Vehicle sound processing method and device, storage medium and program product

Abstract

The invention relates to a vehicle sound processing method and device, a storage medium and a program product. The vehicle sound processing device obtains a processed vehicle sound material; performing feature extraction processing on the processed vehicle sound material by using the trained evaluation model to obtain feature data, and performing analysis processing on sound parameters according to the feature data to obtain a parameter analysis result; wherein the parameter analysis result comprises at least one of a tone analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result and a sound source length analysis result; determining an evaluation question according to the sound type and the material quantity of the vehicle sound material and the quantity of the evaluation objects; determining an evaluation result of the processed vehicle sound material based on the evaluation question; determining a sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result; therefore, objectivity and accuracy of vehicle sound evaluation can be effectively improved.

Description

Technical Field

[0001] This application relates to the technical field of vehicles, and in particular, to a vehicle sound processing method, apparatus, storage medium, and program product. Background Art

[0002] Vehicles play various sound elements to improve the experience of vehicle drivers and passengers, such as beep sounds, etc.; through comprehensive evaluation of the sounds in the vehicle, directional improvement suggestions can be provided for the vehicle sound design; however, in the current related processing methods for vehicle sounds, they generally rely on user feedback, resulting in poor performance in terms of objectivity and accuracy of the evaluation results. Summary of the Invention

[0003] Embodiments of this application provide a vehicle sound processing method, apparatus, storage medium, and program product, which can effectively improve the objectivity and accuracy of vehicle sound evaluation.

[0004] In a first aspect, embodiments of this application provide a vehicle sound processing method, the method comprising:

[0005] Obtain processed vehicle sound materials;

[0006] Use a trained evaluation model to perform feature extraction processing on the processed vehicle sound materials to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result;

[0007] Performing analysis processing on sound parameters based on the feature data to obtain a parameter analysis result includes at least one of the following:

[0008] Perform fundamental frequency analysis based on the spectral feature information in the feature data to obtain a pitch analysis result;

[0009] Perform pitch analysis based on the timing feature information in the feature data to obtain a pitch analysis result;

[0010] Perform rhythm analysis based on the sound paragraph feature information in the feature data to obtain a rhythm analysis result;

[0011] Perform loudness analysis based on the time-frequency feature information in the feature data to obtain a loudness analysis result;

[0012] Perform timbre analysis based on the sound source feature information in the feature data to obtain a timbre analysis result;

[0013] Perform sound source length analysis based on the sound intensity change timing feature information in the feature data to obtain a sound source length analysis result;

[0014] Determine the evaluation questions according to the sound type and the number of sound materials of the processed vehicle sound materials, as well as the number of evaluation objects;

[0015] Determine the evaluation results of the processed vehicle sound materials based on the evaluation questions;

[0016] Determine the sound evaluation results of the processed vehicle sound materials according to the parameter analysis results and the evaluation results.

[0017] In this embodiment, after obtaining the vehicle sound materials, the vehicle sound processing device can use the trained evaluation model to extract features from the vehicle sound materials. The extracted feature data can include different types of features. Furthermore, different types of sound parameters can be analyzed for different types of features respectively to obtain objective parameter analysis results. The parameter analysis results can include at least one of pitch analysis results, tone analysis results, rhythm analysis results, loudness analysis results, timbre analysis results, and sound source length analysis results, improving the comprehensiveness of parameter analysis and thus the objectivity of parameter analysis results. As a result, objective parameter analysis results in different dimensions can be obtained. At the same time, the evaluation questions matching the vehicle sound materials can be determined. The vehicle sound processing device determines the evaluation questions according to the sound type and the number of sound materials of the vehicle sound materials, as well as the number of evaluation objects, thereby improving the matching degree between the evaluation questions and the vehicle sound materials. And based on these evaluation questions, the evaluation results of the vehicle sound materials can be obtained. Finally, the sound evaluation results of the vehicle sound materials can be obtained by referring to both the evaluation results and the parameter analysis results at the same time, effectively improving the objectivity and accuracy of vehicle sound evaluation.

[0018] Further, in some embodiments, the sound type includes reminder sound type, engine sound wave type, and music type; the evaluation object represents the object answering the evaluation questions.

[0019] In this embodiment, the matching evaluation questions can be determined based on different types of these sound materials such as reminder sound type, engine sound wave type, and music type, the number of evaluation objects, and the number of the above different types of sound materials, improving the reliability of evaluating the vehicle sound materials.

[0020] Further, in some embodiments, obtaining the processed vehicle sound materials includes:

[0021] Obtain the vehicle sound materials;

[0022] Perform noise elimination processing and audio unification processing on the vehicle sound materials to obtain the processed vehicle sound materials.

[0023] In this embodiment, when the vehicle sound processing device analyzes the sound parameters of the vehicle sound material, it can first perform noise cancellation processing and audio unification processing on the vehicle sound material to process the sound material into a format that can support subsequent processing, thereby improving the processing efficiency of subsequent operations such as feature extraction on the processed vehicle sound material.

[0024] Further, in some embodiments, before using the trained evaluation model to perform feature extraction processing on the processed vehicle sound material to obtain feature data and performing analysis processing on the sound parameters based on the feature data to obtain a parameter analysis result, the method further includes:

[0025] Obtain vehicle audio data;

[0026] Determine audio positive samples and audio negative samples according to the vehicle audio data;

[0027] Use the audio positive samples and audio negative samples to train the initial evaluation model to obtain the trained evaluation model.

[0028] In this embodiment, the vehicle sound processing device can also obtain some vehicle audio data, determine the positive and negative samples therein, and thus use the positive and negative samples to train the initial evaluation model to obtain a trained evaluation model that can objectively analyze the sound parameters of the vehicle sound material, thereby ensuring the objectivity of vehicle sound evaluation.

[0029] Further, in some embodiments, determining audio positive samples and audio negative samples according to the vehicle audio data includes:

[0030] Classify the vehicle audio data according to the functional attributes of the vehicle audio data to obtain the classified audio data;

[0031] Perform quality verification on the classified audio data to obtain a quality verification result;

[0032] Determine audio positive samples and audio negative samples based on the quality verification result.

[0033] In this embodiment, when the vehicle sound processing device determines the positive and negative samples, it can first classify the vehicle audio data into various classified audio data according to the functional attributes of the vehicle audio data, and then can determine the positive and negative samples according to the quality verification results of these classified audio data to realize the construction of the positive and negative samples, thereby improving the training effect of the evaluation model.

[0034] Further, in some embodiments, determining audio positive samples and audio negative samples based on the quality verification result includes:

[0035] Label the audio data with a passed quality verification result as audio positive samples;

[0036] Label the audio data that fails the quality verification as audio negative samples.

