CN113504891B - Volume adjusting method, device, equipment and storage medium - Google Patents
Volume adjusting method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN113504891B CN113504891B CN202110805665.XA CN202110805665A CN113504891B CN 113504891 B CN113504891 B CN 113504891B CN 202110805665 A CN202110805665 A CN 202110805665A CN 113504891 B CN113504891 B CN 113504891B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- features
- feature extraction
- volume
- volume adjustment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 69
- 230000007613 environmental effect Effects 0.000 claims abstract description 10
- 230000001276 controlling effect Effects 0.000 claims abstract description 8
- 230000001105 regulatory effect Effects 0.000 claims abstract description 5
- 238000000605 extraction Methods 0.000 claims description 106
- 230000009471 action Effects 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 238000001228 spectrum Methods 0.000 claims description 7
- 230000010354 integration Effects 0.000 claims description 5
- 239000013598 vector Substances 0.000 description 15
- 230000008569 process Effects 0.000 description 11
- 230000006870 function Effects 0.000 description 8
- 238000011176 pooling Methods 0.000 description 6
- 238000001914 filtration Methods 0.000 description 5
- 238000009432 framing Methods 0.000 description 4
- 230000004913 activation Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000005236 sound signal Effects 0.000 description 3
- 230000003595 spectral effect Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 235000019800 disodium phosphate Nutrition 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000005484 gravity Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/16—Sound input; Sound output
- G06F3/165—Management of the audio stream, e.g. setting of volume, audio stream path
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Fittings On The Vehicle Exterior For Carrying Loads, And Devices For Holding Or Mounting Articles (AREA)
- Circuit For Audible Band Transducer (AREA)
Abstract
The embodiment of the application provides a volume adjusting method, a volume adjusting device, volume adjusting equipment and a storage medium, and relates to the field of vehicle-mounted environment recognition, wherein the volume adjusting method comprises the following steps: obtaining a plurality of types of vehicle features, the plurality of types of vehicle features comprising: vehicle environmental characteristics and vehicle control state characteristics; adopting a pre-trained prediction model to carry out prediction processing on the various vehicle characteristics to obtain a target volume adjustment prediction value; and regulating and controlling the playing volume of the vehicle-mounted audio equipment according to the target volume regulation predicted value. By the method, the safety of the vehicle interior personnel in operating the vehicle machine during driving is guaranteed, manual operation of the vehicle interior personnel on the vehicle machine such as volume control during driving is reduced, the burden of the vehicle interior personnel is reduced, the potential safety hazard of driving is reduced, and the concentration of the driver in driving the vehicle is improved.
Description
Technical Field
The invention relates to the field of vehicle-mounted environment identification, in particular to a volume adjusting method, a device, equipment and a storage medium.
Background
Along with the integrated function of vehicle-mounted system is more and more intelligent and entertaining, the user also is more and more through the scene of vehicle-mounted system broadcast music or other multi-media, not only can provide the amusement for the way that the driver was driving for a long time through vehicle-mounted system broadcast music or other multi-media, can also carry out shielding to a certain extent to the noise inside and outside the car, avoids the interference of noise, improves the degree of concentration that the driver drove the vehicle.
In the related art, the volume is often controlled by the person in the vehicle actively through a volume adjustment function in the vehicle.
However, in the related art, the volume of the vehicle system needs to be manually adjusted by the vehicle occupant actively controlling the volume through the volume adjusting function in the vehicle, and the manual control of the volume inevitably increases the burden on the vehicle occupant, which not only results in complex operation, but also reduces the concentration of the driver in driving the vehicle and increases the driving safety hazard.
Disclosure of Invention
The present invention aims to provide a volume adjustment method, apparatus, device and storage medium to reduce the volume control and other vehicle operations of the vehicle interior personnel during the driving process, reduce the operation burden of the vehicle interior personnel, improve the concentration of the driver in driving the vehicle, and reduce the driving safety hazard.
In order to achieve the above purpose, the technical solutions adopted in the embodiments of the present application are as follows:
in a first aspect, an embodiment of the present application provides a volume adjustment method, where the method includes:
obtaining a plurality of types of vehicle features, the plurality of types of vehicle features comprising: vehicle environmental characteristics and vehicle control state characteristics;
adopting a pre-trained prediction model to carry out prediction processing on the multiple types of vehicle characteristics to obtain a target volume adjustment prediction value;
and regulating and controlling the playing volume of the vehicle-mounted audio according to the target volume regulation predicted value.
