In-vehicle audio adjusting method, system and device and storage medium

Abstract

The invention discloses an in-vehicle audio adjusting method, system and device and a storage medium. The method comprises the steps that audio signals of areas where different persons are located in a vehicle are acquired; performing sound quality prediction on the audio signal by adopting the trained score prediction model to obtain a score result of the audio signal; the trained score prediction model is determined based on preference information of people in the vehicle for the sound quality; determining a first preset scoring condition corresponding to the current environment information of the vehicle; under the condition that the scoring result does not meet the first preset scoring condition, adjusting the audio signal for at least one time; the adjusted scoring result of the audio signal meets a preset scoring condition corresponding to the current environment information. In this way, fine and intelligent adjustment of the audio in the vehicle is achieved, and personalized preferences of drivers and passengers are met.

Description

Technical Field

[0001] The present application relates to the field of sound quality control and acoustic environment improvement, and in particular to an in-vehicle audio adjustment method, system, device and storage medium. Background Art

[0002] At present, the control methods of in-car sound quality mainly focus on passive noise reduction technology and the processing and analysis of sound signals to optimize the in-car acoustic environment. However, these methods generally lack sufficient consideration of the dynamic changes in sound quality under different road conditions, and fail to flexibly adjust the in-car acoustic environment according to the real-time changes in environmental information such as road conditions. Therefore, there are limitations in achieving refined and intelligent control of the in-car acoustic environment, making it difficult to ensure the best riding experience for passengers. In addition, existing methods often only focus on a single dimension of sound quality, and therefore cannot meet the personalized needs of different customer groups with different preferences. Summary of the invention

[0003] In view of this, the embodiments of the present application at least provide an in-vehicle audio adjustment method, system, device and storage medium.

[0004] The technical solution of the embodiment of the present application is as follows:

[0005] In a first aspect, an embodiment of the present application provides an in-vehicle audio adjustment method, which includes: obtaining audio signals of different areas where people are located in the vehicle; using a trained scoring prediction model to predict the sound quality of the audio signal to obtain a scoring result of the audio signal; the trained scoring prediction model is determined based on the preference information of people in the vehicle for sound quality; determining a first preset scoring condition corresponding to the current environmental information of the vehicle; if the scoring result does not meet the first preset scoring condition, adjusting the audio signal at least once; the scoring result of the adjusted audio signal meets the preset scoring condition corresponding to the current environmental information.

[0006] In this way, the audio signals of different areas where different people are located in the vehicle, as well as the preference information of the people in the vehicle for sound quality, are obtained, and the sound quality of the obtained audio signals is predicted using the trained scoring prediction model to obtain the scoring results of the audio signals; the first preset scoring condition corresponding to the current environmental information of the vehicle is determined; if the scoring result does not meet the first preset scoring condition, the audio signal is adjusted at least once so that the scoring result of the adjusted audio signal meets the preset scoring condition corresponding to the current environmental information. By using the trained scoring prediction model to predict the audio signal in real time, the sound quality of the audio signal can be analyzed and evaluated in real time, thereby satisfying the personalized preferences of the drivers and passengers in a dynamic environment, and making adjustments according to different environmental information, so as to achieve refined and intelligent adjustment of the audio in the car.

[0007] Furthermore, the in-vehicle audio adjustment method also includes: obtaining preference information of people in the vehicle for sound quality; the preference information represents the degree of preference of people in the vehicle for sound quality in different dimensions; based on the preference information of people in the vehicle for sound quality, determining a trained scoring prediction model.

[0008] In this way, by obtaining the preference information of people in the vehicle for sound quality and determining the scoring prediction model corresponding to the preference information, a user-centered design can be achieved, so that the audio adjustment is more in line with the individual needs of the passengers and the personalized preferences of the driver and passengers are met.

[0009] Furthermore, determining a first preset scoring condition corresponding to the current environmental information of the vehicle includes: obtaining road surface image information of the vehicle on the current driving road; based on the road surface image information, identifying the road type of the current driving road; and determining the first preset scoring condition based on the road type of the current driving road.

[0010] In this way, since different road types have different impacts on passengers' preferences and audio signals, by adjusting the corresponding scoring conditions, it is possible to achieve refined and intelligent adjustment of the in-car audio, meet the personalized preferences of drivers and passengers, and optimize the riding experience of drivers and passengers.

[0011] Further, determining the first preset scoring condition based on the road type of the current driving road includes: determining the first preset scoring condition based on the driving state information of the vehicle and the road type of the current driving road.

[0012] In this way, by comprehensively considering the vehicle's driving status and road type, the system can formulate scoring conditions that are more in line with the actual driving environment, thereby optimizing the riding experience of the driver and passengers.

[0013] Furthermore, the first preset scoring condition includes multiple scoring intervals and a scoring mark corresponding to each scoring interval; when the scoring result does not meet the first preset scoring condition, the audio signal is adjusted at least once, including: determining a target scoring interval of the scoring result in the multiple scoring intervals; when the scoring mark corresponding to the target scoring interval is a preset scoring mark, determining that the scoring result does not meet the first preset scoring condition, and adjusting the audio signal at least once.

[0014] In this way, the system monitors the scoring results in real time and makes adjustments to ensure that the audio signal always remains in the best condition in a dynamic driving environment, thereby optimizing the riding experience of the driver and passengers.

[0015] Furthermore, the audio signal is adjusted at least once, including: determining a first audio suppression signal of the audio signal; adjusting the audio signal based on the first audio suppression signal to obtain a first audio signal; when the current environmental information of the vehicle changes, determining a second preset scoring condition corresponding to the changed current environmental information of the vehicle; when the first audio signal does not meet the second preset scoring condition, adjusting the first audio signal based on a second audio suppression signal corresponding to the first audio signal to obtain a second audio signal.

[0016] In this way, the scoring conditions can be dynamically adjusted according to changes in the vehicle's current environmental information, making the system more intelligent and adaptable in different driving scenarios, and providing drivers and passengers with a better user experience.

[0017] Furthermore, the training process of the trained rating prediction model includes: collecting multiple sample audio information in the vehicle under different scenarios and / or different vehicle speeds; for each sample audio information, obtaining multiple rating label information of the sample audio information under multiple sound quality rating standards; the multiple sound quality rating standards correspond one-to-one with the preference information of people in multiple vehicles for sound quality; based on each sample audio information and the multiple rating label information corresponding to each sample audio information, training an initial rating prediction model to obtain multiple trained rating prediction models; the multiple trained rating prediction models correspond one-to-one with the preference information of people in multiple vehicles for sound quality.

[0018] In this way, by adopting multiple sound quality scoring standards, the model can comprehensively consider different acoustic characteristics, thereby obtaining a more comprehensive evaluation result of the audio signal. In addition, by continuously updating and training the scoring prediction model, it is possible to provide drivers and passengers with changes in preferences and new environments, meet their personalized preferences, and optimize their riding experience.

[0019] In a second aspect, an embodiment of the present application provides an in-vehicle audio adjustment device, which includes: an acquisition module for acquiring audio signals of areas where different people are located in the vehicle; a prediction module for using a trained scoring prediction model to predict the sound quality of the audio signal to obtain a scoring result of the audio signal; the trained scoring prediction model is determined based on the preference information of the people in the vehicle for sound quality; a determination module is used to determine a first preset scoring condition corresponding to the current environmental information of the vehicle; an adjustment module is used to adjust the audio signal at least once when the scoring result does not meet the first preset scoring condition; the scoring result of the adjusted audio signal meets the preset scoring condition corresponding to the current environmental information.

[0020] In a third aspect, an embodiment of the present application provides an in-vehicle audio adjustment device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the program, some or all of the steps in the above method are implemented.

[0021] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having a computer program stored thereon, which implements some or all of the steps in the above method when executed by a processor.