[0037] In this embodiment, the audio data that fails the quality verification can be labeled as audio negative samples, while the audio data that passes the quality verification can be labeled as audio positive samples, so as to distinguish between positive and negative samples.

[0038] Furthermore, in some embodiments, the sound evaluation result of the processed vehicle sound material is determined according to the parameter analysis result and the evaluation result, including:

[0039] Perform weighted processing on the first weight information corresponding to the parameter analysis result, the second weight information corresponding to the evaluation result, the parameter analysis result, and the evaluation result to obtain the sound evaluation result.

[0040] In this embodiment, the vehicle sound processing device can perform weighted processing on the first weight information corresponding to the parameter analysis result, the second weight information corresponding to the evaluation result, the parameter analysis result, and the evaluation result to obtain the final sound evaluation result, improving the objectivity and accuracy of the sound evaluation.

[0041] In a second aspect, an embodiment of the present application provides a vehicle sound processing device, including an acquisition unit, an analysis unit, and a determination unit;

[0042] The acquisition unit is configured to acquire the processed vehicle sound material;

[0043] The analysis unit is configured to perform feature extraction processing on the processed vehicle sound material using the trained evaluation model to obtain feature data, and perform analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; and is configured to perform at least one of the following: perform fundamental frequency analysis on the spectral feature information in the feature data to obtain a pitch analysis result; perform pitch analysis on the timing feature information in the feature data to obtain a pitch analysis result; perform rhythm analysis on the sound paragraph feature information in the feature data to obtain a rhythm analysis result; perform loudness analysis on the time-frequency feature information in the feature data to obtain a loudness analysis result; perform timbre analysis on the sound source feature information in the feature data to obtain a timbre analysis result; perform sound source length analysis on the sound intensity change timing feature information in the feature data to obtain a sound source length analysis result;

[0044] A determination unit is configured to determine an evaluation question based on the sound type and the number of vehicle sound materials after processing, as well as the number of evaluation objects; and determine the evaluation result of the vehicle sound materials after processing based on the evaluation question, and determine the sound evaluation result of the vehicle sound materials after processing according to the parameter analysis result and the evaluation result.

[0045] In this embodiment, after obtaining the vehicle sound materials, the vehicle sound processing device can use the trained evaluation model to extract features from the vehicle sound materials. The extracted feature data can include different types of features. Then, different types of sound parameters can be analyzed for different types of features respectively to obtain an objective parameter analysis result. The parameter analysis result can include at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result, which improves the comprehensiveness of the parameter analysis and thus the objectivity of the parameter analysis result. As a result, objective parameter analysis results in different dimensions can be obtained. At the same time, an evaluation question matching the vehicle sound materials can be determined. The vehicle sound processing device determines the evaluation question according to the sound type and the number of vehicle sound materials, as well as the number of evaluation objects, which can improve the matching degree between the evaluation question and the vehicle sound materials. And based on these evaluation questions, the evaluation result of the vehicle sound materials can be obtained. Finally, the sound evaluation result of the vehicle sound materials can be obtained by referring to both the evaluation result and the parameter analysis result at the same time, which can effectively improve the objectivity and accuracy of the vehicle sound evaluation.

[0046] In a third aspect, an embodiment of the present application provides a vehicle sound processing device, which includes a processor and a memory storing processor-executable instructions; when the instructions are executed by the processor, the above vehicle sound processing method is implemented.

[0047] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above vehicle sound processing method is implemented.

[0048] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program or instructions. When the computer program or instructions are executed by a processor, the steps in the above vehicle sound processing method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Schematic diagram of the implementation process of the vehicle sound processing method provided by the embodiment of the present application Figure 1 ;

[0050] Figure 2 Schematic diagram of the implementation process of the vehicle sound processing method provided by the embodiment of the present application Figure 2 ;

[0051] Figure 3 Schematic diagram of the composition structure of the vehicle sound processing device provided by the embodiment of the present application Figure 1 ;

[0052] Figure 4 Schematic diagram of the composition structure of the vehicle sound processing device provided by the embodiment of the present application Figure 2 。 Detailed implementation manners

[0053] The following will describe the implementation manners of the present application with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in this specification. The present application can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, rather than for limiting the protection scope of the present application.

[0054] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0055] Currently, in order to improve the experience of drivers and passengers in a vehicle, various vehicle sound elements are usually set in the vehicle, such as car reminder sounds, etc.; by comprehensively evaluating the sounds in the vehicle, directional improvement suggestions can be provided for the sound design of the vehicle; however, in the related methods for evaluating vehicle sounds, the evaluation of vehicle sounds is mainly based on user feedback, which will be affected by individual preferences and emotions, limiting the objectivity of the evaluation results. Moreover, since vehicle reminder sounds involve multiple dimensions, such as the pitch, loudness, rhythm, etc. of the sound, the current evaluation methods are difficult to capture all-dimensional information at the same time, resulting in poor reliability of the evaluation results; at the same time, since users are not familiar with professional terms, it is also difficult to accurately express their feelings when evaluating sound dimensions. Therefore, relying on user feedback will also lead to errors in the evaluation results.

[0056] To solve the above problems, in the embodiments of the present application, a vehicle sound processing method, device, storage medium, and program product are proposed. The vehicle sound processing device can obtain the processed vehicle sound materials, perform feature extraction processing on the processed vehicle sound materials by using the trained evaluation model to obtain feature data, and perform analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result. Among them, the parameter analysis result includes at least one of a pitch analysis result, a tone analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result. Determine the evaluation questions according to the sound type and the number of the vehicle sound materials, as well as the number of evaluation objects. Determine the evaluation result of the processed vehicle sound materials based on the evaluation questions. Determine the sound evaluation result of the vehicle sound materials according to the parameter analysis result and the evaluation result. Based on the above solution, the objectivity and accuracy of vehicle sound evaluation can be effectively improved.

[0057] Next, the technical solutions in the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application.

[0058] Figure 1 Schematic diagram of the implementation process of the vehicle sound processing method provided in the embodiments of the present application Figure 1 , such as Figure 1 shown, in the embodiments of the present application, the vehicle sound processing method of the vehicle sound processing device may include the following steps:

[0059] Step 101, obtain the processed vehicle sound materials.

[0060] In the embodiments of the present application, the vehicle sound processing device can obtain the processed vehicle sound materials.

[0061] In some embodiments of the present application, when the vehicle sound processing device obtains the processed vehicle sound materials, it can first obtain the vehicle sound materials, and then perform noise elimination processing and audio unification processing on the vehicle sound materials to obtain the processed vehicle sound materials.

[0062] In the embodiments of the present application, the vehicle sound materials represent the sound materials to be evaluated.

[0063] In the embodiments of the present application, the number of the vehicle sound materials is not limited in the present application. For example, the vehicle sound processing device obtains 40 vehicle sound materials, so that subsequent vehicle sound evaluation can be performed on the 40 vehicle sound materials to obtain the sound evaluation results of the 40 vehicle sound materials.