Optionally, the acquiring various types of vehicle features includes:
carrying out feature extraction on the vehicle environment image to obtain vehicle image features;
carrying out feature extraction on the in-vehicle input voice to obtain in-vehicle voice features;
integrating the vehicle image features and the in-vehicle voice features to obtain the vehicle environment features;
and carrying out feature extraction on the vehicle control state to obtain the vehicle control state features.
Optionally, the performing feature extraction on the vehicle environment image to obtain the vehicle image feature includes:
and adopting a pre-trained image feature extraction model to extract the features of the vehicle environment image to obtain the vehicle image features.
Optionally, the performing feature extraction on the in-vehicle input speech to obtain the in-vehicle speech feature includes:
respectively extracting the features of the in-vehicle input voice by adopting a plurality of audio feature extraction modes to obtain voice features corresponding to the plurality of audio feature extraction modes;
and performing feature integration on the voice features corresponding to the multiple feature extraction modes to obtain the in-vehicle voice features.
Optionally, the multiple audio feature extraction manners include at least two audio feature extraction manners: a Mel frequency cepstrum coefficient feature extraction mode, a zero-crossing rate feature extraction mode and a spectrum centroid feature extraction mode.
Optionally, the performing feature extraction on the vehicle control state to obtain the vehicle control state feature includes:
carrying out feature extraction on the in-vehicle audio playing volume in a continuous value mode to obtain vehicle control volume features;
carrying out feature extraction on the skylight opening state in a discrete value mode to obtain the vehicle control skylight features;
carrying out feature extraction on the opening state of the vehicle window in a discrete value mode to obtain the features of the vehicle control vehicle window;
the vehicle control state characteristic comprises at least one of the following characteristics: the vehicle control volume characteristic, the vehicle control skylight characteristic and the vehicle control window characteristic.
Optionally, the prediction model includes: each characteristic classification tree corresponds to one type of vehicle characteristic, and each leaf node on each characteristic classification tree is a vehicle characteristic corresponding to a volume adjustment action;
the method for predicting the vehicle environmental characteristics and the vehicle control state characteristics by adopting the pre-trained prediction model to obtain the target volume adjustment prediction value comprises the following steps:
according to each type of vehicle feature, determining a leaf node where each type of vehicle feature is located as a target leaf node from a feature classification tree corresponding to each type of vehicle feature in the prediction model, and determining a volume adjustment parameter value of a volume adjustment action corresponding to the target leaf node as a volume adjustment predicted value corresponding to each type of vehicle feature;
and obtaining a target volume adjusting parameter value according to the volume adjusting predicted value corresponding to the various types of vehicle characteristics.
In a second aspect, an embodiment of the present application further provides a volume adjustment device, where the device includes:
an acquisition module to acquire a plurality of types of vehicle characteristics, the plurality of types of vehicle characteristics including: vehicle environmental characteristics and vehicle control state characteristics;
the prediction module is used for carrying out prediction processing on the various types of vehicle characteristics by adopting a pre-trained prediction model to obtain a target volume adjustment prediction value;
and the adjusting module is used for adjusting and controlling the playing volume of the vehicle-mounted audio according to the target volume adjusting predicted value.
In a third aspect, an embodiment of the present application further provides a computer device, including: a memory and a processor, wherein the memory stores a computer program executable by the processor, and the processor implements the volume adjusting method according to the first aspect when executing the computer program.
In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, the computer program performs the volume adjusting method according to the first aspect.