[0022] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program or instructions, which implement some or all of the steps in the above method when executed by a processor.

[0023] It should be understood that the above general description and the following detailed description are merely exemplary and explanatory, and are not intended to limit the technical solutions of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0025] Figure 1 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 1 ;

[0026] Figure 2 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 2 ;

[0027] Figure 3 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 3 ;

[0028] Figure 4 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 4 ;

[0029] Figure 5 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 5 ;

[0030] Figure 6 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 6 ;

[0031] Figure 7 A schematic diagram of the architecture of an in-vehicle audio adjustment method provided in an embodiment of the present application;

[0032] Figure 8 A schematic diagram of a hybrid model structure of an in-car audio adjustment method provided in an embodiment of the present application;

[0033] Fig. 9 A schematic diagram of the structure of an in-car audio adjustment device provided in an embodiment of the present application;

[0034] Fig.10 A hardware entity schematic diagram of an in-vehicle audio adjustment device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0035] The following will describe the implementation methods 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 contents disclosed in this specification. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways 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, not for limiting the scope of protection of the present application.

[0036] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application, and thus the drawings only show components related to the present application rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.

[0037] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0038] The terms "first / second / third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It is understandable that "first / second / third" can be interchanged with a specific order or sequence where permitted so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing this application and are not intended to limit this application.

[0040] Before describing the sequence processing method in the embodiment of the present application, the professional terms involved in the embodiment of the present application are first explained:

[0041] Active Noise Cancellation (ANC): Reduces the noise in the current environment by generating sound waves that are opposite to the noise signals in the current environment.

[0042] Convolutional Neural Network (CNN): As a representative algorithm of deep learning, it has excellent performance in image classification, target detection and other fields, and has attracted much attention and broad application prospects. Convolutional neural networks are directly driven by data, and establish a mapping relationship between input and output through multi-dimensional nonlinear feature extraction. They have extremely strong data representation and mapping capabilities.

[0043] Recurrent Neural Network (RNN): It is a neural network architecture for processing sequence data, which can retain and transmit information in time series.

[0044] Long Short-Term Memory Network (LSTM): LSTM is a network structure of recurrent neural network in deep learning. It is mostly used for recognition and classification tasks with time series characteristics, such as machine translation, weather forecasting, and audio recognition. It solves the gradient vanishing and gradient exploding problems of RNN. By introducing the gating mechanism, namely the input gate, forget gate, and output gate, it effectively controls the data flow and information preservation, improves the learning ability and generalization performance of the model, and breaks through the limitations of RNN.

[0045] With the rapid development of the automobile industry, consumers have higher and higher requirements for the comfortable driving experience of automobiles. As one of the important factors affecting the driving experience, the in-car sound quality has received widespread attention, and the control and design of the in-car sound quality has become an important part of vehicle research and development. In the related technology, the original noise signal is first obtained, that is, the real-time noise signal of the engine and the noise signal of the human ear head area; then, according to the set target sound quality noise signal, the original noise signal is processed by active noise reduction and acoustic compensation methods; finally, the processed noise signal is synthesized with the original noise signal to achieve active control of the sound quality in the vehicle.

[0046] It can be seen that most of the in-car sound quality control methods are based on passive noise reduction technology or processing and analysis of sound signals to improve the in-car sound environment. However, they often lack consideration of changes in sound quality under different road conditions, and cannot adjust the in-car sound environment in real time according to road conditions, making it difficult to achieve refined and intelligent adjustment of the in-car sound environment to provide the best riding experience. In addition, existing methods often only focus on a single dimension of sound quality, such as comfort, while ignoring the comprehensive consideration of multi-dimensional sound quality such as power and sportiness, and cannot meet the personalized needs of different customer groups.

[0047] In order to solve the problems of the related art, an embodiment of the present application provides an in-vehicle audio adjustment method. Figure 1 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 1 The in-vehicle audio adjustment method in the embodiment of the present application can be executed by an audio adjustment device running in a vehicle computer, such as Figure 1 As shown, the in-car audio adjustment method can be implemented through steps S101 to S104:

[0048] Step S101, obtaining audio signals of areas where different persons are located in the vehicle.

[0049] Among them, audio signals are information carriers of regular frequency and amplitude changes of sound waves with voice, music and sound effects. These sound waves are transmitted through a medium (such as air) and are perceived by the human auditory system.

[0050] In some embodiments, the audio signal includes not only clear and orderly signals (such as music and speech), but also noise, wherein noise is a random, generally considered useless audio signal that may interfere with the transmission or perception of useful signals but is still part of the audio signal.

[0051] It is understandable that inside a vehicle, noise may come from multiple sources, such as the engine, tires, wind noise, and conversations between passengers. By acquiring audio signals from different areas, the location and characteristics of each noise source can be accurately identified.

[0052] In some embodiments, obtaining audio signals of different areas where people are located in the vehicle may include obtaining audio signals of different areas where people are located in the vehicle based on a microphone array. The microphone array is a group of omnidirectional microphones located at different positions in space and arranged into an array according to certain shape rules. Exemplarily, 4 distributed microphones are deployed in the car, which are installed on the roof handrails near the main driver's seat, the co-driver's seat, the left rear row, and the right rear row passengers, respectively, to achieve a distributed 4-microphone sound reception in the car, corresponding to the main driver's seat, the co-driver's seat, the left rear row, and the right rear row, respectively, so as to obtain the audio signal of each passenger's area.

[0053] Step S102: Use the trained score prediction model to predict the sound quality of the audio signal to obtain a score result of the audio signal.

[0054] In the embodiment of the present application, the trained score prediction model is determined based on the preference information of the people in the vehicle regarding the sound quality.

[0055] In the embodiment of the present application, the score prediction model is a mathematical model that can predict the user or system's score on sound quality through training and learning. For example, the score prediction model can be trained by using a large amount of data, automatically extracting features and patterns using an algorithm, and then improving model performance through manual annotation, adjusting model parameters, and optimizing algorithms.

[0056] Among them, sound quality refers to the characteristics and quality of sound, which is usually composed of multiple factors. For example, sound quality can include A-weighted total sound pressure level, order sound pressure level, loudness, roughness, sharpness, fluctuation, language clarity, etc.

[0057] The preference information of sound quality refers to the subjective preference of the occupants of the vehicle for different sound characteristics and qualities. For example, the preference information of sound quality may include comfort, power and sportiness.

[0058] In some embodiments, each preference information may correspond to a score prediction model. For example, when it is determined that the user's preference information is comfort, the score prediction model trained for comfort is selected to predict the audio signal in the car; when it is determined that the user's preference information is dynamic, the score prediction model trained for dynamic is selected to predict the audio signal in the car.

[0059] It is understandable that different preference information has significant differences in features and influencing factors. Therefore, it is necessary to model each preference information separately to obtain a score prediction model corresponding to the preference information, which can better capture the complexity and uniqueness of each score prediction model, thereby improving the prediction accuracy. In addition, training each preference information separately facilitates optimization and adjustment for each indicator, simplifying model design and analysis.

[0060] In some embodiments, the score prediction model may include a hybrid model composed of a convolutional neural network (CNN) and a knowledge vector regression (KVR). Exemplarily, the hybrid model may be CNN-KVR.

[0061] In some embodiments, the method of obtaining the preference information of sound quality may include collecting the preference information of sound quality input by the driver and passenger. For example, the driver and passenger may set the preference degree for multiple perception dimensions such as comfort, power and sportiness on the multimedia system or vehicle setting interface of the vehicle, and the audio adjustment device running in the vehicle computer may obtain the preference information of sound quality input by the driver and passenger.