[0064] In the embodiments of the present application, the vehicle sound processing device can be any electronic device with communication and computing functions, such as a computer, a server, etc.

[0065] In an embodiment of the present application, noise cancellation processing can be used to remove ambient noise that interferes with vehicle sound materials.

[0066] In some embodiments of the present application, the vehicle sound processing device can use beamforming technology to directionally enhance the target sound source in the vehicle sound material and suppress ambient noise, such as wind noise and tire noise, thereby completing the noise cancellation processing.

[0067] In an embodiment of the present application, audio unification processing can be used to unify the format of vehicle sound materials.

[0068] In some embodiments of the present application, the vehicle sound processing device can unify the sampling rate, loudness base, etc. of the vehicle sound material.

[0069] It can be understood that in an embodiment of the present application, the processed vehicle sound material represents the sound material that has undergone noise cancellation processing and audio unification processing.

[0070] Step 102: Use the trained evaluation model to perform feature extraction processing on the processed vehicle sound material to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result can include at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result.

[0071] In an embodiment of the present application, after obtaining the processed vehicle sound material, the vehicle sound processing device can use the trained evaluation model to perform feature extraction processing on the vehicle sound material to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result.

[0072] In some embodiments of the present application, the sound parameters can include parameters of multiple dimensions; the sound parameters can include at least one of pitch, pitch, rhythm, loudness, timbre, and sound source length.

[0073] In an embodiment of the present application, when the vehicle sound processing device analyzes and processes sound parameters based on the feature data to obtain a parameter analysis result, it may include at least one of the following: performing fundamental frequency analysis based on the spectral feature information in the feature data to obtain a pitch analysis result; performing pitch analysis based on the temporal feature information in the feature data to obtain a pitch analysis result; performing rhythm analysis based on the sound segment feature information in the feature data to obtain a rhythm analysis result; performing loudness analysis based on the time-frequency feature information in the feature data to obtain a loudness analysis result; performing timbre analysis based on the sound source feature information in the feature data to obtain a timbre analysis result; performing sound source length analysis based on the temporal feature information of sound intensity change in the feature data to obtain a sound source length analysis result.

[0074] In an embodiment of the present application, the trained evaluation model may be obtained by training based on an initial evaluation model. The initial evaluation model may be an evaluation model constructed based on a deep neural network, which can be used to extract high-dimensional abstract features of the input sound material and output key feature data.

[0075] In some embodiments of the present application, the trained evaluation model may include a feature extraction layer and a temporal modeling layer. The feature extraction layer may be a network layer constructed based on a convolutional module, which can be used to extract the time-frequency features of the sound material. The temporal modeling layer may be a network layer constructed based on a bidirectional long short-term memory network (LSTM), which can be used to extract the dynamic change features of the sound material.

[0076] In an embodiment of the present application, the feature data may include features in different dimensions, so as to be able to measure the performance of the processed vehicle sound material in terms of parameters in different dimensions based on the feature data, and improve the reliability and comprehensiveness of the sound parameter analysis.

[0077] In some embodiments of the present application, the feature data may include at least one of spectral feature information, temporal feature information, sound segment feature information, time-frequency feature information, sound source feature information, and temporal feature information of sound intensity change.

[0078] It can be understood that in an embodiment of the present application, the vehicle sound processing device can perform precise calculations of sound parameters in different dimensions through the feature data, and can achieve efficient and high-precision sound parameter analysis.

[0079] In some embodiments of the present application, the vehicle sound processing device may perform fundamental frequency analysis based on the spectral feature information in the feature data to obtain a pitch analysis result.

[0080] Exemplarily, the spectral feature information can be output in the form of a Mel spectrogram, so that the vehicle sound processing device can use the Mel spectrogram for fundamental frequency analysis to obtain a pitch analysis result; among them, the fundamental frequency is the core concept of pitch parameters, referring to the harmonic component with the lowest frequency in the sound wave vibration, and is also the key factor determining the pitch perception.

[0081] In some embodiments of the present application, the vehicle sound processing device can perform pitch analysis based on the timing feature information in the feature data to obtain a pitch analysis result.

[0082] Exemplarily, the timing feature information is presented as a high-precision pitch curve of 100 frames per second. The vehicle sound processing device calculates the dynamic fluctuation frequency and fluctuation amplitude according to the timing feature information, so that a pitch analysis result can be obtained based on the dynamic fluctuation frequency and fluctuation amplitude; among them, the dynamic fluctuation frequency can calculate the change frequency by counting the number of pitch extreme points within a unit time, so as to obtain the calculation result of the dynamic fluctuation frequency; the fluctuation amplitude can be calculated by calculating the ratio of the standard deviation to the mean of the pitch curve in the timing feature information to detect abnormal jumps, so as to obtain the calculation result of the fluctuation amplitude.

[0083] In some embodiments of the present application, the vehicle sound processing device can perform rhythm analysis based on the sound paragraph feature information in the feature data to obtain a rhythm analysis result.

[0084] Exemplarily, the effective sound paragraphs can be marked in the sound paragraph feature information, so that the vehicle sound processing device can count the pause duration and interval frequency between the effective paragraphs in the sound paragraph feature information, so as to obtain a rhythm analysis result based on the counted pause duration and interval frequency.

[0085] In some embodiments of the present application, the vehicle sound processing device can perform loudness analysis based on the time-frequency feature information in the feature data to obtain a loudness analysis result.

[0086] Exemplarily, the time-frequency feature information can be presented in the form of a time-frequency energy distribution map. The vehicle sound processing device can calculate the loudness according to the time-frequency feature information, such as short-term loudness and comprehensive loudness, etc., to determine the loudness analysis result.

[0087] In some embodiments of the present application, the vehicle sound processing device can perform timbre analysis based on the sound source feature information in the feature data to obtain a timbre analysis result.

[0088] Exemplarily, the sound source feature information can be presented in the form of a 128-dimensional timbre embedding vector, so that the vehicle sound processing device can distinguish different sound source characteristics according to the sound source feature information and use the sound source feature information for calculation and analysis of timbre-related information.

[0089] In some embodiments of the present application, the vehicle sound processing device may perform sound source length analysis based on the sound intensity change timing feature information in the feature data to obtain a sound source length analysis result.

[0090] Exemplarily, the sound intensity change timing feature information may be presented in the form of a voice activity curve, and the sound intensity change timing feature information may include timing marks of sound intensity changes. Thus, the vehicle sound processing device may detect a continuous interval in which the sound intensity exceeds a threshold based on the sound intensity change timing feature information to obtain an effective sound source length, which is the sound source length analysis result.

[0091] Step 103: Determine a test question according to the sound type and the number of sound materials of the processed vehicle sound materials, and the number of evaluation objects.