The beneficial effects of the invention are:
the embodiment of the application provides a volume adjusting method, a device, equipment and a storage medium. The method comprises the steps of adopting a pre-trained prediction model to carry out prediction processing on various vehicle characteristics to obtain a target volume adjustment predicted value, processing acquired necessary data through the pre-trained model, outputting the target volume adjustment predicted value, and regulating and controlling the playing volume of the vehicle-mounted audio according to the volume adjustment predicted value, so that the safety of controlling a vehicle machine by a person in the vehicle during driving is ensured, the manual operation of the vehicle machine such as volume control during driving of the person in the vehicle is reduced, the burden of the person in the vehicle is reduced, the potential safety hazard of driving is reduced, and the concentration degree of the driver in driving the vehicle is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a flowchart of a volume adjustment method according to an embodiment of the present disclosure;
fig. 2 is a flowchart of obtaining vehicle characteristics in a volume adjustment method according to an embodiment of the present disclosure;
fig. 3 is a flowchart illustrating a method for adjusting a volume according to an embodiment of the present disclosure to extract a vehicle voice feature;
fig. 4 is a flowchart illustrating a vehicle control state feature extraction method according to an embodiment of the present disclosure;
fig. 5 is a flowchart illustrating obtaining a target volume adjustment predicted value in a volume adjustment method according to an embodiment of the present disclosure;
fig. 6 is a schematic view of a volume adjustment device according to an embodiment of the present application;
fig. 7 is a block diagram of a computer device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments.
Fig. 1 is a flowchart of a volume adjustment method according to an embodiment of the present disclosure. The execution sequence of the steps of the volume adjustment method is not limited by the sequence disclosed in the present embodiment. The volume adjusting method can be implemented by a computer device, which can be, for example, a car machine device, a vehicle central control device, or other vehicle-mounted control device in a vehicle. As shown in fig. 1, the volume adjusting method includes:
s100, acquiring various types of vehicle characteristics, wherein the various types of vehicle characteristics comprise: vehicle environmental characteristics and vehicle control status characteristics.
For example, different feature extraction methods may be used to obtain different types of vehicle features, that is, the method used to obtain the environmental features of the vehicle may be different from the method used to obtain the vehicle control state features.
The vehicle environment characteristic can be an environment characteristic inside the vehicle and/or an environment characteristic outside the vehicle, and the vehicle control state characteristic can be a state characteristic of a control component inside the vehicle.
S200, adopting a pre-trained prediction model to perform prediction processing on various vehicle characteristics to obtain a target volume adjustment prediction value.
The prediction model may be a model obtained by training in advance using a plurality of types of vehicle sample features of a vehicle, where the plurality of types of vehicle sample features may include: the vehicle environment sample characteristics and the vehicle control state sample characteristics, wherein each type of vehicle sample characteristics are marked with volume adjustment sample values of the vehicle.
In the process of carrying out model training by adopting various types of vehicle sample characteristics, the internal association between various types of vehicle sample characteristics and the volume adjustment sample values can be learned by the prediction model through multiple times of training and learning.
After training is completed and a prediction model is obtained, the prediction model can be adopted to carry out prediction processing on various vehicle characteristics only by inputting various vehicle characteristics into the prediction model, and the result output by the prediction model is the target volume adjustment prediction value. The target volume adjustment predicted value may be a relative volume value, i.e., a volume deviation between the current playing volume and the target playing volume, or may be an absolute volume value, i.e., the target volume adjustment predicted value is the target playing volume.
S300, adjusting and controlling the playing volume of the vehicle-mounted audio equipment according to the target volume adjusting and predicting value.
In a possible implementation process, a volume regulation instruction is generated according to the target volume regulation predicted value, and the volume regulation instruction is sent to the vehicle-mounted audio equipment, so that the vehicle-mounted audio equipment regulates and controls the playing volume based on the volume regulation instruction. The volume control command carries a target volume control predicted value.
If the target volume adjustment predicted value is a relative volume value, the target volume adjustment predicted value ranges from a negative maximum volume value to a positive maximum volume value, if the target volume adjustment predicted value is negative, a volume adjustment instruction is generated through the target volume adjustment predicted value, the volume adjustment instruction is sent to the vehicle-mounted audio equipment to reduce the volume, if the target volume adjustment predicted value is positive, the volume adjustment instruction is generated through the target volume adjustment predicted value, the volume adjustment instruction is sent to the vehicle-mounted audio equipment to increase the volume, and if the target volume adjustment predicted value is zero, the current volume is kept unchanged.