[0062] In some embodiments, the method of obtaining the preference information of sound quality may also include obtaining the preference information of sound quality by analyzing the historical data of the preference information. Exemplarily, the historical data of the preference information is counted and analyzed using correlation analysis to identify the user's sound quality preference trend. Correlation analysis is a statistical method for evaluating the strength and direction of the relationship between two or more variables, and determining the linear relationship between the variables by calculating the coefficients between the variables.

[0063] In some embodiments, the method of obtaining the preference information of sound quality may also include, when the preference information of the sound quality of the number of drivers and passengers is not obtained, or the historical data of the preference information is not obtained, using the preset preference information as the preference information of the sound quality of the people in the vehicle. For example, the preset preference information may be comfort.

[0064] Step S103, determining a first preset scoring condition corresponding to the current environmental information of the vehicle.

[0065] The current environment information of the vehicle refers to various data and states of the surrounding environment perceived by the vehicle during driving. For example, the current environment of the vehicle may include the type of road on which the vehicle is driving, the driving state information of the vehicle, environmental conditions, and traffic conditions.

[0066] It can be understood that the first preset scoring condition is used to determine whether the scoring result output by the scoring prediction model meets the scoring standard corresponding to the current environmental information. The environmental information corresponds to the first preset scoring condition one by one, and the first preset scoring condition corresponding to different environmental information may be different. Exemplarily, when the environmental information of the vehicle is that the road type is dry, because the contact between the car tires and the road surface is stable when the ground is dry, and the sound generated during driving is small, the scoring requirement for the sound is high, and the corresponding first pre-scoring condition can be set to a scoring result greater than or equal to 8. When the environmental information of the vehicle is that the road type is slippery, because the moisture on the ground will reduce the friction between the tires and the ground, it is inevitable to generate noise, so the scoring requirement for the sound is low, and the corresponding first preset scoring condition can be set to a scoring result greater than or equal to 6.

[0067] In some embodiments, obtaining the current environment information of the vehicle may include obtaining the current environment information of the vehicle by analyzing sensor data collected by a sensor.

[0068] For example, sensor data containing current environmental information of the vehicle can be collected based on vehicle-mounted sensors (such as radar, camera, and microphone, etc.), and feature extraction can be performed on the sensor data. Speed ​​characteristics, road type characteristics, and location characteristics of the vehicle in the sensor data can be extracted.

[0069] In some embodiments, obtaining the current environment information of the vehicle may also include obtaining the current environment information of the vehicle through cloud data and a global positioning system (GPS). The cloud data includes global real-time traffic information and road information. Exemplarily, the audio adjustment device obtains the current location information of the vehicle through the global positioning system, and queries the cloud data through the current location information to obtain road information (such as road type) in the current area.

[0070] In some embodiments, obtaining the current environment information of the vehicle may also include using at least one artificial intelligence model to analyze the image information collected during the vehicle's driving process to obtain the current environment information of the vehicle. Exemplarily, the camera in the vehicle inputs the collected image information into a hybrid model CNN-LSTM model composed of a convolutional neural network CNN and a long short-term memory network LSTM, wherein the convolutional neural network is used to extract the spatial features of the image information, that is, the geometric and distribution characteristics of an object in the image information, and the long short-term memory network is used to capture the time dependency of the image information, that is, the characteristics of a certain feature in the image information that changes over time. Through the CNN-LSTM hybrid model, the current environment information of the vehicle can be efficiently identified.

[0071] Step S104: When the scoring result does not satisfy the first preset scoring condition, the audio signal is adjusted at least once.

[0072] In the embodiment of the present application, the scoring result of the adjusted audio signal satisfies the preset scoring condition corresponding to the current environmental information. Wherein, when the current environmental information remains unchanged, the preset scoring condition may be the first preset scoring condition, and when the current environmental information changes, the preset scoring condition needs to be re-determined.

[0073] It is understandable that when the scoring result of the audio signal does not meet the preset scoring conditions corresponding to the current environmental information, it means that the current audio signal does not meet the scoring criteria corresponding to the current environmental information. At this time, the sound environment in the car may affect the driving experience of the occupants. Therefore, the audio signal needs to be adjusted to meet the preset scoring conditions corresponding to the current environmental information.

[0074] Among them, adjusting the audio signal refers to the process of obtaining ideal sound quality by adjusting various parameters of the audio signal, which may include volume adjustment, timbre adjustment, equalizer setting, and reverberation effect adjustment. For example, volume adjustment can directly affect the audience's perception of sound, and the loudness of the sound can be changed by increasing or decreasing the amplitude of the audio signal to achieve volume adjustment of the audio signal.

[0075] In some embodiments, adjusting the audio signal at least once may include adjusting the volume of different frequency bands in the audio signal at least once. The frequency band refers to a sound or signal within a specific frequency range in the audio signal. Exemplarily, the volume of the frequency band between 60 Hz (Hz) and 250 Hz is reduced to reduce the intensity of the sound, thereby making the sound quality more balanced.

[0076] In some embodiments, adjusting the audio signal at least once may include determining a parameter of the audio signal that needs to be adjusted, and adjusting the parameter of the audio signal at least once. The parameter of the audio signal is used to describe and analyze the characteristics of the audio signal, and may include frequency, amplitude, coincidence, and sampling rate. Exemplarily, when the scoring result does not meet the first preset scoring condition, it is determined that the loudness parameter of the audio signal needs to be adjusted, and the loudness parameter of the audio signal is adjusted at least once.

[0077] In some embodiments, adjusting the audio signal at least once may also include adjusting the audio signal at least once using active noise reduction technology. Among them, active noise reduction technology is a technology that achieves noise reduction by generating a sound wave opposite to the noise signal and superimposing it with the noise signal to cancel each other out. Exemplarily, after the audio signal is propagated, an additional secondary sound source signal is generated in the environment. The secondary sound source signal has an opposite phase to the audio signal. The opposite phase at a certain position will cause the signal at the same position on the audio signal to attenuate. Therefore, by outputting a parameter adjustment instruction for adjusting the secondary sound source signal and sending the secondary sound source signal through the speaker, the audio signal is adjusted in an active noise reduction manner.

[0078] In some embodiments, adjusting the audio signal at least once may further include adjusting the audio signal at least once using a filter, wherein a filter is an electronic device or system that allows a specific frequency signal in the audio signal to pass through while blocking or attenuating other frequency signals in the audio signal to achieve noise reduction technology.

[0079] In the embodiment of the present application, the audio signals of the areas where different people are located in the vehicle, as well as the preference information of the people in the vehicle for sound quality are obtained, and the sound quality of the obtained audio signals is predicted using the trained scoring prediction model to obtain the scoring result of the audio signal; the first preset scoring condition corresponding to the current environmental information of the vehicle is determined; if the scoring result does not meet the first preset scoring condition, the audio signal is adjusted at least once so that the scoring result of the adjusted audio signal meets the preset scoring condition corresponding to the current environmental information. By using the trained scoring prediction model to predict the audio signal in real time, the sound quality of the audio signal can be analyzed and evaluated in real time, thereby satisfying the personalized preferences of the driver and passengers in a dynamic environment, and making adjustments according to different environmental information, so as to achieve refined and intelligent adjustment of the audio in the car.

[0080] In some embodiments, Figure 2 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 2 ,like Figure 2 As shown, before step S101 "obtaining audio signals of different areas where people are located in the vehicle", steps S201 and S202 may be included. Figure 2 The steps shown are explained.

[0081] Step S201, obtaining the preference information of the persons in the vehicle for the sound quality.

[0082] In the embodiment of the present application, the preference information represents the preference of the occupants of the vehicle for the sound quality in different dimensions.

[0083] In some embodiments, obtaining the preference information of the vehicle occupants on the sound quality may include obtaining the preference information of the vehicle occupants on the sound quality through social interaction with the vehicle occupants. For example, multiple passengers in the vehicle are allowed to participate in joint selection or voting to determine the type of music to be played or the sound effect setting, and more accurate preference information is obtained through such interaction.