[0092] In an embodiment of the present application, after obtaining the processed vehicle sound materials, before determining the evaluation result of the vehicle sound materials based on a test question matching the vehicle sound materials, the vehicle sound processing device may determine a test question according to the sound type and the number of sound materials of the processed vehicle sound materials, and the number of evaluation objects.

[0093] In some embodiments of the present application, a large number of test questions may be generated based on the principles of safety priority, standard consistency, functional beneficiary, intuitive communication, moderate arousal, sensory comfort, moderate sensory channel, emotional care, and imagery association. Thus, a test question bank may be constituted according to the large number of test questions. Further, when evaluating the sound of any vehicle sound material subsequently, a test question matching the vehicle sound material may be selected from the test question bank.

[0094] In some embodiments of the present application, the principle of safety first means that the sound design must take ensuring driving safety as the core goal, and the playback of sound shall not interfere with the driver's perception of the road, traffic signals or other key driving information; the principle of standard consistency means that the sound needs to maintain a high degree of unity in core parameters such as timbre, loudness, and rhythm to avoid causing cognitive confusion to users; the principle of functional beneficiary means that the sound must be oriented towards the actual needs of users and verify its value through three dimensions: comfort, necessity, and experience; the principle of intuitive communication means that the sound should have a clear scene directivity so that the driver can understand its meaning without thinking; the principle of moderate arousal means that the playback frequency and duration of the prompt sound need to be strictly matched with the actual needs to avoid excessive disturbance; the principle of sensory comfort means that the acoustic parameters of the sound need to conform to the human auditory comfort curve to avoid physical discomfort; the principle of moderate sensory channel means that the spatial propagation path of the sound needs to be accurately matched with the functional scene to enhance the perception accuracy; the principle of emotional care means that the sound needs to match the emotional needs of users in different usage scenarios and convey emotional value; the principle of image association means that the sound can achieve cross-scene semantic expansion and associative mapping. It can be understood that in the embodiments of the present application, the vehicle sound processing device can determine the evaluation questions matching the vehicle sound materials in the evaluation question bank according to the sound type and quantity of the vehicle sound materials, as well as the number of evaluation objects.

[0095] In some embodiments of the present application, the sound types include prompt sound types, engine sound wave types, and music types.

[0096] In some embodiments of the present application, the evaluation object represents the object answering the evaluation questions.

[0097] In the embodiments of the present application, the quantity and content of the evaluation questions can be matched with the sound type, quantity of the vehicle sound materials, and the number of evaluation objects. The evaluation questions can include questions for comprehensively evaluating a certain type of vehicle sound materials. And when the quantity of the materials and the number of evaluation objects are small, the quantity of the obtained evaluation questions can be larger and the content of the questions can be richer to obtain a more comprehensive and accurate evaluation result.

[0098] Exemplarily, for 40 sound materials of the prompt sound type and 40 evaluation objects, the obtained evaluation questions can include 200 questions. One of the questions can be: Is the listening feeling of this sound pleasant? Please select on a scale of 1 to 5 according to the pleasantness level, where 5 represents the highest pleasantness level; for 10 sound materials of the engine sound wave type and 20 evaluation objects, the obtained evaluation questions can include 500 questions. One of the questions can be: Is this sound sharp? Please select on a scale of 1 to 5 according to the sharpness level, where 5 represents the highest sharpness level.

[0099] Step 104: Determine the evaluation result of the processed vehicle sound material based on the evaluation questions.

[0100] In an embodiment of the present application, after determining the evaluation questions according to the sound type and quantity of the vehicle sound material and the quantity of evaluation objects, the vehicle sound processing device may determine the evaluation result of the processed vehicle sound material based on the evaluation questions.

[0101] Exemplarily, the vehicle sound materials to be evaluated include 30 sound materials of the prompt sound type, and there are 20 evaluation objects currently evaluating these 30 sound materials. Then, the vehicle sound processing device may select 60 evaluation questions from the evaluation question bank according to the prompt sound type (the sound type of the vehicle sound material), 30 (quantity of materials), and 20 (quantity of evaluation objects) as the evaluation questions matching these 30 sound materials of the prompt sound type for the evaluation objects to perform evaluation and obtain the evaluation result.

[0102] Step 105: Determine the sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result.

[0103] In an embodiment of the present application, after the vehicle sound processing device performs feature extraction processing on the processed vehicle sound material using the trained evaluation model to obtain feature data, performs analysis processing on the sound parameters according to the feature data to obtain the parameter analysis result, and determines the evaluation result of the processed vehicle sound material based on the evaluation questions, it may determine the sound evaluation result of the vehicle sound material according to the parameter analysis result and the evaluation result.

[0104] In some embodiments of the present application, the parameter analysis result may be understood as the objective evaluation content obtained by performing parameter analysis on the vehicle sound material in multiple dimensions, while the evaluation result may be understood as the subjective evaluation content manually performed on the vehicle sound material. Therefore, determining the sound evaluation result based on the parameter analysis result and the evaluation result can comprehensively consider the subjective and objective evaluation contents and improve the accuracy and reliability of the sound evaluation.

[0105] In some embodiments of the present application, when the vehicle sound processing device determines the sound evaluation result of the vehicle sound material according to the parameter analysis result and the evaluation result, it may perform weighted processing according to the first weight information corresponding to the parameter analysis result, the second weight information corresponding to the evaluation result, the parameter analysis result, and the evaluation result to obtain the sound evaluation result.

[0106] In an embodiment of the present application, the specific values of the first weight information and the second weight information are not limited in the present application.

[0107] Exemplarily, the first weight information corresponding to the parameter analysis result is denoted as a, and the second weight information corresponding to the evaluation result is denoted as b. Then, the vehicle sound processing device can perform weighted processing based on a, b, the parameter analysis result, and the evaluation result to obtain the final sound evaluation result.

[0108] In some embodiments of the present application, after obtaining the sound evaluation result, the vehicle sound can be designed using the sound evaluation result to improve the sound effect of the vehicle sound.

[0109] In some embodiments of the present application, before the vehicle sound processing device performs feature extraction processing on the processed vehicle sound material using the trained evaluation model to obtain feature data and performs analysis processing on the sound parameters based on the feature data to obtain the parameter analysis result, that is, before step 102, the following steps may further be included:

[0110] Step 106: Obtain vehicle audio data.

[0111] In an embodiment of the present application, before the vehicle sound processing device performs feature extraction processing on the processed vehicle sound material using the trained evaluation model to obtain feature data, it may first obtain vehicle audio data.

[0112] In an embodiment of the present application, the vehicle audio data may be audio data for constructing the training evaluation model, including audio positive samples and audio negative samples, and the vehicle audio data may include audio data of different types and different formats.