If the target volume adjustment predicted value is the absolute value of the volume, the range of the target volume adjustment predicted value is between zero and the positive maximum volume value, a volume adjustment instruction is generated through the target volume adjustment predicted value, and the volume adjustment instruction is sent to the audio equipment in the vehicle, so that the current playing volume is adjusted to be consistent with the target volume adjustment predicted value.
According to the volume adjusting method, various types of vehicle features including vehicle environment features and vehicle control state features are collected, collected feature information is processed through a prediction model, a target volume adjustment prediction value is output, vehicle audio is regulated according to the volume adjustment prediction value, automatic regulation and control of playing volume of vehicle audio equipment are achieved based on various vehicle features, manual operation of vehicle interior personnel on vehicle machines such as volume control in the driving process is reduced, burden of the vehicle interior personnel is reduced, driving potential safety hazards are reduced, and the concentration degree of the driver in driving the vehicle is improved.
Optionally, on the basis of the method shown in fig. 1, the embodiment of the present application further provides another possible implementation example of obtaining the vehicle characteristic in the volume adjustment method, which is described below with reference to the accompanying drawings. Fig. 2 is a flowchart of acquiring vehicle characteristics in a volume adjustment method according to an embodiment of the present disclosure. As shown in fig. 2, the acquiring of multiple types of vehicle characteristics in S100 in the volume adjustment method may include:
and S110, carrying out feature extraction on the vehicle environment image to obtain vehicle image features.
Before feature extraction is performed on the vehicle environment image, the vehicle environment image can be acquired. The vehicle environmental characteristics may include: an in-vehicle environment image and/or an out-vehicle environment image. The in-vehicle environment image may be in-vehicle image information acquired by a camera in the vehicle, and the out-vehicle environment image may be an out-vehicle image acquired by a camera installed outside the vehicle, for example, at least one position of the vehicle body. If a plurality of cameras are installed at a plurality of positions of the vehicle, such as the head, the tail, the body and the like, and images of the surrounding environment of the vehicle can be collected completely, the images of the environment outside the vehicle can be collected vehicle panoramic images.
For example, the vehicle image feature may be obtained by performing feature extraction on the vehicle interior environment image, and the vehicle exterior image feature may be obtained by performing feature extraction on the vehicle exterior environment image, and therefore, the vehicle image feature may include: an in-vehicle image feature and/or an out-of-vehicle image feature.
For example, a pre-trained image feature extraction model may be adopted to perform feature extraction on the vehicle environment image to obtain the vehicle image features. The image feature extraction model can be an image feature extraction model of a neural network structure, and the vehicle image feature is a one-dimensional vector.
The image feature extraction model includes: input layer, convolution layer, pooling layer, full connection layer. The input of the convolutional layer is that the input layer scales the input vehicle environment image to a vehicle environment image with a preset size, and the output of the convolutional layer is used as the input of the pooling layer. The last layer outputs the vehicle image features. The input layer can scale the input vehicle environment image to a preset size such as 256 pixels by 256 pixels; the convolution layer can perform feature extraction by performing inner product on each window unit in the vehicle environment image subjected to input layer scaling through a plurality of convolution checks, and the specific convolution operation is to perform inner product on each window unit of the image through convolution checks.
The image information features of the pooling layer are compressed, so that the input data volume can be reduced, important features in the features can be extracted, redundant information is eliminated, when maximum pooling is used, only the local maximum features can be reserved in a result returned by pooling, and the important feature extraction and feature simplification are performed in the pooling layer.
And performing memorability integration and identification on the local feature information extracted by the full connection layer, performing parameter optimization on the previous three-layer model through identification, extracting all local features, and integrating the local features to form a complete image input feature.
The image feature extraction model may further include: the activation layer performs feature conversion through an activation function, the activation function is realized through a nonlinear function, the neural network can be stronger by introducing the nonlinearity, fitting of auxiliary data is guaranteed, the realization process is that each feature is calculated through a function, such as a sigmiod function, and the obtained result is the activated result.
And S120, performing feature extraction on the in-vehicle input voice to obtain in-vehicle voice features.
The feature extraction can be carried out on the in-vehicle input voice by adopting a preset voice feature extraction mode to obtain the in-vehicle voice feature. The in-vehicle voice feature is the voice of the person in the vehicle speaking. The in-vehicle voice feature may be voice capture through a microphone disposed inside the vehicle.