[0084] In some embodiments, obtaining the preference information of the occupants of the vehicle for the sound quality may also include performing multi-modal interaction with the occupants of the vehicle to obtain the preference information of the occupants of the vehicle for the sound quality. Multi-modal interaction refers to the process of communication between a user and a computer system through multiple interaction methods, which may include but are not limited to voice, gesture, touch, text, image, etc. Exemplarily, passengers are allowed to directly adjust the audio settings through voice commands or by touching the vehicle display screen, such as adjusting the bass, treble, and sound effect mode, and their preference information is collected through feedback.

[0085] Step S202: determining a trained score prediction model based on the preference information of the occupants of the vehicle regarding the sound quality.

[0086] It is understandable that each score prediction model is trained for different preference information of people in the vehicle for sound quality, and different score prediction models correspond to different model parameters. Then, when scoring audio signals for different preference information, it is necessary to first determine the trained score prediction model, and then determine the score of the audio information on the dimension corresponding to the preference information based on the corresponding trained score prediction model. Exemplarily, when the preference information of people in the vehicle for sound quality is comfort, the score prediction model is determined to be a comfort type score prediction model.

[0087] In some embodiments, based on the preference information of the vehicle personnel for the sound quality, determining the trained score prediction model may include: collecting the scores of the vehicle personnel for different sound qualities, deriving the preference information for the sound quality, and determining the trained score prediction model. Exemplarily, the vehicle personnel score different sound qualities, the vehicle system collects the scores and determines the closest type of preference as the preference information through data visualization and statistical analysis.

[0088] In the embodiment of the present application, by obtaining the preference information of the people in the vehicle for sound quality and determining a scoring prediction model corresponding to the preference information, a user-centered design can be implemented to make the audio adjustment more in line with the individual needs of the passengers, thereby satisfying the personalized preferences of the driver and passengers.

[0089] In some embodiments, Figure 3 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 3 ,like Figure 3 As shown, the "determining the first preset scoring condition corresponding to the current environmental information of the vehicle" in step S103 may include steps S301 to S303. Figure 3 The steps shown are explained.

[0090] Step S301, obtaining road surface image information of the current road on which the vehicle is traveling.

[0091] Among them, road surface image information refers to visual data related to the road surface captured by cameras or other imaging devices. This information is used to analyze and evaluate the condition, performance and safety of the road surface.

[0092] In some embodiments, obtaining the road surface image information of the vehicle on the current road includes obtaining the road surface state information, road surface type and road surface markings of the vehicle on the current road. For example, the road surface state signal may be defects such as the degree of wear of the road surface and cracks on the road surface. The road surface markings may be the clarity and integrity of lane lines, stop lines, pedestrian crossing lines, etc.

[0093] In some embodiments, obtaining the road surface image information of the vehicle on the road currently being driven may also include using a vehicle-mounted camera to capture the road surface image in real time. For example, a vehicle-mounted wide-angle camera may be used to obtain a larger field of view, and infrared or night vision technology may be used to monitor the road surface at night or in low light conditions.

[0094] In some embodiments, obtaining the road image information of the vehicle on the road currently being driven may also include obtaining the road image information of the vehicle on the road currently being driven based on the current positioning information. For example, multiple users mark and report the road image information of the current positioning in the smartphone application, and the road image information can be shared by the vehicle system and can be used as the road image information of the vehicle on the road currently being driven.

[0095] Step S302: Identify the road type of the current driving road based on the road surface image information.

[0096] In some embodiments, the method for classifying the road type of the current driving road may include classifying the road type into eight road types from A to H according to the roughness of the road. For example, when the road is extremely flat, the road type is determined to be A, and when the road is extremely uneven, the road type is determined to be H.

[0097] In some embodiments, the method for dividing the road type of the current driving road may also include dividing the road type into multiple road types according to the road surface material and road surface characteristics. Exemplarily, according to the road surface material and characteristics, the road surface type is first divided into asphalt road, cement road, and gravel road. Each road surface is then combined with the road surface state (dry or slippery) at the time to be divided into dry road surface and slippery road surface. For example, asphalt road can be subdivided into dry asphalt road and slippery asphalt road, three types of road surfaces, each road surface has two road surface states, and a total of 6 road surface types are planned according to the road surface material and road surface state.

[0098] In some embodiments, the method for classifying the road type of the current driving road may also include classifying the road into multiple road types according to road surface materials, construction technology, and road surface characteristics.

[0099] In some embodiments, based on the road image information, identifying the road type of the current driving road may include using at least one artificial intelligence model to analyze the road image information to identify the road type of the current driving road. Exemplarily, the camera in the vehicle inputs the collected image information into a hybrid model CNN-LSTM model composed of a convolutional neural network CNN and a long short-term memory network LSTM, wherein the convolutional neural network is used to extract the spatial features of the road image information, that is, the geometric and distribution characteristics of an object in the road image information, and the long short-term memory network is used to capture the time dependency of the road image information, that is, the characteristics of a certain feature in the road image information that changes over time. Through the CNN-LSTM hybrid model, the current road type of the vehicle can be efficiently identified.

[0100] Step S303: determining a first preset scoring condition based on the road type of the current driving road.

[0101] It is understandable that different road types correspond to different first preset scoring conditions. When a vehicle is traveling on different types of roads, even if the scoring results obtained by predicting the audio signal based on the trained scoring prediction model are the same, the first preset score corresponding to the current road type is not satisfied on every type of road. For example, for a pothole road, since potholes inevitably generate noise, even if noise reduction is achieved to the greatest extent, the impact of noise on sound quality cannot be completely eliminated. Therefore, when the evaluation result ranges from 0 to 10, the first preset evaluation condition for potholes can be set to be greater than or equal to 5; similarly, since smooth ground generates less noise, the first preset evaluation condition for smooth ground can be set to be greater than or equal to 8.

[0102] In some embodiments, based on the road type of the current driving road, determining the first preset scoring condition may include querying a preset mapping rule based on the road type of the current driving road to determine the first preset scoring condition. Exemplarily, the mapping rule defines a mapping rule between each type of road and the first preset scoring condition. When the road type is A, the corresponding first preset scoring condition is that the scoring result is greater than or equal to 8; when the road type is H, the corresponding first preset scoring condition is that the scoring result is greater than or equal to 5.

[0103] In the embodiment of the present application, since different road types have different effects on passengers' preferences and audio signals, by adjusting the corresponding scoring conditions, refined and intelligent adjustment of the in-car audio can be achieved to meet the personalized preferences of the driver and passengers and optimize the riding experience of the driver and passengers.

[0104] In some embodiments, "determining a first preset scoring condition based on the road type of the current driving road" in step S303 may include:

[0105] Step S3031, determining a first preset scoring condition based on the driving state information of the vehicle and the road type of the current driving road.

[0106] It is understandable that different road conditions and vehicle driving conditions will have a direct impact on driving safety. By evaluating this information in real time, the driving strategy can be adjusted in time to reduce the risk of accidents. For example, on a slippery road, an increase in vehicle speed or acceleration will cause the loudness of the audio signal in the vehicle to increase, resulting in a change in the first preset scoring condition, so the first preset condition needs to be determined.

[0107] In some embodiments, the vehicle's driving status information may include acceleration, deceleration, increased acceleration, decreased acceleration, and stationary.

[0108] For example, the first preset evaluation condition for a vehicle traveling at a constant speed on a bumpy road can be set to be greater than or equal to 5. However, when the vehicle decelerates on the bumpy road, the friction between the tires and the ground is reduced, resulting in a corresponding reduction in noise. At this time, the first preset evaluation condition may change and can be set to be greater than or equal to 6.

