard audio quality index and repairing it based on it, the quality parameters such as the clarity and fidelity of the audio in the target scene can be effectively improved, bringing a better listening experience to the user. Whether in scenarios such as remote voice calls, music playback or video conferencing, the audio effect can be significantly improved.

[0014] Corresponding standard audio quality indicators and repair modes are determined for different target scenarios, so that the method can be widely applied to various scenarios, improve the adaptability and pertinence to the environment, and improve the accuracy and flexibility of audio quality analysis and repair. For example, for quiet indoor environments, noisy outdoor environments, professional recording studios, and different users with different hearing, specific audio needs in different scenarios can be met.

[0015] After repairing the audio to be transmitted, adverse factors such as noise and distortion in the audio can be reduced, thereby reducing the transmission volume of audio data and improving transmission efficiency, which is of great significance especially when the network bandwidth is limited. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A flowchart of a universal audio quality analysis method provided by an embodiment of the present disclosure; Figure 2 A structural block diagram of a universal audio quality analysis system provided by an embodiment of the present disclosure; Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0018] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present disclosure. However, it should be clear to those skilled in the art that the present disclosure may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obstructing the description of the present disclosure with unnecessary details.

[0019] In order to make the purpose, technical solutions and advantages of the present disclosure more clear, specific embodiments will be described below in conjunction with the accompanying drawings.

[0020] Please refer to Figure 1 , Figure 1 This is a flow chart of a universal audio quality analysis method provided in an embodiment of the present disclosure. The method may include S101 to S103.

[0021] S101: Determine a standard audio quality index corresponding to a target scene.

[0022] In this embodiment, determining the standard audio quality index corresponding to the target scene includes: The initial audio quality indicator is updated based on the environmental information of the target scene to determine a first audio quality indicator.

[0023] The initial audio quality indicator is updated based on the user information of the target scenario to determine a second audio quality indicator.

[0024] The first audio quality indicator and the second audio quality indicator are integrated to determine a standard audio quality indicator.

[0025] In this embodiment, the standard audio quality index may include at least one of signal-to-noise ratio, total harmonic distortion, frequency response, audio dynamic range, clarity, timbre, and spatial sense (the spatial effect created by the audio, including the direction, distance, depth, etc. of the sound). Environmental information includes environmental noise information and environmental network information. User information includes user auditory information and user preference information. The initial audio quality index may be a more general audio quality index (fixed constant), which has the problem of being relatively broad and difficult to provide a more accurate reference index for specific scenarios and users. For example, some users have a poor auditory experience for audio that meets the initial audio quality index.

[0026] In this embodiment, the target scene can be clearly defined, and the environmental information and user information in the scene can be collected. For environmental information, the environmental noise information and environmental network information are analyzed. The environmental noise information can be sampled in the scene for a period of time by an audio acquisition device to determine the intensity, frequency distribution and other characteristics of the noise. Environmental network information can be obtained through network monitoring tools, including parameters such as network bandwidth, delay, and packet loss rate.

[0027] For user information, user hearing information and user preference information are collected. User hearing information can be obtained through user feedback questionnaires, professional hearing tests, or the establishment of human hearing models to understand the user's sensitivity to different audio quality aspects. User preference information can be obtained by analyzing the user's historical audio usage records in this scenario, such as favorite audio types, volume settings, etc.

[0028] The initial audio quality indicators are updated based on environmental information and user information respectively. For environmental noise information, if the noise is greater than the noise threshold (that is, the noise is large), the signal-to-noise ratio indicator of the audio needs to be improved. If the environmental network information shows that the network bandwidth is less than the bandwidth threshold (that is, the network bandwidth is low), the audio bit rate needs to be reduced to ensure smooth transmission, and the audio compression ratio and other indicators need to be adjusted. For user auditory information, if the user's sensitivity to high-frequency sounds is greater than the sensitivity threshold (that is, the user is more sensitive to high-frequency sounds), the frequency response indicator of the audio in the high frequency band needs to be improved. If there is a difference in the user's left and right ears, different standard audio quality indicators need to be determined for the auditory characteristics of the left and right ears. For user preference information, if the user likes a higher volume, the upper limit of the dynamic range of the audio needs to be adjusted.

[0029] The updated first audio quality index and the second audio quality index are merged. A weighted average method can be used to assign different weights according to the importance of environmental information and user information in different scenarios to determine the final standard audio quality index.