[0113] Exemplarily, the vehicle audio data may include vehicle prompt tone data, Acoustic Vehicle Alert System (AVAS) data, music data, MP4 format audio data, and Waveform Audio File Format (WAV); taking the prompt tone as an example, it supports single sound source or entire set of sound source input, and the sound source corresponds one by one to the actual vehicle application sound generation scenario and unit.

[0114] Step 107: Determine audio positive samples and audio negative samples according to the vehicle audio data.

[0115] In an embodiment of the present application, after the vehicle sound processing device obtains the vehicle audio data, it can determine audio positive samples and audio negative samples according to the vehicle audio data.

[0116] In some embodiments of the present application, when the vehicle sound processing device determines audio positive samples and audio negative samples according to vehicle audio data, it may classify the vehicle audio data according to the functional attributes of the vehicle audio data to obtain the classified audio data; then perform quality verification on the classified audio data to obtain a quality verification result; and further determine the audio positive samples and audio negative samples based on the quality verification result.

[0117] In the embodiments of the present application, the application scenario or function of the vehicle audio data can be determined according to the functional attributes of the vehicle audio data; for example, when classifying the vehicle audio data according to the functional attributes of the vehicle audio data, the obtained classified audio data may include three types of audio data: prompt tone type, music type, and engine sound wave type.

[0118] In some embodiments of the present application, the quality verification result may include two results: passing the verification and failing the verification.

[0119] In some embodiments of the present application, the audio data that passes the verification can be applied to the vehicle, and the audio data that fails the verification can be understood as the audio data that is vetoed from getting on the vehicle; that is to say, the audio data that passes the verification can be understood as having a good evaluation and can be applied to the audio data in the vehicle, while the audio data that fails the verification can be understood as having a poor evaluation and is not applied to the audio data in the vehicle.

[0120] In the embodiments of the present application, the specific operation of the quality verification is not limited in the present application. For example, the result of manual evaluation of the classified audio data can be obtained as the quality verification result, or the result of whether the audio data is allowed to be applied to the vehicle can be obtained as the quality verification result.

[0121] In some embodiments of the present application, when the vehicle sound processing device determines audio positive samples and audio negative samples based on the quality verification result, it may label the audio data with a quality verification result of passing the verification as audio positive samples; and label the audio data with a quality verification result of failing the verification as audio negative samples.

[0122] Step 108: Train the initial evaluation model using the audio positive samples and audio negative samples to obtain a trained evaluation model.

[0123] In the embodiments of the present application, after the vehicle sound processing device determines audio positive samples and audio negative samples according to the vehicle audio data, it may train the initial evaluation model using the audio positive samples and audio negative samples to obtain a trained evaluation model.

[0124] In some embodiments of the present application, the data volumes of the audio positive samples and the audio negative samples can be set in a certain proportion. For example, the data volumes of the audio positive samples and the audio negative samples can be 3:1.

[0125] In some embodiments of the present application, when training an initial evaluation model with audio positive samples and audio negative samples to obtain a trained evaluation model, a loss function between the feature information output by the initial evaluation model and the input audio positive samples and audio negative samples can be calculated, so as to update the parameters of the initial evaluation model based on the loss function to obtain a trained evaluation model.

[0126] An embodiment of the present application provides a vehicle sound processing method. A vehicle sound processing device can obtain processed vehicle sound materials; use a trained evaluation model to perform feature extraction processing on the processed vehicle sound materials to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; performing analysis processing on sound parameters based on the feature data to obtain a parameter analysis result includes at least one of the following: performing fundamental frequency analysis on the spectral feature information in the feature data to obtain a pitch analysis result; performing pitch analysis on the temporal feature information in the feature data to obtain a pitch analysis result; performing rhythm analysis on the sound segment feature information in the feature data to obtain a rhythm analysis result; performing loudness analysis on the time-frequency feature information in the feature data to obtain a loudness analysis result; performing timbre analysis on the sound source feature information in the feature data to obtain a timbre analysis result; performing sound source length analysis on the temporal feature information of the sound intensity change in the feature data to obtain a sound source length analysis result; determining a test question according to the sound type and the number of sound materials of the processed vehicle sound materials, and the number of evaluation objects; determining an evaluation result of the vehicle sound materials based on the test question; determining a sound evaluation result of the vehicle sound materials according to the parameter analysis result and the evaluation result. It can be seen that after obtaining the vehicle sound materials, the vehicle sound processing device can use the trained evaluation model to perform feature extraction on the vehicle sound materials. The extracted feature data can include different types of features, and then different types of sound parameters can be analyzed for different types of features respectively to obtain an objective parameter analysis result. The parameter analysis result can include at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result, improving the comprehensiveness of parameter analysis, and then improving the objectivity of the parameter analysis result. Thus, objective parameter analysis results in different dimensions can be obtained; at the same time, a test question matching the vehicle sound materials can be determined. The vehicle sound processing device determines the test question according to the sound type and the number of sound materials of the vehicle sound materials, and the number of evaluation objects, thereby improving the matching degree between the test question and the vehicle sound materials; and obtaining an evaluation result of the vehicle sound materials based on these test questions. Finally, the sound evaluation result of the vehicle sound materials can be obtained by referring to the evaluation result and the parameter analysis result at the same time, effectively improving the objectivity and accuracy of vehicle sound evaluation.

[0127] Based on the above embodiment, in another embodiment of the present application, exemplarily, as Figure 2 shown, the sound processing method may include the following steps:

[0128] Step 201: Obtain the sound material to be evaluated.

[0129] In the embodiments of the present application, the sound material to be evaluated can be audio files of any type and any quantity.

[0130] In the embodiments of the present application, the sound material to be evaluated can be subject to normalization processing, including noise elimination processing and audio unification processing.

[0131] Exemplarily, a microphone array can be used to collect the vehicle sound material after being subject to normalization processing.

[0132] Step 202: Conduct parameter analysis on the sound material to be evaluated in different dimensions.

[0133] In some embodiments of the present application, the pitch, tone height, rhythm, loudness, timbre, and sound source length of the sound material to be evaluated can be analyzed to obtain the parameter analysis results of these different types.

[0134] In some embodiments of the present application, a trained evaluation model can be used to perform feature extraction processing on the vehicle sound material, and then based on the obtained feature data, analysis processing of sound parameters such as pitch, tone height, rhythm, loudness, timbre, and sound source length can be carried out to obtain the corresponding parameter analysis results.

[0135] In some embodiments of the present application, a trained evaluation model can be used to extract high-dimensional abstract features of the vehicle sound material and output key feature data. Since the feature data can include features in different dimensions, the performance of the processed vehicle sound material in terms of parameters in different dimensions can be measured based on the feature data, improving the reliability and comprehensiveness of sound parameter analysis.