[0109] In the embodiment of the present application, by comprehensively considering the vehicle driving status and road type, the system can formulate scoring conditions that are more in line with the actual driving environment, thereby optimizing the riding experience of the driver and passengers.

[0110] In some embodiments, the first preset scoring condition includes a plurality of scoring intervals and a scoring mark corresponding to each scoring interval. Figure 4 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 4 ,like Figure 4 As shown, in step S104, "if the scoring result does not meet the first preset scoring condition, adjusting the audio signal at least once" may include step S401 and step S402. Figure 4 The steps shown are explained.

[0111] Step S401, determining a target scoring interval of a scoring result in a plurality of scoring intervals.

[0112] It is understandable that dividing the first preset scoring condition into multiple intervals can provide consistency for different evaluation criteria, and different interval classifications can make statistical work more convenient. Exemplarily, the first preset scoring condition may include [0,6), [6,8) and [8,10). When the evaluation result is in the interval [0,6), it means that the scoring mark corresponding to the current evaluation result is "unqualified", when the evaluation result is in the interval [6,8), it means that the scoring mark corresponding to the current evaluation result is "qualified", and when the evaluation result is in the interval [8,10], it means that the scoring mark corresponding to the current evaluation result is "excellent".

[0113] Step S402: when the scoring mark corresponding to the target scoring interval is a preset scoring mark, it is determined that the scoring result does not meet the first preset scoring condition, and the audio signal is adjusted at least once.

[0114] The preset scoring mark is a scoring mark corresponding to the interval that does not meet the first preset scoring condition. For example, the preset scoring mark is "failed", and when the scoring mark of the evaluation result in the corresponding target scoring interval is "failed", it is determined that the scoring result does not meet the first preset scoring condition.

[0115] In the embodiment of the present application, the system monitors the scoring results in real time and makes adjustments to ensure that the audio signal always remains in the best state in a dynamic driving environment, thereby optimizing the riding experience of the driver and passengers.

[0116] In some embodiments, Figure 5 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 5 ,like Figure 5 As shown, the step S402 of "adjusting the audio signal at least once" may include steps S501 to S504. Figure 5 The steps shown are explained.

[0117] Step S501: determine a first audio suppression signal of an audio signal.

[0118] It is understandable that background noise will affect the quality of the audio signal. In order to solve this problem, the present application adopts active noise reduction technology, which is a technology that generates sound waves opposite to the noise signal and superimposes them with the ambient noise in the vehicle to cancel each other out and achieve noise reduction. Compared with other noise reduction technologies, it has a wider range of applications and a significant and intelligent noise reduction effect. Therefore, before achieving noise reduction of the audio signal, it is necessary to generate a suppression signal for the audio signal collected in the vehicle to reduce the noise of the audio signal.

[0119] The first suppression signal refers to a suppression signal generated to cancel a specific noise source (audio signal).

[0120] In some embodiments, determining the first audio suppression signal of the audio signal may include performing an inversion process on the audio signal to generate a first audio suppression signal having a phase opposite to that of the original noise signal. Inversion process is a signal processing technology mainly used to reverse the phase of the sound signal by 180 degrees to suppress the sound signal.

[0121] Step S502: adjust the audio signal based on the first audio suppression signal to obtain a first audio signal.

[0122] In some embodiments, adjusting the audio signal based on the first audio suppression signal to obtain the first audio signal may include performing noise suppression on the audio signal based on the first audio suppression signal to obtain the first audio signal, wherein the noise suppression is performed by identifying and reducing the impact of background noise on the audio signal.

[0123] Exemplarily, a noise suppression algorithm (such as spectral subtraction and Wiener filtering) is used to improve the clarity of the audio signal to obtain a first audio signal with a more prominent audio signal. Spectral subtraction is a frequency domain noise suppression technology that reduces noise by processing the spectrum of the signal; Wiener filtering is a linear filtering technology based on a statistical model that suppresses noise by modeling the statistical characteristics of the signal.

[0124] In some embodiments, adjusting the audio signal based on the first audio suppression signal to obtain the first audio signal may include performing dynamic range control on the audio signal based on the first audio suppression signal to obtain the first audio signal. The dynamic range control is to compress the dynamic range of the audio signal so that the larger volume (such as the peak part) in the signal is suppressed while improving the hearing of the lower volume part. Exemplarily, a compressor for implementing dynamic range control can set a threshold value. When the audio signal exceeds this threshold value, the compressor will reduce the gain of the audio, thereby making the sound smoother and avoiding distortion.

[0125] Step S503: when the current environment information of the vehicle changes, determine a second preset scoring condition corresponding to the changed current environment information of the vehicle.

[0126] It is understandable that the audio adjustment method of the present application is performed for real-time monitoring of the vehicle. When the current environmental information of the vehicle changes, the first preset scoring condition corresponding to the environmental information will change. At this time, the re-determined first preset scoring condition will be regarded as the second preset scoring condition, so as to re-determine the in-vehicle audio adjustment method adapted to the current environmental information.

[0127] Step S504: When the first audio signal does not satisfy the second preset scoring condition, the first audio signal is adjusted based on a second audio suppression signal corresponding to the first audio signal to obtain a second audio signal.

[0128] In the embodiment of the present application, the scoring conditions can be dynamically adjusted according to the changes in the vehicle's current environmental information, making the system more intelligent and adaptable in different driving scenarios, thereby providing a better user experience for the driver and passengers.

[0129] In some embodiments, Figure 6 A schematic diagram of the implementation process of an in-car audio adjustment method provided in an embodiment of the present application Figure 6 ,like Figure 6 As shown, the training process of the trained score prediction model in the in-vehicle audio adjustment method may include steps S601 to S603. Figure 6 The steps shown are explained.

[0130] Step S601 , collecting multiple sample audio information in the vehicle under different scenes and / or different vehicle speeds.

[0131] Among them, the sample audio information is the audio information used to train the score prediction model. By conducting in-depth analysis and inference on these sample audio information, a model can be constructed that can make reasonable and reliable predictions and inferences on the overall audio information.

[0132] It is understandable that the training of the scoring preset model needs to rely on a large amount of sample audio information to ensure that the model can accurately learn the relationship between input and output. The diversity of samples is crucial to the performance of the model, because only through a wide range of samples can the model capture the underlying features and regularities, thereby providing reliable scoring results in practical applications.

[0133] In some embodiments, the multiple sample audio information collected in the vehicle under different scenes and / or different vehicle speeds may include natural sound effects, environmental sound effects, waveforms and spectra of audio signals, etc.

[0134] Step S602: for each sample audio information, obtain multiple scoring label information of the sample audio information under multiple sound quality scoring standards.

[0135] In the embodiment of the present application, a plurality of sound quality scoring standards correspond one-to-one to a plurality of preference information of persons in the vehicle regarding sound quality.

[0136] It is understandable that the same sample audio information may give different results in the score prediction models of different dimensions. For example, when the preference information of the person in the vehicle for the sound quality is comfort, the score label information of the sample audio information under the corresponding sound quality scoring standard of the comfort type may be comfort 9 points. When the preference information of the person in the vehicle for the sound quality is dynamic, the score label information of the sample audio information under the corresponding sound quality scoring standard of the dynamic type may be dynamic 8 points.

[0137] Step S603: Based on each sample audio information and multiple score label information corresponding to each sample audio information, an initial score prediction model is trained to obtain multiple trained score prediction models.