[0030] In this embodiment, determining the standard audio quality index corresponding to the target scene may also include: The initial audio quality indicator is updated based on the environmental information of the target scene to determine a first audio quality indicator.

[0031] The first audio quality indicator is updated based on the user information of the target scenario to determine a standard audio quality indicator.

[0032] In this embodiment, determining the standard audio quality index corresponding to the target scene may also include: The initial audio quality indicator is updated based on the user information of the target scenario to determine a second audio quality indicator.

[0033] The second audio quality indicator is updated based on the environmental information of the target scene to determine the standard audio quality indicator.

[0034] For example, in a voice call scenario, firstly, the target scenario is clarified as a voice call, and then the environment information and user information in the scenario are collected.

[0035] For environmental information: use the audio acquisition device to sample the user's environment for a period of time. If the user is on a noisy street, the sampling result shows that the environmental noise intensity is greater than the noise intensity threshold (that is, the environmental noise intensity is large) and the frequency distribution is mainly concentrated in the medium and low frequency bands. At this time, according to the environmental noise information, it is necessary to improve the audio signal-to-noise ratio index, for example, by enhancing the strength of the voice signal or using a more advanced noise suppression algorithm to reduce the impact of environmental noise.

[0036] At the same time, obtain environmental network information through network monitoring tools. If the network bandwidth is lower than the bandwidth threshold, the delay is higher than the delay threshold (i.e. the delay is high), and the packet loss rate is higher than the packet loss rate threshold (i.e. the packet loss rate is high), then it is necessary to reduce the audio bit rate to ensure smooth transmission of the call, and adjust the audio compression ratio to reduce the amount of data while maintaining the intelligibility of the voice as much as possible.

[0037] For user information: Understand user hearing information through user feedback questionnaires and professional hearing tests. If the user has low recognition or sensitivity to high-frequency sounds during voice calls, it may be due to age or hearing problems. Then it is necessary to appropriately increase the frequency response index of the audio in the high frequency band to make the high-frequency part of the voice clearer. If it is detected that there are differences in the user's left and right ears, such as the left ear is more sensitive to low frequencies and the right ear is more sensitive to high frequencies, then different standard audio quality indicators are determined for the hearing characteristics of the left and right ears, and different audio enhancement strategies are provided for the left and right ears.

[0038] Analyze the user's historical usage records in voice call scenarios to obtain user preference information. If the user often sets a high volume during calls, it is necessary to adjust the upper limit of the audio dynamic range to meet the user's demand for high volume while avoiding sound distortion.

[0039] Finally, the first audio quality index and the second audio quality index are fused by weighted average. In the voice call scenario, if the environmental noise is greater than the noise threshold and the network stability is less than the stability threshold, the weight of the environmental information can be relatively high. If the user has specific requirements for audio quality and has obvious historical preferences, the weight of the user information will be increased accordingly. By reasonably allocating weights, the standard audio quality index that best suits the voice call scenario is determined, thereby improving the quality of voice calls and user experience.

[0040] Exemplarily, the hearing test can be conducted through pure tone audiometry, where the user wears professional headphones and listens to pure tones of different frequencies to determine the sensitivity threshold to sounds of each frequency. If the user has a higher hearing threshold at certain frequencies, it means that the perception of the sound at that frequency is weak. Speech recognition tests can also be conducted, where words and sentences of different volumes and clarity are played, and the user is asked to repeat or answer related questions to assess their ability to understand speech under different conditions. In addition, binaural hearing tests are conducted, where different sounds are played to the left and right ears at the same time, allowing the user to judge the difference and understand the hearing differences between the left and right ears. These test results can provide a basis for audio quality analysis. For example, if the user is insensitive to a specific frequency, the frequency component can be appropriately enhanced in audio processing to better meet the user's hearing needs and enhance the personalization and comfort of the audio experience.

[0041] S102: Extract features of target audio to obtain multiple audio feature data, where the target audio is audio in a target scene.