For example, a specific audio feature extraction mode can be adopted to perform feature extraction on the in-vehicle input speech, and the extracted speech features are fused and stored in a one-dimensional vector form through the audio feature extraction mode.
And S130, integrating the vehicle image characteristics and the in-vehicle voice characteristics to obtain vehicle environment characteristics.
Integrating the vehicle image features and the in-vehicle voice features to obtain one-dimensional vectors of the vehicle image features and one-dimensional vectors of the in-vehicle voice features, splicing the one-dimensional vectors of the vehicle image features and the one-dimensional vectors of the in-vehicle voice features, wherein the splicing process is similar to that the two vectors of [1,0, 2] and [3,4,3] are spliced to form a vector of [1,0,2,3,4,3], and then obtaining the vehicle environment features.
And S140, carrying out feature extraction on the vehicle control state to obtain vehicle control state features.
After the vehicle environment characteristics are obtained, the current vehicle state characteristics can be obtained by performing characteristic extraction on the vehicle control state and can be used as auxiliary characteristics to facilitate the regulation and control of the playing volume of subsequent vehicle audio equipment.
The execution sequence of S110, S120, and S140 is not limited.
According to the method provided by the embodiment, the acquisition of various types of vehicle characteristics can be enriched by providing the acquisition mode of the vehicle characteristics, so that the situation of the current vehicle can be more accurately acquired through the acquired various types of vehicle characteristics, the most accurate characteristic collection is carried out on the situation of the current vehicle, and the vehicle can be more accurately operated according to the situation of the current vehicle.
On the basis of the foregoing embodiments, the embodiments of the present application may further provide a possible implementation example of extracting a vehicle voice feature in a volume adjustment method, which is described below with reference to the accompanying drawings. Fig. 3 is a flowchart illustrating a method for adjusting a volume according to an embodiment of the present disclosure to extract a vehicle voice feature. As shown in fig. 3, the extracting the features of the in-vehicle input speech in S120 in the volume adjustment method to obtain the in-vehicle speech features includes:
and S121, respectively extracting the features of the input voice in the vehicle by adopting a plurality of audio feature extraction modes to obtain the voice features corresponding to the plurality of feature extraction modes.
Illustratively, the plurality of audio feature extraction methods include at least two audio feature extraction methods: a Mel frequency cepstrum coefficient feature extraction mode, a zero-crossing rate feature extraction mode and a spectrum centroid feature extraction mode.
The Mel Frequency Cepstrum Coefficient (MFCC) feature extraction method is a common speech feature extraction method, and the main process comprises the following steps: framing, windowing, FFT, taking absolute values, Mel filtering, taking logarithms, DCT, and dynamic features. MFCC processes audio signals by pre-emphasizing some signals, framing, windowing, filtering model human auditory principles, enhancing important information, and filtering useless information. The realization process is as follows: performing framing processing on a voice signal; performing power spectrum (power spectrum) estimation by using a periodogram (periodogram) method; filtering the power spectrum by using a Mel filter bank, and calculating the energy in each filter; taking log of the energy of each filter; performing a Discrete Cosine Transform (DCT) transform; the 2 nd to 13 th coefficients of the DCT are reserved, and other main processes are removed, including: framing, windowing, FFT, taking absolute values, Mel-filtering, taking logarithms, DCT, and dynamic features.
The Zero Crossing Rate (Zero Crossing Rate) feature extraction mode refers to the number of Zero Crossing points of a voice signal in each frame, is widely used in the fields of voice recognition and music information retrieval, reflects the Rate of signal symbol change, and is specifically realized by calculating the number of Zero Crossing points of the voice signal in each frame.
The Spectral Centroid (Spectral Centroid) feature extraction mode is an important physical parameter for describing tone color attributes, is the center of gravity of frequency components, is a weighted average of energy in a certain frequency range, is important information of frequency distribution and energy distribution of sound signals, and is specifically realized by calculating average values 0011,0012 and 0013 of each frame of frequency, namely performing feature processing on the sound signals, and performing encoding operation on sound physical information to obtain a group of vectors.
And S122, performing feature integration on the voice features corresponding to the multiple feature extraction modes to obtain the in-vehicle voice features.