[0138] It is understandable that during the training process, each sample audio information and the corresponding multiple scoring label information with different dimensions (such as comfort, power and sportiness) are imported into the preset model. After processing these sound data, the model will generate corresponding scores as output results. Subsequently, these automatically generated scores are carefully compared with the scores obtained by professionals for subjective evaluation of the same training set. This comparison link aims to reveal the difference between model prediction and actual perception. Based on this comparative analysis, by manually adjusting the parameter configuration of the model and continuously iterating and optimizing, we strive to make the scores output by the model gradually approach the subjective scores of professionals. This process is repeated until the model output results are highly consistent with the manually input data. Exemplarily, the error between the model output results and the manually input data can be less than 2%.

[0139] In the embodiment of the present application, a plurality of trained rating prediction models correspond one-to-one to the preference information of the persons in a plurality of vehicles for the sound quality.

[0140] Exemplarily, based on each sample audio information, and the 9 points of comfort, 8 points of movement, and 5 points of power corresponding to each sample audio information, the initial score prediction model corresponding to comfort, the score prediction model corresponding to movement, and the score prediction model corresponding to power are trained respectively to obtain multiple trained score prediction models.

[0141] In the embodiment of the present application, multiple sound quality scoring standards are adopted so that the model can comprehensively consider different acoustic characteristics, thereby obtaining a more comprehensive evaluation result of the audio signal. In addition, by continuously updating and training the scoring prediction model, it is possible to provide the driver and passengers with changes in preferences and new environments, meet the personalized preferences of the driver and passengers, and optimize the riding experience of the driver and passengers.

[0142] In some embodiments, Figure 7 A schematic diagram of the architecture of an in-car audio adjustment method provided in an embodiment of the present application, such as Figure 7 As shown, the architecture may include a data acquisition and preprocessing module 701, a road surface recognition module 702, a sound quality preference acquisition module 703, a multi-dimensional active acoustic design and control module 704, and a feedback and optimization module 705. Figure 7 The architecture shown is described.

[0143] The data acquisition and preprocessing module 701 is responsible for collecting road data, capturing the road image information in front of the vehicle during the vehicle's driving process through the vehicle-mounted camera, laser radar, etc., and transmitting the image data to the road recognition module 702 of the vehicle-mounted computer for processing.

[0144] The road surface recognition module 702 recognizes the road surface type (road type) through a preset algorithm and transmits the recognition result to the control system.

[0145] In some embodiments, the road surface recognition model used in the road surface recognition module 702 can be a hybrid model composed of multiple intelligent models. Exemplarily, the road surface recognition model can be a CNN-LSTM hybrid model based on a convolutional neural network (CNN) and a long short-term memory network (LSTM) for joint calculation. Among them, the convolutional neural network is a deep learning model used to process data with a grid structure, such as images, and is used to extract the spatial features of road surface data; the long short-term memory network is a powerful sequence modeling tool that can effectively capture long-term dependencies and is applied to various sequence data processing tasks to capture the time dependency of road surface data.

[0146] In some embodiments, in order to improve the modeling accuracy, before the road surface recognition module 702 recognizes the road surface type, the data acquisition and preprocessing module 701 needs to preprocess the data. The preprocessing method mainly includes the following steps 1 to 3:

[0147] Step 1: Data cleaning.

[0148] Among them, data cleaning refers to the processing of raw data to improve the quality and consistency of the data.

[0149] In some embodiments, data cleaning may include deduplication and text processing. Exemplarily, deduplication may include deleting duplicate records in the original data; text processing may include removing punctuation marks in the original data, performing stemming or lemmatization.

[0150] Step 2: Data enhancement.

[0151] Among them, data enhancement is to diversify the data during the training process to increase the generalization ability of the model.

[0152] In some embodiments, data enhancement may include image data enhancement and audio data enhancement. For example, image data enhancement includes operations such as rotating, flipping, and scaling the image; audio data enhancement includes adding noise, adding background noise to the audio, and time stretching, stretching or compressing the audio.

[0153] Step 3: Data normalization.

[0154] Among them, data normalization is the process of scaling data to a specific range (such as [0, 1] or [-1, 1]), which helps improve the training efficiency and convergence speed of the model.

[0155] In some embodiments, data normalization includes minimum-maximum normalization, which is a method of linearly transforming data to a specified range to eliminate differences between features so that the influence of each feature on the model is relatively balanced.

[0156] After completing data preprocessing, the data acquisition and preprocessing module 701 inputs the filtered data into the hybrid model, trains the data based on the model, and evaluates the prediction results until the prediction results meet the set accuracy requirements.

[0157] The sound quality preference acquisition module 703 collects the occupant's preference information on sound quality through the vehicle-mounted human-computer interaction system. The occupant can set the degree of preference in dimensions such as comfort, power, and sportiness, and the system uses this information as the basis for subsequent acoustic design. At the same time, a sound quality preference database is established to store the corresponding sound quality preference settings under different road types.

[0158] The multi-dimensional active acoustic design and control module 704 processes the noise signal in the human ear area in the vehicle, compares it with the preset target sound quality signal, and when the noise signal in the human ear area in the vehicle does not meet the preset target sound quality signal, according to the road type identified by the road surface recognition module 702 and the sound quality preference information of the driver and passenger obtained by the sound quality preference acquisition module 703, combined with the current speed, acceleration and other driving state information of the vehicle, through the multi-dimensional active acoustic design and control module 704, by adjusting the influencing parameters of the in-vehicle sound quality, according to the parameter adjustment instruction output by the control system, the secondary sound source signal is adjusted in real time and sent out through the speaker, so as to realize active noise reduction or active acoustic compensation of multi-dimensional sound quality, adjust the in-vehicle acoustic environment in real time, and provide the occupants with an in-vehicle acoustic environment that adapts to the road conditions and meets their personal preferences.

[0159] In some embodiments, the influencing parameters may include A-weighted total sound pressure level, order sound pressure level, loudness, roughness, sharpness, fluctuation, speech clarity, etc.

[0160] The feedback and optimization module 705 dynamically adjusts and optimizes the active noise reduction of multi-dimensional sound quality through real-time feedback from passengers and continuous monitoring of road types, so as to meet the ever-changing needs of passengers and adapt to different road conditions, and continuously improve the sound quality experience of passengers.

[0161] In some embodiments, Figure 8 A schematic diagram of a hybrid model structure of an in-car audio adjustment method provided in an embodiment of the present application, such as Figure 8 As shown, the hybrid model structure may include steps S801 to S805. Figure 8 The architecture shown is described.

[0162] Step S801, input sample data.

[0163] In some embodiments, the sample data is audio signals of areas where different persons are located in the vehicle in the present application.

[0164] Step S802: data preprocessing.

[0165] In some embodiments, the process of data preprocessing includes data cleaning, data enhancement, and data normalization. Among them, data cleaning is used to remove invalid data, as well as unclear or missing data. Data enhancement is used to expand the sample size, overcome the overfitting problem caused by insufficient sample size, and thus improve the accuracy of the model. The data set can be enhanced according to actual needs. For numerical data, more samples are obtained by interpolating two samples and target values ​​in the data set. For image data, cropping, flipping, light brightness changes, etc. can be performed to obtain more sample sizes. Data normalization is used to uniformly scale the data to avoid the impact of data of different scales on the model.

[0166] Step S803: feature extraction.

[0167] In some embodiments, after data preprocessing is completed, the screened data is input into the CNN-LSTM hybrid model, the data is trained based on the model, and the prediction results are evaluated until the prediction results meet the set accuracy requirements.

[0168] Among them, the CNN-LSTM hybrid model includes CNN layer and LSTM layer. The CNN layer includes convolution layer and pooling layer. The convolution layer is used to extract image data features, and the pooling layer is used to filter the feature map to reduce network parameters. The feature map output by the CNN layer is flattened to obtain the feature vector and input into the LSTM layer.

[0169] Step S804, classification.

[0170] In some embodiments, the feature vector extracted in step S803 is input into the fully connected layer and the Softmax layer of the LSTM to classify the road surface.