[0042] In this embodiment, the multiple audio feature data may include: the amplitude data of the audio signal (including the amplitude of the signal part and the noise part) and the background noise amplitude when there is no signal, which is used to calculate the signal-to-noise ratio and the audio dynamic range. The audio signal is Fourier transformed to obtain spectrum data, and the amplitude of the fundamental frequency and each harmonic frequency is determined to calculate the total harmonic distortion. The time domain waveform data of the audio signal is collected, and the high-frequency components and transient response of the audio signal are analyzed to calculate the clarity. The audio signal is Fourier transformed to obtain spectrum data, and the harmonic structure, resonance peak and other features are analyzed. The timbre characteristics are described by comparing the relative intensity and distribution of different frequency components. For stereo audio, the signal data of the left and right channels are collected, and the time difference, phase difference and amplitude difference of the signal are analyzed. The azimuth of the audio source can be calculated based on the time difference, and the phase difference and amplitude difference can be used to infer the sense of distance of the audio source and to describe the sense of space.

[0043] For example, by using deep learning techniques such as convolutional neural networks, recurrent neural networks and their variants (such as long short-term memory networks and gated recurrent units) or generative adversarial networks, large-scale audio data is trained to improve the accuracy and universality of the model. The transfer learning method is used to fine-tune the model parameters for different scenarios to adapt to different real-time communication scenarios.

[0044] In this embodiment, the abnormal patterns and noise in the audio data can be analyzed through anomaly detection and repair technology to detect and locate audio quality problems. The cause of the audio quality problem can also be inferred by fusing audio with multimodal information obtained by other sensors (such as video and location), thereby improving the accuracy and comprehensiveness of audio quality analysis.

[0045] Exemplary, audio and video fusion: using visual information in the video to assist audio feature extraction, such as combining the character's lip shape, movements, etc. with the audio for joint analysis to improve the accuracy and robustness of audio processing.

[0046] Audio and sensor data fusion: Combine sensor data such as accelerometers and gyroscopes to infer the environment and scene where the audio is produced, so as to better extract and process audio features.

[0047] S103: Determine a target repair mode corresponding to a target scene based on a standard audio quality index and a plurality of audio feature data, and repair the audio to be transmitted based on the target repair mode.

[0048] In this embodiment, the target repair mode corresponding to the target scene is determined based on the standard audio quality index and multiple audio feature data, including: The quality level of the target audio is calculated based on a standard audio quality index and a plurality of audio feature data.

[0049] Target repair mode based on quality level.

[0050] For example, in a voice call scenario, the standard audio quality index is first determined, such as requiring a high signal-to-noise ratio to reduce background noise interference, clear voice frequency response, low latency, etc. Feature extraction is performed on the target audio to obtain audio feature data such as the signal-to-noise ratio value and frequency distribution of the current audio.

[0051] The quality level is calculated based on the standard audio quality index and these audio feature data. If the current audio signal-to-noise ratio is lower than the audio signal-to-noise ratio threshold (i.e., the signal-to-noise ratio is low), there is obvious noise, and the frequency response is unbalanced, it can be judged as a lower quality level (such as secondary audio).

[0052] Determine the target repair mode based on the quality level. For low-quality audio, if the signal-to-noise ratio is low, a noise suppression algorithm can be used, such as identifying and removing background noise through a deep learning model. If the frequency response is uneven, the audio equalizer can be adjusted to enhance the signal in the voice frequency range. Before transmission, the audio is processed according to the target repair mode to ensure that the other party receives clear, high-quality voice, improving the quality of voice calls and user experience.

[0053] It can be concluded from the above that this embodiment determines the standard audio quality index by collecting the environmental information and user information of the target scene, and fully considers the auditory differences and preferences of different users. This enables each user to obtain an audio experience that meets their own needs, greatly improving user satisfaction and personalization.

[0054] Taking into account environmental information, such as ambient noise and network conditions, can ensure that audio quality remains high in a variety of complex scenarios. Whether in noisy streets or areas with unstable networks, it can provide users with high-quality audio that is as clear and understandable as possible. In some scenarios with low audio quality requirements, it can also avoid unnecessary high-performance processing, saving computing costs and resources, and adapting to different actual scenario requirements.

[0055] Advanced technologies such as deep learning are used to extract and repair audio features, which improves the accuracy and universality of the model. At the same time, anomaly detection and repair technology and multimodal information fusion can more accurately locate audio quality problems and infer the causes of problems, thereby making more effective repairs. For example, audio and video fusion, audio and sensor data fusion, improve the accuracy and robustness of audio processing.