[0136] In some embodiments of the present application, the feature data can include spectral feature information, temporal feature information, sound segment feature information, time-frequency feature information, sound source feature information, and sound intensity change temporal feature information.

[0137] In some embodiments of the present application, exemplarily, some tools can also be used to implement the parameter analysis of the sound material. For example, the Mel Frequency Cepstral Coefficients (MFCC) of the sound material to be evaluated can be extracted through a Python library such as LibROSA for audio and music analysis, and then the pitch parameter analysis can be carried out based on the Mel Frequency Cepstral Coefficients.

[0138] Exemplarily, some communication protocols such as WebSocket can also be combined for timely data feedback to improve the analysis efficiency of different-dimensional sound parameters according to different features.

[0139] Step 203: Obtain the manual evaluation result of the voice material to be evaluated.

[0140] In some embodiments of the present application, the manual evaluation result may be the answer result of the voice material to be evaluated for the evaluation questions.

[0141] In some embodiments of the present application, the evaluation questions correspond to the voice material to be evaluated, and the matching evaluation questions can be determined from the questions in the evaluation question bank.

[0142] In some embodiments of the present application, a large number of evaluation questions can be generated based on the principles of safety priority, standard consistency, functional beneficiary, intuitive communication, moderate arousal, sensory comfort, moderate sensory channel, emotional care, and image association, and an evaluation question bank can be constituted according to the large number of evaluation questions. Furthermore, the evaluation questions matching the voice material to be evaluated can be selected from the evaluation question bank for the evaluation object to answer the evaluation questions, so as to obtain the manual evaluation result.

[0143] In some embodiments of the present application, the evaluation questions can be determined according to the voice type and the number of materials of the voice material to be evaluated, as well as the number of evaluation objects.

[0144] It can be understood that in the embodiments of the present application, the execution order of Step 202 and Step 203 is not limited in the present application. For example, Figure 2 as shown, Step 202 is executed first, and then Step 203; Step 203 can also be executed first, and then Step 202; Step 203 and Step 202 can also be executed simultaneously.

[0145] Step 204: Weight the parameter analysis result and the evaluation result to obtain the voice evaluation result.

[0146] In some embodiments of the present application, the voice evaluation result can be obtained by performing a weighting process according to the weight corresponding to the parameter analysis result, the weight corresponding to the evaluation result, the parameter analysis result, and the evaluation result.

[0147] In the embodiments of the present application, by evaluating the voice through the parameter analysis result and the evaluation result, the parameters of the voice material from the objective evaluation perspective and the evaluation result of the voice material from the subjective evaluation perspective can be comprehensively considered, thereby improving the accuracy and reliability of the voice evaluation result.

[0148] In some embodiments of the present application, after obtaining the voice evaluation result, the voice evaluation result can be summarized and analyzed to obtain the feedback information of the voice evaluation result, so as to design the vehicle voice according to the feedback information to improve the voice effect of the vehicle voice.

[0149] In summary, by introducing the user experience principle and combining the subjective evaluation of the evaluation object with the usability principle, the embodiments of the present application make the evaluation more in line with the actual usage situation. Thus, based on the user experience principle, the physical parameters of the sound, and the functional attributes, a comprehensive evaluation of the vehicle sound is realized, which can comprehensively evaluate the multi-dimensional characteristics of the sound, including emotions, information transmission effects, etc. Among them, due to the introduction of sound parameters, the influence of subjective evaluation on the evaluation results can be reduced, and data deviation caused by the evaluation object's unfamiliarity with terms can be minimized. Furthermore, the evaluation results can be used to design and improve the vehicle sound to meet the user's expectations for the vehicle sound and enhance the driving experience.

[0150] An embodiment of the present application provides a vehicle sound processing method. A vehicle sound processing device can obtain processed vehicle sound materials; use a trained evaluation model to perform feature extraction processing on the processed vehicle sound materials to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; performing analysis processing on sound parameters based on the feature data to obtain a parameter analysis result includes at least one of the following: performing fundamental frequency analysis on the spectral feature information in the feature data to obtain a pitch analysis result; performing pitch analysis on the temporal feature information in the feature data to obtain a pitch analysis result; performing rhythm analysis on the sound segment feature information in the feature data to obtain a rhythm analysis result; performing loudness analysis on the time-frequency feature information in the feature data to obtain a loudness analysis result; performing timbre analysis on the sound source feature information in the feature data to obtain a timbre analysis result; performing sound source length analysis on the temporal feature information of the sound intensity change in the feature data to obtain a sound source length analysis result; determining a test question based on the sound type and the number of the vehicle sound materials after processing, and the number of evaluation objects; determining an evaluation result of the vehicle sound materials based on the test question; determining a sound evaluation result of the vehicle sound materials based on the parameter analysis result and the evaluation result. It can be seen that after obtaining the vehicle sound materials, the vehicle sound processing device can use the trained evaluation model to perform feature extraction on the vehicle sound materials. The extracted feature data can include different types of features. Furthermore, different types of features can be respectively analyzed for different types of sound parameters to obtain an objective parameter analysis result. The parameter analysis result can include at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result, improving the comprehensiveness of the parameter analysis, and further improving the objectivity of the parameter analysis result. Thus, objective parameter analysis results in different dimensions can be obtained; at the same time, a test question matching the vehicle sound materials can be determined. The vehicle sound processing device determines the test question according to the sound type and the number of the vehicle sound materials, and the number of evaluation objects. Thus, the matching degree between the test question and the vehicle sound materials can be improved; and an evaluation result of the vehicle sound materials can be obtained based on these test questions. Finally, the sound evaluation result of the vehicle sound materials can be obtained by referring to both the evaluation result and the parameter analysis result, effectively improving the objectivity and accuracy of the vehicle sound evaluation.

[0151] Based on the above embodiment, in another embodiment of the present application, Figure 3 is a schematic composition structure of the vehicle sound processing device provided by the embodiment of the present application Figure 1 , as Figure 3As shown in the figure, the vehicle sound processing device 1 provided by the embodiment of the present application may include an acquisition unit 11, an analysis unit 12, a determination unit 13, and a training unit 14.

[0152] The acquisition unit 11 can be used to acquire vehicle sound materials.

[0153] The analysis unit 12 can be used to perform feature extraction processing on the processed vehicle sound materials by using the trained evaluation model to obtain feature data, and perform analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; and is used to perform at least one of the following: perform fundamental frequency analysis according to the spectral feature information in the feature data to obtain a pitch analysis result; perform pitch analysis according to the timing feature information in the feature data to obtain a pitch analysis result; perform rhythm analysis according to the sound paragraph feature information in the feature data to obtain a rhythm analysis result; perform loudness analysis according to the time-frequency feature information in the feature data to obtain a loudness analysis result; perform timbre analysis according to the sound source feature information in the feature data to obtain a timbre analysis result; perform sound source length analysis according to the sound intensity change timing feature information in the feature data to obtain a sound source length analysis result.