[0171] Step S805: output the classification result.

[0172] The technical solution of the present application can combine the advantages of convolutional neural networks in image feature extraction and the ability of long short-term memory networks in processing time series data to identify road types in real time, and actively design the in-vehicle sound environment based on road conditions and the sound quality preferences of drivers and passengers, so as to provide the best driving experience for different customer groups.

[0173] This application proposes a method and system for active acoustic design of multi-dimensional sound quality in a car based on convolutional neural network and long short-term memory network to identify road surface. This method actively designs the acoustic environment in the car by identifying the road surface type in real time and according to the road surface conditions and the sound quality preferences of the occupants, so as to provide the best driving experience for different customer groups. The present invention includes a matching system, including a data acquisition and preprocessing module, a CNN-LSTM road surface recognition module, a sound quality preference acquisition module, a multi-dimensional active acoustic design module, and a feedback and optimization module, and each module works together to implement the above method.

[0174] The advantages of the in-car audio adjustment method listed in the embodiment of the present application include: 1) Real-time road surface recognition: Through the deep learning model, real-time and accurate recognition of road surface types is achieved, which improves the pertinence and effectiveness of the in-car sound quality design. 2) Multi-dimensional sound quality design: Comprehensively consider multi-dimensional sound quality characteristics such as comfort, power, and sportiness to provide customers with a richer and more delicate driving experience. 3) Personalized experience: According to the sound quality preferences and real-time feedback of the occupants, personalized in-car sound quality design is achieved to meet the needs of different occupants. 4) Real-time feedback and optimization: The present invention can adjust the sound quality design strategy in real time according to road surface changes and occupant feedback to achieve continuous optimization of the riding experience. Through the built-in feedback mechanism, the system of the present invention can continuously learn and optimize, and its control accuracy and user experience will continue to improve with the increase of usage time. 5) Wide range of applications: The present invention is applicable to various types of cars and road conditions, and has broad application prospects.

[0175] At this point, the refined and intelligent adjustment of in-car audio and the in-car audio adjustment method that meets the personalized preferences of drivers and passengers have been completed.

[0176] Based on the same inventive concept, the embodiment of the present application further provides an in-vehicle audio adjustment device, which operates in a vehicle computer. Fig. 9A schematic diagram of the structure of an in-car audio adjustment device is provided for an embodiment of the present application. Fig. 9 As shown, the in-car audio adjustment device 900 includes an acquisition module 901, a prediction module 902, a determination module 903 and an adjustment module 904, wherein:

[0177] An acquisition module 901 is used to acquire audio signals of different areas where people are located in the vehicle;

[0178] A prediction module 902 is used to use a trained score prediction model to predict the sound quality of the audio signal to obtain a score result of the audio signal; the trained score prediction model is determined based on the preference information of the person in the vehicle for sound quality;

[0179] A determination module 903, configured to determine a first preset scoring condition corresponding to the current environmental information of the vehicle;

[0180] The adjustment module 904 is used to adjust the audio signal at least once when the scoring result does not meet the first preset scoring condition; the scoring result of the audio signal after adjustment meets the preset scoring condition corresponding to the current environment information.

[0181] In some embodiments, the in-vehicle audio adjustment device 900 also includes: a first acquisition unit, used to obtain the preference information of the people in the vehicle for sound quality; the preference information represents the degree of preference of the people in the vehicle for sound quality in different dimensions; a first determination unit, used to determine the trained score prediction model based on the preference information of the people in the vehicle for sound quality.

[0182] In some embodiments, the determination unit 903 includes: a second acquisition unit, used to acquire road surface image information of the vehicle on the current road; an identification unit, used to identify the road type of the current road based on the road surface image information; and a second determination unit, used to determine the first preset scoring condition based on the road type of the current road.

[0183] In some embodiments, the second determination unit includes: a third determination unit, configured to determine the first preset scoring condition based on the driving state information of the vehicle and the road type of the current driving road.

[0184] In some embodiments, the first preset scoring condition includes multiple scoring intervals and a scoring mark corresponding to each of the scoring intervals; the adjustment module 904 includes: a fourth determination unit, used to determine a target scoring interval of the scoring result in the multiple scoring intervals; a fifth determination unit, used to determine that the scoring result does not meet the first preset scoring condition when the scoring mark corresponding to the target scoring interval is a preset scoring mark, and adjust the audio signal at least once.

[0185] In some embodiments, the fifth determination unit includes: a sixth determination unit, used to determine a first audio suppression signal of the audio signal; a first adjustment unit, used to adjust the audio signal based on the first audio suppression signal to obtain a first audio signal; a seventh determination unit, used to determine a second preset scoring condition corresponding to the changed current environmental information of the vehicle when the current environmental information of the vehicle changes; and a second adjustment unit, used to adjust the first audio signal based on a second audio suppression signal corresponding to the first audio signal to obtain a second audio signal when the first audio signal does not meet the second preset scoring condition.

[0186] In some embodiments, the in-vehicle audio adjustment device 900 also includes: a collection unit, which is used to collect multiple sample audio information in the vehicle under different scenarios and / or different vehicle speeds; a third acquisition unit, which is used to obtain multiple score label information of the sample audio information under multiple sound quality scoring standards for each of the sample audio information; the multiple sound quality scoring standards correspond one-to-one to the preference information of multiple people in the vehicle for sound quality; a training unit, which is used to train an initial score prediction model based on each of the sample audio information and the multiple score label information corresponding to each of the sample audio information, to obtain multiple trained score prediction models; the multiple trained score prediction models correspond one-to-one to the preference information of multiple people in the vehicle for sound quality.

[0187] The above description of the system embodiment for in-car audio adjustment is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment. In some embodiments, the functions or modules included in the system provided by the embodiment of the present disclosure can be used to execute the method described in the above method embodiment. For technical details not disclosed in the system embodiment of the present application, please refer to the description of the method embodiment of the present application for understanding.

[0188] It should be noted that in the embodiment of the present application, if the above-mentioned resource configuration method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, which is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a disk or an optical disk. In this way, the embodiment of the present application is not limited to any specific hardware, software or firmware, or any combination of hardware, software, and firmware.

[0189] Fig.10 The hardware entity diagram of an in-car audio adjustment device provided in an embodiment of the present application is shown in FIG. Fig.10 As shown, the in-car audio adjustment device 1000 can be the electronic device in the above-mentioned embodiment. The in-car audio adjustment device 1000 may include a processor 1001, which can perform various appropriate steps and processes according to computer program instructions stored in a read-only memory 1002 or computer program instructions loaded from a memory 1008 to a random access memory (RAM) 1003. In RAM 1003, various programs and data required for the operation of the in-car audio adjustment device 1000 can also be stored. The processor 1001, ROM 1002 and RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0190] Multiple components in the in-car audio adjustment device 1000 are connected to the I / O interface 1005, including: an input unit 1006, such as a keyboard, a mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a memory 1008, such as a disk, an optical disk, etc.; and a communication unit 1009, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1009 allows the in-car audio adjustment device 1000 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0191] The processor 1001 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 1001 include, but are not limited to, a central processing unit, a graphics processor, various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSP), and any appropriate processors, controllers, microcontrollers, etc. The processor 1001 may execute the various methods and processes described above, such as executing the resource configuration method described above. For example, in some embodiments, the resource configuration method described above may be implemented as a computer software program, which is stored in a machine-readable medium, such as a memory 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on the in-vehicle audio adjustment device 1000 via ROM 1002 and / or the communication unit 1009. When the computer program is loaded into RAM 1003 and executed by the processor 1001, one or more steps in the resource configuration method described above may be executed. Alternatively, in other implementations, the processor 1001 may be configured to execute one or more steps in the resource configuration method described above in any other appropriate manner (for example, by means of firmware).