[0056] In the voice call scenario, this embodiment can significantly improve the call quality. By determining reasonable standard audio quality indicators and target repair modes, problems such as low signal-to-noise ratio and uneven frequency response are solved, ensuring that the other party receives clear, high-quality voice. This not only improves communication efficiency, but also enhances the user's immersion and comfort in voice calls.

[0057] In one embodiment of the present disclosure, the quality level includes primary audio and secondary audio.

[0058] The quality level of the target audio is calculated based on standard audio quality indicators and multiple audio feature data, including: Calculate the similarity between standard audio quality indicators and multiple audio feature data.

[0059] If the similarity is greater than or equal to the quality analysis index, the quality level of the target audio is determined to be level 1 audio.

[0060] If the similarity is less than the quality analysis index, the quality level of the target audio is determined to be secondary audio.

[0061] In this embodiment, the quality analysis index represents the key value for determining the audio quality level division, which can be understood as a threshold for measuring the closeness of the target audio to the standard audio quality index. Different weights can be assigned to the standard audio quality index and each parameter in the audio feature data, and the sum of the weighted difference values ​​can be calculated. Through the analysis and experiment of a large amount of sample data, a difference value that can better distinguish between the first-level audio and the second-level audio is determined as the quality analysis index. This index can also be dynamically adjusted based on expert experience or machine learning algorithms.

[0062] In this embodiment, in a voice call scenario, a standard audio quality index is first calculated based on the environmental information of the target scene (such as environmental noise information, environmental network information) and user information (such as user auditory information, user preference information). Feature extraction is performed on the target audio to obtain multiple audio feature data, which reflect the characteristics of the target audio in different aspects. The similarity between the standard audio quality index and the multiple audio feature data is calculated, and the similarity can be determined by comparing the degree of difference in the values ​​of each specific indicator, such as comparing the proximity of indicators such as signal-to-noise ratio, frequency response, and audio dynamic range. If the similarity is greater than or equal to the quality analysis index, the quality level of the target audio is determined as primary audio. If the similarity is less than the quality analysis index, it is determined as secondary audio.

[0063] In this embodiment, setting different quality levels helps to classify and manage audio and process it in a targeted manner. The quality analysis index is set as a judgment threshold to distinguish the degree of closeness of the audio to the standard. If the audio has a high degree of similarity to the standard audio quality index, it means that the audio is closer to the ideal state in all aspects and can meet higher quality requirements, so it is determined to be first-level audio. When the similarity is low, it means that the audio has a large gap with the standard in some aspects and needs further optimization and repair, so it is determined to be second-level audio. Such a division can more specifically adopt different processing strategies for audio of different quality levels.

[0064] For example, in a voice call, assume that the quality analysis index is set to 85% similarity. User A is making a call in a relatively quiet indoor environment with good network conditions. First, standard audio quality indicators are determined based on the environment and user information, such as a high signal-to-noise ratio, clear frequency response, etc. After feature extraction of the target audio of the current call, it is found that its various audio features are relatively close to the standard indicators. After calculation, the similarity is 90%, which is greater than the quality analysis index. At this time, the quality level of the target audio is determined to be level 1 audio.

[0065] User B is talking in a noisy outdoor environment, and the network has a certain delay. After extracting the target audio features, it is found that the audio is stuck and distorted due to the high environmental noise, low signal-to-noise ratio, and network delay. The calculated similarity is 70%, which is less than the quality analysis index, and the target audio is determined to be secondary audio. The system can take some measures, such as enhancing noise suppression and adjusting the bit rate to adapt to network conditions, to improve audio quality.

[0066] This embodiment can quickly evaluate audio quality through quality level division and provide a clear processing direction for audio repair. For level 1 audio, unnecessary processing can be reduced and call efficiency can be improved. For level 2 audio, effective repair measures can be taken in a timely manner to improve call quality. At the same time, through quality level division, users can better understand the status of call audio, and service providers can also allocate and optimize resources according to different levels to improve the overall service quality. The adjustability of the quality analysis index enables this method to adapt to different call scenarios and needs, and has high flexibility and practicality.

[0067] In one embodiment of the present disclosure, the target repair mode includes a first repair mode or a second repair mode.

[0068] Target repair modes based on quality level, including: In response to the quality level of the target audio being level 1 audio, the target repair mode is determined to be the first repair mode.

[0069] In response to the quality level of the target audio being the secondary audio, the target repair mode is determined to be the second repair mode.