The frontal speech features extracted by the Mel frequency cepstrum coefficient feature extraction mode, the zero-crossing rate feature extraction mode and the spectral centroid feature extraction mode are processed into three groups of one-dimensional vectors, the three groups of one-dimensional vectors can be 0011,0012,0013, the three groups of one-dimensional vectors are combined into one-dimensional vector in a splicing mode, and finally the in-vehicle speech features are obtained.
According to the method provided by the embodiment, the characteristics of the in-vehicle input voice are extracted through various audio characteristic extraction modes, the extracted characteristics are integrated, the vehicle voice characteristics are obtained, the accuracy of voice characteristic extraction is guaranteed, the intention of people in the vehicle is correctly reflected through accurate voice characteristic extraction, and therefore the vehicle can be operated more accurately according to the current in-vehicle voice environment.
On the basis of the foregoing embodiments, the embodiments of the present application may further provide a possible implementation example of extracting the vehicle control state feature in the volume adjustment method, which is described below with reference to the accompanying drawings. Fig. 4 is a flowchart of a vehicle control status feature in a volume adjustment method according to an embodiment of the present disclosure. As shown in fig. 4, the extracting the feature of the vehicle control state in S140 in the volume adjustment method to obtain the vehicle control state feature may include:
and S141, performing feature extraction on the in-vehicle audio playing volume in a continuous value mode to obtain vehicle control volume features.
The feature extraction is performed on the in-vehicle audio playback volume in a continuous value manner, for example, the value of the volume feature extraction is 0.5 when the volume is in the middle.
And S142, carrying out feature extraction on the skylight opening state in a discrete value mode to obtain the vehicle control skylight feature.
And (3) carrying out feature extraction on the opening state of the skylight in a discrete value mode, wherein the opening state of the skylight is 1, the closing state of the skylight is 0, and each skylight is independently judged.
And S143, carrying out feature extraction on the opening state of the vehicle window in a discrete value mode to obtain the vehicle control window features.
And (3) carrying out feature extraction on the opening state of the car window in a discrete value mode, wherein the opening state of the car window is 1, the closing state of the car window is 0, and each car window is independently judged.
The vehicle control state characteristic comprises at least one of the following characteristics: a vehicle control volume characteristic, a vehicle control skylight characteristic, and a vehicle control window characteristic.
In a possible implementation manner, the vehicle control state feature may be one of a vehicle control volume feature, a vehicle control skylight feature and a vehicle control window feature, may be a vehicle control feature combined in pairs, and may be a vehicle control state feature included in all three vehicle control features.
The execution order of S141, S142, and S143 is not limited.
The method for adjusting the volume extracts the vehicle control state characteristics, the vehicle control volume characteristics, the vehicle control skylight characteristics, the vehicle control window characteristics and other characteristics are extracted to serve as auxiliary characteristics to provide the current most real state of the vehicle, the vehicle control state characteristics are calculated and processed, the fact that the volume control of the vehicle is closer to an ideal control state is guaranteed, and the accuracy of the vehicle volume control is improved.
Optionally, on the basis of the method shown in fig. 1, an embodiment of the present application further provides a possible implementation example of obtaining a target volume adjustment predicted value in a volume adjustment method. In this implementation example, the predictive model may include, for example: and each leaf node on each characteristic classification tree is a vehicle characteristic corresponding to a volume adjustment action.
The prediction model may be an XGBoost (eXtreme Gradient Boosting) machine learning model, and the pre-training model may also be any one or a combination of a support vector machine model, a neural network model, a deep learning model, and the like.
Through a boosting algorithm in the XGboost, the XGboost can generate a plurality of feature classification trees, each feature classification tree corresponds to one type of vehicle feature, through a plurality of times of training, the next feature classification tree is fit to the residual error of the previous feature classification tree, when the training is completed, the feature classification trees with the preset number are the feature classification trees corresponding to the plurality of features, and leaf nodes of each feature classification tree store corresponding volume adjusting parameters.