[0192] It should be further noted that the present application may include methods, systems and / or computer program products. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for executing various aspects of the present application.

[0193] The embodiment of the present application provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, some or all of the steps in the above method can be implemented. The computer-readable storage medium can be transient or non-transient.

[0194] An embodiment of the present application provides a computer program, including computer-readable program instructions. When the computer-readable program instructions are executed in a computer device, a processor in the computer device executes some or all of the steps in the above method.

[0195] The present application embodiment provides a computer program product, including a computer program or an instruction, which, when executed by a processor, can implement some or all of the steps in the above method. The computer program product can be implemented specifically by hardware, software, or a combination thereof. In some embodiments, the computer program product is specifically embodied as a computer storage medium, and in other embodiments, the computer program product is specifically embodied as a software product, such as a software development kit (SDK).

[0196] Computer readable storage medium can be a tangible device that can hold and store instructions used by an instruction execution device. Computer readable storage medium can be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination thereof. More specific examples (non-exhaustive list) of computer readable storage medium include: a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM), a flash memory, a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a protruding structure in a groove on which instructions are stored, and any suitable combination thereof. The computer-readable storage medium used in this document is not to be interpreted as a transient signal itself, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagated through a waveguide or other transmission medium (for example, light pulses through a fiber optic cable), or an electrical signal transmitted through wires.

[0197] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or to a processing device, or to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0198] The computer program instructions for performing the operation of the present application may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as C language or similar programming languages. The computer-readable program instructions may be executed entirely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, by utilizing the state information of computer-readable program instructions to personalize an electronic circuit, such as a programmable logic circuit, a field programmable gate array, or a programmable logic array (PLA), the electronic circuit can execute the computer-readable program instructions to implement various aspects of the present application.

[0199] Various aspects of the present application are described herein with reference to the flow chart and / or step diagram of the method, device (system), and computer program product according to the exemplary embodiments of the present application. It should be understood that each step of the flow chart and / or step diagram and the combination of each step in the flow chart and / or step diagram can be implemented by computer-readable program instructions.

[0200] These computer-readable program instructions can be provided to a processor in a voice interaction device, a general-purpose computer, a special-purpose computer, or a processing unit of other programmable data processing devices, thereby producing a machine, so that when these instructions are executed by a processing unit of a computer or other programmable data processing device, a device that implements the functions / steps specified in one or more steps in a flowchart and / or a step diagram is generated. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions enable a computer, a programmable data processing device, and / or other devices to work in a specific manner, so that a computer-readable medium storing instructions includes a manufactured product, which includes instructions for implementing various aspects of the functions / steps specified in one or more steps in a flowchart and / or a step diagram.

[0201] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device so that a series of operating steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to implement the functions / steps specified in one or more steps in the flowchart and / or step diagram.

[0202] The flow chart and step diagram in the accompanying drawings show the possible architecture, function and operation of the equipment, method and computer program product according to multiple embodiments of the present application. In this regard, each step in the flow chart or step diagram can represent a part of a module, program segment or instruction, and a part of the module, program segment or instruction includes one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the steps can also occur in a sequence different from that marked in the accompanying drawings. For example, two consecutive steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each step in the step diagram and / or flow chart, and the combination of the steps in the step diagram and / or flow chart, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0203] The above embodiments are only preferred embodiments for fully illustrating the present application, and the protection scope of the present application is not limited thereto. Any equivalent substitution or change made by a person skilled in the art based on the present application is within the protection scope of the present application.

Claims

1. A method for adjusting in-vehicle audio, characterized in that: The in-car audio adjustment method comprises: Obtain audio signals from different areas of the vehicle where different people are located; Using the trained score prediction model to predict the sound quality of the audio signal to obtain a score result of the audio signal; the trained score prediction model is determined based on the preference information of the person in the vehicle for the sound quality; Determining a first preset scoring condition corresponding to current environmental information of the vehicle; When the scoring result does not satisfy the first preset scoring condition, the audio signal is adjusted at least once; and the scoring result of the adjusted audio signal satisfies the preset scoring condition corresponding to the current environment information.

2. The in-vehicle audio adjustment method according to claim 1, characterized in that: The in-car audio adjustment method further includes: Acquiring preference information of the person in the vehicle for sound quality; the preference information represents the degree of preference of the person in the vehicle for sound quality in different dimensions; Based on the preference information of the persons in the vehicle regarding the sound quality, the trained score prediction model is determined.

3. The in-vehicle audio adjustment method according to claim 1, characterized in that: The determining of a first preset scoring condition corresponding to the current environmental information of the vehicle includes: Acquiring road surface image information of the vehicle on the road currently being driven; Based on the road surface image information, identifying the road type of the current driving road; The first preset scoring condition is determined based on the road type of the current driving road.

4. The in-vehicle audio adjustment method according to claim 3, characterized in that: The determining the first preset scoring condition based on the road type of the current driving road includes: The first preset scoring condition is determined based on the driving state information of the vehicle and the road type of the current driving road.

5. The in-vehicle audio adjustment method according to claim 1, characterized in that: The first preset scoring condition includes a plurality of scoring intervals and a scoring mark corresponding to each of the scoring intervals; When the scoring result does not satisfy the first preset scoring condition, adjusting the audio signal at least once includes: Determine a target scoring interval for the scoring result in the multiple scoring intervals; When the scoring mark corresponding to the target scoring interval is a preset scoring mark, it is determined that the scoring result does not meet the first preset scoring condition, and the audio signal is adjusted at least once.

6. The in-vehicle audio adjustment method according to claim 5, characterized in that: The step of adjusting the audio signal at least once comprises: determining a first audio suppression signal for the audio signal; Adjust the audio signal based on the first audio suppression signal to obtain a first audio signal; In the case where the current environmental information of the vehicle changes, determining a second preset scoring condition corresponding to the changed current environmental information of the vehicle; In a case where the first audio signal does not satisfy the second preset scoring condition, the first audio signal is adjusted based on a second audio suppression signal corresponding to the first audio signal to obtain a second audio signal.

7. The in-vehicle audio adjustment method according to any one of claims 1 to 6, characterized in that: The training process of the trained rating prediction model includes: Collect multiple sample audio information in the vehicle under different scenes and / or different vehicle speeds; For each of the sample audio information, multiple scoring label information of the sample audio information under multiple sound quality scoring standards are obtained; the multiple sound quality scoring standards correspond one to one with the preference information of the multiple persons in the vehicle for sound quality; Based on each of the sample audio information and multiple score label information corresponding to each of the sample audio information, an initial score prediction model is trained to obtain multiple trained score prediction models; the multiple trained score prediction models correspond one-to-one to the preference information of multiple people in the vehicle for sound quality.

8. An in-car audio adjustment device, characterized in that: include: An acquisition module, used to acquire audio signals of areas where different people are located in the vehicle; A prediction module, configured to use a trained score prediction model to predict the sound quality of the audio signal to obtain a score result of the audio signal; the trained score prediction model is determined based on the preference information of the occupants of the vehicle for the sound quality; A determination module, configured to determine a first preset scoring condition corresponding to current environmental information of the vehicle; An adjustment module is used to adjust the audio signal at least once when the scoring result does not meet the first preset scoring condition; the scoring result of the audio signal after adjustment meets the preset scoring condition corresponding to the current environment information.

9. An in-car audio adjustment device, characterized in that: include: A memory for storing executable instructions; A processor, configured to implement the method according to any one of claims 1 to 7 when executing the executable instructions stored in the memory.

10. A computer-readable storage medium, characterized in that: The storage medium stores executable instructions, and when the executable instructions are executed by the processor, the steps of the in-vehicle audio adjustment method described in any one of claims 1 to 7 are implemented.

Images