[0070] In this embodiment, audios of different quality levels usually require different degrees and types of repair measures. Level 1 audio is relatively closer to standard audio quality and only requires some fine-tuning or light optimization processing, so the first repair mode is adopted. Level 2 audio is far from the standard and requires more powerful repair measures, so the second repair mode is adopted. Such logic can ensure targeted processing of audios of different qualities, improve repair efficiency and effect, and meet the audio quality requirements of users in voice calls.

[0071] Exemplarily, assume that the quality analysis index is set to a certain similarity value. User A is talking in a quiet environment with a stable network. After analyzing the target audio, it is determined that its quality level is level 1 audio. At this time, the first repair mode may include making slight equalization adjustments to the audio to further optimize the frequency response of the audio, or enhancing the signal strength by a smaller amplitude to improve clarity, so that the quality of the target audio reaches a higher value similar to the standard audio index, such as the first similarity.

[0072] User B is in a noisy environment, and the network also has certain delays and fluctuations. After analysis, the target audio is determined to be secondary audio. The second repair mode may include a powerful noise suppression algorithm to reduce the interference of environmental noise. Adjust the bit rate and compression ratio to adapt to unstable network conditions, and perform more significant signal enhancement processing on the audio to improve the intelligibility and quality of the audio. The quality of the target audio is similar to the standard audio index to the second similarity. The second similarity is less than the first similarity.

[0073] In this embodiment, the method further includes: determining a target repair mode based on a target trigger instruction. The target trigger instruction includes a first trigger instruction or a second trigger instruction.

[0074] In response to the target trigger instruction being the first trigger instruction, the target repair mode is determined to be the first repair mode.

[0075] In response to the target trigger instruction being the second trigger instruction, the target repair mode is determined to be the second repair mode.

[0076] Exemplarily, the user can independently select the corresponding target trigger instruction based on the needs. For example, if the user has low requirements for audio quality and pays attention to efficiency, the second repair mode can be adopted by sending the second trigger instruction. If the user has high requirements for audio quality, the first repair mode can be adopted by sending the first trigger instruction. The repair process of the first repair mode is more refined, the adjustment step is smaller, and the number of adjustments is more. The repair process of the second repair mode is relatively broad, for example, a larger adjustment step is used to correct certain audio parameters, and the number of adjustments is less.

[0077] This embodiment automatically determines the repair mode according to the audio quality level, performs targeted processing on audios of different qualities, improves the repair efficiency and effect, and meets the user's audio quality requirements. In addition, the user can independently select the trigger instruction to determine the repair mode, taking into account the personalized needs of different users for audio quality and efficiency. Whether the user focuses on quality or efficiency, they can get the appropriate treatment, which improves the user experience, and also makes audio repair more flexible and intelligent, adapting to various usage scenarios.

[0078] In one embodiment of the present disclosure, repairing the audio to be transmitted based on the target repair mode includes: In response to the target repair mode being the first repair mode: The audio processing parameters are adjusted based on the first step length.

[0079] The audio to be transmitted is repaired based on the adjusted audio processing parameters and the target repair frequency.

[0080] In response to the target repair mode being the second repair mode: The audio processing parameters are adjusted based on a second step length, wherein the second step length is greater than the first step length.

[0081] The audio to be transmitted is repaired based on the adjusted audio processing parameters and the target repair frequency.

[0082] In this embodiment, the first step length and the second step length represent the adjustment amplitude in the process of adjusting the audio processing parameters. The audio processing parameters may include at least one of the audio equalization parameters, gain, noise suppression strength, and compression ratio, and the audio quality can be repaired and improved by adjusting these parameters.

[0083] In this embodiment, it also includes: calculating a quality stability index of the target audio based on multiple audio feature data.

[0084] Determine the target repair frequency based on the quality stability index.

[0085] The quality stability index can reflect the stability of the target audio quality. For example, the quality stability index can be calculated by comprehensively considering factors such as audio fluctuations and signal strength changes.

[0086] The target repair frequency indicates the frequency of performing segment-by-segment repair on the audio to be transmitted.

[0087] In this embodiment, determining the target repair frequency based on the quality stability index includes: If the quality stability index is less than the first threshold, the target repair frequency is determined to be the first repair frequency.