[0154] The determination unit 13 can be used to determine a test question according to the sound type and the number of materials of the processed vehicle sound materials, and the number of evaluation objects; and determine the evaluation result of the processed vehicle sound materials based on the test question, and determine the sound evaluation result of the processed vehicle sound materials according to the parameter analysis result and the evaluation result.

[0155] In some embodiments of the present application, the sound type includes a prompt sound type, an engine sound wave type, and a music type; the evaluation object represents an object that answers the test question.

[0156] In some embodiments of the present application, the acquisition unit 12 can also be used to acquire vehicle sound materials; and perform noise elimination processing and audio unification processing on the vehicle sound materials to obtain the processed vehicle sound materials.

[0157] The training unit 14 can be used to acquire vehicle audio data before the analysis unit 12 performs feature extraction processing on the processed vehicle sound materials by using the trained evaluation model to obtain feature data, and perform analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result; and determine audio positive samples and audio negative samples according to the vehicle audio data; and use the audio positive samples and the audio negative samples to train the initial evaluation model to obtain the trained evaluation model.

[0158] In some embodiments of the present application, the training unit 14 may also be used to classify the vehicle audio data according to the functional attributes of the vehicle audio data to obtain the classified audio data; and perform quality verification on the classified audio data to obtain a quality verification result; and determine audio positive samples and audio negative samples based on the quality verification result.

[0159] In some embodiments of the present application, the training unit 14 may also be used to label the audio data with a passed quality verification result as an audio positive sample; and label the audio data with a failed quality verification result as an audio negative sample.

[0160] In some embodiments of the present application, the determination unit 13 may also be used to perform weighted processing according to the first weight information corresponding to the parameter analysis result, the second weight information corresponding to the evaluation result, the parameter analysis result, and the evaluation result to obtain a voice evaluation result.

[0161] In the embodiments of the present application, further Figure 4 is a schematic composition structure of the vehicle sound processing device provided by the embodiments of the present application Figure 2 , as Figure 4 shown, the vehicle sound processing device 1 provided by the embodiments of the present application may further include a processor 15, a memory 16 storing executable instructions of the processor 15; further, the vehicle sound processing device 1 may further include a communication interface 17, and a bus 18 for connecting the processor 15, the memory 16, and the communication interface 17.

[0162] In an embodiment of the present application, the above-mentioned processor 15 may be at least one of an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a Central Processing Unit (CPU), a controller, a microcontroller, and a microprocessor. It can be understood that for different devices, the electronic devices used to implement the functions of the above-mentioned processor may also be others, and the embodiments of the present application do not make specific limitations. The vehicle sound processing device 1 may further include a memory 16, and the memory 16 may be connected to the processor 15. Among them, the memory 16 is used to store executable program codes, and the program codes include computer operation instructions. The memory 16 may include a high-speed RAM memory, and may also include a non-volatile memory, for example, at least two disk memories.

[0163] In an embodiment of the present application, the bus 18 is used to connect the communication interface 17, the processor 15, and the memory 16 and for mutual communication between these devices.

[0164] In an embodiment of the present application, the memory 16 is used to store instructions and data.

[0165] Further, in the embodiments of the present application, the above-mentioned processor 15 is configured to obtain the processed vehicle sound material; perform feature extraction processing on the processed vehicle sound material by using the trained evaluation model to obtain feature data, and perform analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result; wherein the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; performing analysis processing on the sound parameters according to the feature data to obtain a parameter analysis result includes at least one of the following: performing fundamental frequency analysis according to the spectral feature information in the feature data to obtain a pitch analysis result; performing pitch analysis according to the temporal feature information in the feature data to obtain a pitch analysis result; performing rhythm analysis according to the sound segment feature information in the feature data to obtain a rhythm analysis result; performing loudness analysis according to the time-frequency feature information in the feature data to obtain a loudness analysis result; performing timbre analysis according to the sound source feature information in the feature data to obtain a timbre analysis result; performing sound source length analysis according to the temporal feature information of the sound intensity change in the feature data to obtain a sound source length analysis result; determining a test question according to the sound type and the number of materials of the processed vehicle sound material, and the number of evaluation objects; determining an evaluation result of the vehicle sound material based on the test question; and determining a sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result.

[0166] In practical applications, the above-mentioned memory 16 may be a volatile memory, such as a random access memory (RAM); or a non-volatile memory, such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); or a combination of the above types of memories, and provide instructions and data to the processor 15.

[0167] In addition, each functional module in this embodiment may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional module.

[0168] When the integrated unit is implemented in the form of a software functional module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method of this embodiment. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs.

[0169] An embodiment of the present application provides a vehicle sound processing device. The vehicle sound processing device can obtain processed vehicle sound materials; use a trained evaluation model to perform feature extraction processing on the processed vehicle sound materials to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; where the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; performing analysis processing on sound parameters based on the feature data to obtain a parameter analysis result includes at least one of the following: performing fundamental frequency analysis on the spectral feature information in the feature data to obtain a pitch analysis result; performing pitch analysis on the temporal feature information in the feature data to obtain a pitch analysis result; performing rhythm analysis on the sound segment feature information in the feature data to obtain a rhythm analysis result; performing loudness analysis on the time-frequency feature information in the feature data to obtain a loudness analysis result; performing timbre analysis on the sound source feature information in the feature data to obtain a timbre analysis result; performing sound source length analysis on the temporal feature information of the sound intensity change in the feature data to obtain a sound source length analysis result; determining a test question based on the sound type and the number of the vehicle sound materials after processing, and the number of evaluation objects; determining an evaluation result of the vehicle sound materials based on the test question; determining a sound evaluation result of the vehicle sound materials based on the parameter analysis result and the evaluation result. It can be seen that after obtaining the vehicle sound materials, the vehicle sound processing device can use the trained evaluation model to perform feature extraction on the vehicle sound materials. The extracted feature data can include different types of features, and then different types of sound parameters can be analyzed for different types of features respectively to obtain an objective parameter analysis result. The parameter analysis result can include at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result, improving the comprehensiveness of the parameter analysis and further improving the objectivity of the parameter analysis result. Thus, objective parameter analysis results in different dimensions can be obtained; at the same time, a test question matching the vehicle sound materials can be determined. The vehicle sound processing device determines the test question according to the sound type and the number of the vehicle sound materials after processing, and the number of evaluation objects. Thus, the matching degree between the test question and the vehicle sound materials can be improved; and an evaluation result of the vehicle sound materials can be obtained based on these test questions. Finally, the sound evaluation result of the vehicle sound materials can be obtained by referring to both the evaluation result and the parameter analysis result, effectively improving the objectivity and accuracy of the vehicle sound evaluation.