As will be explained below in conjunction with the drawings. Fig. 5 is a flowchart illustrating obtaining a target volume adjustment predicted value in a volume adjustment method according to an embodiment of the present disclosure. As shown in fig. 5, in the above method, the predicting the multiple types of vehicle features by using the pre-trained prediction model in S200 to obtain the target volume adjustment prediction value may include:
s210, according to the vehicle environment characteristics and the vehicle control state characteristics, determining leaf nodes where the vehicle characteristics of each type are located from the characteristic classification trees corresponding to the vehicle characteristics of each type in the prediction model as target leaf nodes, and determining volume adjustment parameter values of volume adjustment actions corresponding to the target leaf nodes as a volume adjustment predicted value corresponding to the vehicle characteristics of each type.
In the prediction model, each feature classification number corresponds to a vehicle feature, each vehicle feature is analyzed and calculated through the pre-trained prediction model, so that each vehicle feature can finally fall into a corresponding leaf node through a feature classification tree, the leaf node where the vehicle feature finally falls is a target leaf node, volume adjusting parameters are stored in the target leaf node, the volume adjusting parameters are one-dimensional vectors, and the value of each vector is between 0 and 1.
And S220, obtaining a target volume adjusting parameter value according to the volume adjusting predicted value corresponding to the various vehicle characteristics.
Adding the volume adjusting parameters of the target leaf nodes in which each vehicle characteristic falls to obtain a volume adjusting parameter, wherein the volume adjusting parameter is a final volume adjusting predicted value, for example, the current volume is 0.4 (the maximum volume is 1, the minimum volume is 0), the model predicted value is 0.1, then the volume is adjusted to 0.5, when the predicted value is-0.2, the sound is adjusted to 0.5 to 0.3, and when the model predicted value is 0, the volume is not adjusted.
According to the prediction model of the volume adjustment method, overfitting of the model is prevented through the prediction model provided by the embodiment of the application, parallelization is supported, when the optimal split point is selected, multithreading parallelization is carried out when gain is calculated on the candidate split points, the training speed is improved, the collected vehicle characteristics are calculated and processed, necessary details for achieving volume adjustment are guaranteed, the hands of a driver are liberated, the burden of people in the vehicle is reduced, the potential safety hazard of driving is reduced, and the concentration degree of the driver for driving the vehicle is improved.
The following describes a device, an apparatus, a storage medium, and the like for executing the volume adjustment method provided in the embodiments of the present application, and specific implementation processes and technical effects thereof are referred to above, and are not described again below.
Fig. 6 is a schematic view of a volume adjustment device according to an embodiment of the present disclosure, and as shown in fig. 6, the volume adjustment device may include:
an acquisition module 11 for acquiring a plurality of types of vehicle characteristics, the plurality of types of vehicle characteristics including: vehicle environmental characteristics and vehicle control status characteristics.
The prediction module 12 is configured to perform prediction processing on multiple types of vehicle characteristics by using a pre-trained prediction model to obtain a target volume adjustment prediction value;
and the adjusting module 13 is configured to adjust and control the playing volume of the vehicle-mounted audio according to the target volume adjustment prediction value.
Optionally, the obtaining module 11 is specifically configured to perform feature extraction on the vehicle environment image to obtain vehicle image features; performing feature extraction on the in-vehicle input voice to obtain in-vehicle voice features; integrating the vehicle image characteristics and the in-vehicle voice characteristics to obtain vehicle environment characteristics; and carrying out feature extraction on the vehicle control state to obtain vehicle control state features.
Optionally, the obtaining module 11 is further specifically configured to perform feature extraction on the vehicle environment image by using a pre-trained image feature extraction model to obtain vehicle image features.
Optionally, the obtaining module 11 is further specifically configured to perform feature extraction on the in-vehicle input speech by using multiple audio feature extraction manners, so as to obtain speech features corresponding to the multiple feature extraction manners; and performing feature integration on the voice features corresponding to the multiple feature extraction modes to obtain the in-vehicle voice features.
Optionally, the multiple audio feature extraction manners include at least two audio feature extraction manners: a Mel frequency cepstrum coefficient feature extraction mode, a zero-crossing rate feature extraction mode and a spectrum centroid feature extraction mode.