[0088] If the quality stability index is greater than or equal to the first threshold, the target repair frequency is determined to be the second repair frequency.

[0089] The first repair frequency is greater than the second repair frequency.

[0090] In this embodiment, repairing the audio to be transmitted based on the target repair frequency includes: In response to the target repair frequency being the first repair frequency, the audio to be transmitted is divided into a plurality of first audio segments based on the first repair frequency, and the first repair frequency represents the lengths of the plurality of first audio segments.

[0091] Each first audio segment is repaired using different audio processing parameters.

[0092] In response to the target repair frequency being the second repair frequency, the audio to be transmitted is divided into a plurality of second audio segments based on the second repair frequency, and the second repair frequency represents the lengths of the plurality of second audio segments.

[0093] Different audio processing parameters are used to repair each second audio segment.

[0094] The first repair frequency is higher and is suitable for situations with a lower quality stability index, which means that the audio is divided into more segments and more targeted repairs are performed; the second repair frequency is lower and is suitable for situations with a higher quality stability index. For example, if the target audio is relatively stable, you can intercept a segment of audio every 10 seconds for analysis and adjust and update the audio processing parameters, or divide the audio to be transmitted according to the data volume, and the data volume of each audio segment is the same. Analyze each audio segment and adjust and update the audio processing parameters, and then repair the subsequent audio to be transmitted based on the adjusted audio processing parameters.

[0095] In this embodiment, in response to the target repair mode being the first repair mode, the method further includes: The audio to be transmitted is repaired based on the adjusted audio processing parameter and the first repair frequency until the first condition is met.

[0096] The first condition is that the similarity between the restored audio feature data and the standard audio quality index is greater than a first similarity.

[0097] In this embodiment, in response to the target repair mode being the second repair mode, the method further includes: The audio to be transmitted is repaired based on the adjusted audio processing parameter and the second repair frequency until the second condition is met.

[0098] The second condition is that the similarity between the restored audio feature data and the standard audio quality index is greater than the second similarity. The first similarity is greater than the second similarity.

[0099] The first similarity and the second similarity respectively represent the similarity requirements between the restored audio feature data and the standard audio quality index in the first restoration mode and the second restoration mode. A higher first similarity indicates a stricter audio quality requirement in the first restoration mode.

[0100] Exemplarily, different audio conditions require different processing strategies. For audio with better quality (primary audio), the first repair mode is adopted. Adjusting the audio processing parameters in small steps can avoid over-processing and maintain the naturalness of the audio. At the same time, the higher first repair frequency and the higher first similarity requirement can ensure that the audio quality reaches a higher level, which is suitable for scenes with higher requirements for audio quality. For audio with poorer quality (secondary audio), the use of large step size adjustment can quickly improve the audio quality. The lower second repair frequency and the lower second similarity requirement take into account that the high quality standard may not be fully met under harsh conditions, but the audio intelligibility can still be improved to a certain extent. The quality stability index determines the repair frequency. If the audio quality is stable, a higher frequency can be used for fine repair; if it is unstable, the repair frequency is reduced to avoid waste of resources and possible over-adjustment.

[0101] Exemplarily, assume that in a voice call, user A is talking in a quiet environment with a stable network. The target audio is determined to be level one audio, and the first step may be to fine-tune the audio equalization parameters by 0.5 dB. Audio processing parameters include moderate high-frequency gain and lower noise suppression intensity. The quality stability index is calculated by analyzing the fluctuations of the audio signal. If the stability index is greater than or equal to the first threshold, the first repair frequency is determined to be once every 2 minutes. The audio is divided into multiple short audio segments according to the first repair frequency, and each audio segment is fine-tuned until the similarity between the repaired audio and the standard audio quality index is greater than the first similarity, such as 90%.

[0102] User B is talking in a noisy environment with unstable network. The target audio is determined to be secondary audio, and the second step can be to increase the noise suppression strength by 3 levels. The audio processing parameters include higher noise suppression and lower bit rate adjustment. The quality stability index is less than the first threshold, and the second repair frequency is determined to be once every 5 minutes. The audio is divided into longer audio segments according to the second repair frequency, and each audio segment is adjusted significantly until the similarity between the repaired audio and the standard audio quality index is greater than the second similarity, such as 80%.