[0170] Specifically, the program instructions corresponding to a vehicle sound processing method in this embodiment can be stored on storage media such as optical discs, hard disks, USB flash drives, etc. When the program instructions corresponding to a vehicle sound processing method in the storage media are read or executed by a vehicle sound processing device, the following steps are included:

[0171] Obtain the processed vehicle sound materials;

[0172] Use the trained evaluation model to perform feature extraction processing on the processed vehicle sound materials to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result;

[0173] Perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result, including at least one of the following:

[0174] Perform fundamental frequency analysis based on the spectral feature information in the feature data to obtain a pitch analysis result;

[0175] Perform pitch analysis based on the timing feature information in the feature data to obtain a pitch analysis result;

[0176] Perform rhythm analysis based on the sound paragraph feature information in the feature data to obtain a rhythm analysis result;

[0177] Perform loudness analysis based on the time-frequency feature information in the feature data to obtain a loudness analysis result;

[0178] Perform timbre analysis based on the sound source feature information in the feature data to obtain a timbre analysis result;

[0179] Perform sound source length analysis based on the sound intensity change timing feature information in the feature data to obtain a sound source length analysis result;

[0180] Determine the evaluation questions according to the sound type and quantity of the processed vehicle sound materials, and the number of evaluation objects;

[0181] Determine the evaluation result of the processed vehicle sound materials based on the evaluation questions;

[0182] Determine the sound evaluation result of the processed vehicle sound materials according to the parameter analysis result and the evaluation result.

[0183] This application embodiment provides a computer program product, including a computer program or instruction. When the computer program or instruction is executed by a processor, the steps in the vehicle sound processing method are implemented.

[0184] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories and optical memories, etc.) containing computer-usable program code.

[0185] The present application is described with reference to the schematic flowcharts and / or block diagrams of the implementation processes of methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the schematic flowcharts and / or block diagrams, as well as the combination of processes and / or blocks in the schematic flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to generate a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0186] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means, and the instruction means implements the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0187] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0188] The above embodiments are only preferred embodiments cited to fully illustrate the present application, and the protection scope of the present application is not limited thereto. Equivalent substitutions or transformations made by those skilled in the art on the basis of the present application are within the protection scope of the present application.

Claims

1. A vehicle sound processing method, characterized in that, The method includes: Obtaining the processed vehicle sound material; Performing feature extraction processing on the processed vehicle sound material by using the trained evaluation model to obtain feature data, and performing analysis processing on sound parameters according to the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; The performing analysis processing on sound parameters according to the feature data to obtain a parameter analysis result includes at least one of the following: Performing fundamental frequency analysis according to the spectral feature information in the feature data to obtain the pitch analysis result; Performing pitch analysis according to the temporal feature information in the feature data to obtain the pitch analysis result; Performing rhythm analysis according to the sound paragraph feature information in the feature data to obtain the rhythm analysis result; Performing loudness analysis according to the time-frequency feature information in the feature data to obtain the loudness analysis result; Performing timbre analysis according to the sound source feature information in the feature data to obtain the timbre analysis result; Performing sound source length analysis according to the temporal feature information of sound intensity change in the feature data to obtain the sound source length analysis result; Determining a test question according to the sound type and the number of the processed vehicle sound materials, and the number of evaluation objects; Determining an evaluation result of the processed vehicle sound material based on the test question; Determining a sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result; 2. The method according to claim 1, wherein The sound type includes a prompt sound type, an engine sound wave type, and a music type; the evaluation object represents an object that answers the test question; 3. The method according to claim 1 or 2, wherein The obtaining the processed vehicle sound material includes: Obtaining vehicle sound material; Performing noise elimination processing and audio unification processing on the vehicle sound material to obtain the processed vehicle sound material; 4. The method according to claim 3, characterized in that, Before performing feature extraction processing on the processed vehicle sound material by using the trained evaluation model to obtain feature data, and performing analysis processing on sound parameters according to the feature data to obtain a parameter analysis result, the method further includes: Obtaining vehicle audio data; Determining an audio positive sample and an audio negative sample according to the vehicle audio data; Training an initial evaluation model by using the audio positive sample and the audio negative sample to obtain the trained evaluation model; 5. The method according to claim 4, characterized in that, The determining an audio positive sample and an audio negative sample according to the vehicle audio data includes: Classifying the vehicle audio data according to the functional attributes of the vehicle audio data to obtain classified audio data; Performing quality verification on the classified audio data to obtain a quality verification result; Determining the audio positive sample and the audio negative sample based on the quality verification result; 6. The method according to claim 5, wherein The determining the audio positive sample and the audio negative sample based on the quality verification result includes: Labeling the audio data with a passed quality verification result as the audio positive sample; Label the audio data with a failed quality verification result as the audio negative sample.

7. The method according to claim 1, characterized in that, Determining the sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result includes: Performing weighted processing on the first weight information corresponding to the parameter analysis result, the second weight information corresponding to the evaluation result, the parameter analysis result, and the evaluation result to obtain the sound evaluation result.

8. A vehicle sound processing device, characterized in that, Including an acquisition unit, an analysis unit, and a determination unit; The acquisition unit is used to acquire the processed vehicle sound material; The analysis unit is used to perform feature extraction processing on the processed vehicle sound material using a trained evaluation model to obtain feature data, and perform analysis processing on sound parameters based on the feature data to obtain a parameter analysis result; wherein, the parameter analysis result includes at least one of a pitch analysis result, a pitch analysis result, a rhythm analysis result, a loudness analysis result, a timbre analysis result, and a sound source length analysis result; and is used to perform at least one of the following: perform fundamental frequency analysis on the spectral feature information in the feature data to obtain the pitch analysis result; perform pitch analysis on the temporal feature information in the feature data to obtain the pitch analysis result; perform rhythm analysis on the sound segment feature information in the feature data to obtain the rhythm analysis result; perform loudness analysis on the time-frequency feature information in the feature data to obtain the loudness analysis result; perform timbre analysis on the sound source feature information in the feature data to obtain the timbre analysis result; perform sound source length analysis on the sound intensity change temporal feature information in the feature data to obtain the sound source length analysis result; The determination unit is used to determine the evaluation questions according to the sound type and the number of materials of the processed vehicle sound material, and the number of evaluation objects; and determine the evaluation result of the processed vehicle sound material based on the evaluation questions, and determine the sound evaluation result of the processed vehicle sound material according to the parameter analysis result and the evaluation result.

9. A vehicle sound processing device, characterized in that, The vehicle sound processing device includes a processor and a memory storing instructions executable by the processor; when the instructions are executed by the processor, the method according to any one of claims 1 to 7 is implemented.

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

11. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the processor, the steps in the vehicle sound processing method according to any one of claims 1 to 7 are implemented.

Images