[0103] This embodiment improves the pertinence and adaptability of audio repair. Different repair modes and parameters are selected according to the actual quality and stability of the audio to meet the needs of users in different scenarios. By reasonably setting the repair frequency and similarity requirements, it not only ensures the fine processing of high-quality audio, but also quickly improves the intelligibility of poor audio when resources are limited. At the same time, this dynamic adjustment method can improve the utilization efficiency of system resources, avoid unnecessary calculations and processing, and improve the overall quality of voice calls and user experience.

[0104] Corresponding to the universal audio quality analysis method of the above embodiment, Figure 2 This is a structural block diagram of a universal audio quality analysis system provided by an embodiment of the present disclosure. For ease of explanation, only the parts related to the embodiment of the present disclosure are shown. Figure 2 , the universal audio quality analysis system 20 comprises: The first quality analysis module 21 is used to determine a standard audio quality index corresponding to a target scene.

[0105] The second quality analysis module 22 is used to extract features from the target audio to obtain a plurality of audio feature data, where the target audio is the audio in the target scene.

[0106] The audio repair module 23 is used to determine a target repair mode corresponding to a target scene based on a standard audio quality index and a plurality of audio feature data, and repair the audio to be transmitted based on the target repair mode.

[0107] In one embodiment of the present disclosure, the first quality analysis module 21 is specifically configured to update the initial audio quality indicator based on the environmental information of the target scene, determine the first audio quality indicator, The initial audio quality indicator is updated based on the user information of the target scenario to determine a second audio quality indicator.

[0108] The first audio quality indicator and the second audio quality indicator are integrated to determine a standard audio quality indicator.

[0109] In an embodiment of the present disclosure, the audio repair module 23 is specifically configured to calculate the quality level of the target audio based on a standard audio quality index and a plurality of audio feature data.

[0110] Target repair mode based on quality level.

[0111] In one embodiment of the present disclosure, the quality level includes primary audio and secondary audio.

[0112] The audio repair module 23 is specifically used to calculate the similarity between the standard audio quality index and multiple audio feature data.

[0113] If the similarity is less than the quality analysis index, the quality level of the target audio is determined to be level one audio.

[0114] If the similarity is greater than or equal to the quality analysis index, the quality level of the target audio is determined to be secondary audio.

[0115] In one embodiment of the present disclosure, the target repair mode includes a first repair mode or a second repair mode.

[0116] The audio repair module 23 is further configured to determine, in response to the quality level of the target audio being level one audio, that the target repair mode is the first repair mode.

[0117] In response to the quality level of the target audio being the secondary audio, the target repair mode is determined to be the second repair mode.

[0118] In one embodiment of the present disclosure, the audio repair module 23 is further configured to: in response to the target repair mode being the first repair mode: The audio processing parameters are adjusted based on the first step length.

[0119] The audio to be transmitted is repaired based on the adjusted audio processing parameter and the first repair frequency.

[0120] In one embodiment of the present disclosure, the audio repair module 23 is further configured to: in response to the target repair mode being the second repair mode: The audio processing parameters are adjusted based on a second step length, wherein the second step length is greater than the first step length.

[0121] The audio to be transmitted is repaired based on the adjusted audio processing parameter and the second repair frequency.

[0122] See also Figure 3 , Figure 3 A schematic block diagram of an electronic device provided by an embodiment of the present disclosure. Figure 3 The electronic device 300 in the embodiment shown may include: one or more processors 301, one or more input devices 302, one or more output devices 303 and one or more memories 304. The processors 301, input devices 302, output devices 303 and memories 304 communicate with each other via a communication bus 305. The memory 304 is used to store computer programs, which include program instructions. The processor 301 is used to execute the program instructions stored in the memory 304. The processor 301 is configured to call the program instructions to execute the functions of the modules in the above-mentioned system embodiments, such as Figure 2 The functions of modules 21 to 23 are shown.

[0123] It should be understood that in the embodiment of the present disclosure, the processor 301 may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0124] The input device 302 may include a touch panel, a fingerprint collection sensor (for collecting the user's fingerprint information and fingerprint direction information), a microphone, etc., and the output device 303 may include a display (LCD, etc.), a speaker, etc.

[0125] The memory 304 may include a read-only memory and a random access memory, and provide instructions and data to the processor 301. A portion of the memory 304 may also include a non-volatile random access memory. For example, the memory 304 may also store information about the device type.

[0126] In a specific implementation, the processor 301, input device 302, and output device 303 described in the embodiments of the present disclosure can execute the implementation methods described in the first and second embodiments of the universal audio quality analysis method provided in the embodiments of the present disclosure, and can also execute the implementation methods of the electronic device 300 described in the embodiments of the present disclosure, which will not be repeated here.

[0127] In another embodiment of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by the processor, all or part of the processes in the above-mentioned embodiment method are implemented, and the computer program can also be completed by instructing the relevant hardware through the computer program. The computer program can be stored in a computer-readable storage medium, and when the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0128] The computer-readable storage medium may be an internal storage unit of the electronic device of any of the aforementioned embodiments, such as a hard disk or memory of the electronic device. The computer-readable storage medium may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device. Furthermore, the computer-readable storage medium may also include both an internal storage unit of the electronic device and an external storage device. The computer-readable storage medium is used to store computer programs and other programs and data required by the electronic device. The computer-readable storage medium may also be used to temporarily store data that has been output or is to be output.

[0129] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this disclosure.

[0130] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the electronic devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0131] In the several embodiments provided in the present application, it should be understood that the disclosed electronic device and method can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces or units, or it can be an electrical, mechanical or other form of connection.

[0132] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present disclosure.

[0133] In addition, each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0134] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in the present disclosure, and these modifications or replacements should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A universal audio quality analysis method, characterized in that: include: Determine the standard audio quality indicators corresponding to the target scene; Extracting features of target audio to obtain multiple audio feature data, wherein the target audio is audio in the target scene; A target repair mode corresponding to the target scene is determined based on the standard audio quality index and the multiple audio feature data, and the audio to be transmitted is repaired based on the target repair mode.

2. The universal audio quality analysis method according to claim 1, characterized in that: The step of determining a standard audio quality indicator corresponding to the target scene includes: Updating the initial audio quality indicator based on the environmental information of the target scene to determine a first audio quality indicator; Updating the initial audio quality indicator based on the user information of the target scene to determine a second audio quality indicator; The first audio quality indicator and the second audio quality indicator are merged to determine a standard audio quality indicator.

3. The universal audio quality analysis method according to claim 1, characterized in that: The determining the target repair mode corresponding to the target scene based on the standard audio quality indicator and the plurality of audio feature data comprises: Calculate the quality level of the target audio based on the standard audio quality index and the plurality of audio feature data; The target repair mode is determined based on the quality level.

4. The universal audio quality analysis method according to claim 3, characterized in that: The quality levels include primary audio and secondary audio; Calculating the quality level of the target audio based on the standard audio quality indicator and the plurality of audio feature data includes: Calculating the similarity between the standard audio quality index and the plurality of audio feature data; If the similarity is less than the quality analysis index, determining the quality level of the target audio to be first-level audio; If the similarity is greater than or equal to the quality analysis index, it is determined that the quality level of the target audio is secondary audio.

5. The universal audio quality analysis method according to claim 4, characterized in that: The target repair mode includes a first repair mode or a second repair mode; The determining the target repair mode based on the quality level comprises: In response to the quality level of the target audio being level 1 audio, determining the target repair mode to be a first repair mode; In response to the quality level of the target audio being level 2 audio, the target repair mode is determined to be a second repair mode.

6. The universal audio quality analysis method according to claim 5, characterized in that: The repairing of the audio to be transmitted based on the target repair mode includes: In response to the target repair mode being the first repair mode: Adjust the audio processing parameters based on the first step length; The audio to be transmitted is repaired based on the adjusted audio processing parameter and the first repair frequency.

7. The universal audio quality analysis method according to claim 6, characterized in that: The repairing of the audio to be transmitted based on the target repair mode includes: In response to the target repair mode being the second repair mode: adjusting the audio processing parameter based on a second step length, wherein the second step length is greater than the first step length; The audio to be transmitted is repaired based on the adjusted audio processing parameter and the second repair frequency.

8. A universal audio quality analysis system, characterized in that: include: A first quality analysis module, used to determine a standard audio quality index corresponding to a target scene; A second quality analysis module is used to extract features of the target audio to obtain a plurality of audio feature data, wherein the target audio is the audio in the target scene; An audio repair module is used to determine a target repair mode corresponding to the target scene based on the standard audio quality index and the multiple audio feature data, and repair the audio to be transmitted based on the target repair mode.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images