Optionally, the obtaining module 11 is further specifically configured to perform feature extraction on the in-vehicle audio playing volume in a continuous value manner, so as to obtain a vehicle control volume feature; carrying out feature extraction on the skylight opening state in a discrete value mode to obtain the vehicle control skylight features; carrying out feature extraction on the opening state of the vehicle window in a discrete value mode to obtain the features of the vehicle control vehicle window; the vehicle control state characteristics include: a vehicle control volume characteristic and/or a vehicle control skylight characteristic and/or a vehicle control window characteristic.
Optionally, the prediction model includes: each characteristic classification tree corresponds to one type of vehicle characteristic, and each leaf node on each characteristic classification tree is a vehicle characteristic corresponding to a volume adjustment action;
the prediction module 12 is specifically configured to determine, according to the vehicle environment characteristic and the vehicle control state characteristic, a leaf node where each type of vehicle characteristic is located from a characteristic classification tree corresponding to each type of vehicle characteristic in the prediction model as a target leaf node, and determine that a volume adjustment parameter value of a volume adjustment action corresponding to the target leaf node is a volume adjustment prediction value corresponding to each type of vehicle characteristic.
The adjusting module 13 is specifically configured to obtain a target volume adjustment parameter value according to the volume adjustment predicted value corresponding to the multiple types of vehicle characteristics.
The above-mentioned apparatus is used for executing the method provided by the foregoing embodiment, and the implementation principle and technical effect are similar, which are not described herein again.
These above modules may be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or one or more microprocessors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs), among others. For another example, when one of the above modules is implemented in the form of a Processing element scheduler code, the Processing element may be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. As another example, these modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Fig. 7 is a schematic diagram of a computer device provided in an embodiment of the present application. The computer device 1000 comprises: memory 1001, processor 1002. The memory 1001 and the processor 1002 are connected by a bus.
The memory 1001 is used for storing programs, and the processor 1002 calls the programs stored in the memory 1001 to execute the above-mentioned method embodiments. The specific implementation and technical effects are similar, and are not described herein again.
Optionally, the invention also provides a program product, for example a computer-readable storage medium, comprising a program which, when being executed by a processor, is adapted to carry out the above-mentioned method embodiments.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a portable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other media capable of storing program codes.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily think of the changes or substitutions within the technical scope of the present application, and shall cover the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (7)
1. A method of volume adjustment, the method comprising:
obtaining a plurality of types of vehicle features, the plurality of types of vehicle features comprising: vehicle environmental characteristics and vehicle control state characteristics;
adopting a pre-trained prediction model to carry out prediction processing on the various vehicle characteristics to obtain a target volume adjustment prediction value;
regulating and controlling the playing volume of the vehicle-mounted audio equipment according to the target volume regulation predicted value;
the acquiring of multiple types of vehicle features includes:
carrying out feature extraction on the vehicle environment image to obtain vehicle image features;
carrying out feature extraction on the in-vehicle input voice to obtain in-vehicle voice features;
integrating the vehicle image features and the in-vehicle voice features to obtain the vehicle environment features;
carrying out feature extraction on a vehicle control state to obtain vehicle control state features;
the prediction model includes: each characteristic classification tree corresponds to one type of vehicle characteristic, and each leaf node on each characteristic classification tree is a vehicle characteristic corresponding to a volume adjustment action;
the adopting a pre-trained prediction model to perform prediction processing on the multiple types of vehicle characteristics to obtain a target volume adjustment prediction value comprises the following steps:
according to the vehicle environment characteristics and the vehicle control state characteristics, determining a leaf node where each type of vehicle characteristics is located as a target leaf node from a characteristic classification tree corresponding to each type of vehicle characteristics in the prediction model, and determining a volume adjustment parameter value of a volume adjustment action corresponding to the target leaf node as a volume adjustment predicted value corresponding to each type of vehicle characteristics;
obtaining a target volume adjustment parameter value according to the volume adjustment predicted value corresponding to the various types of vehicle characteristics;
the feature extraction of the vehicle control state to obtain the vehicle control state features comprises the following steps:
carrying out feature extraction on the in-vehicle audio playing volume in a continuous value mode to obtain vehicle control volume features;
carrying out feature extraction on the skylight opening state in a discrete value mode to obtain the characteristics of the vehicle control skylight;
carrying out feature extraction on the opening state of the vehicle window in a discrete value mode to obtain the features of the vehicle control vehicle window;
the vehicle control state characteristic comprises at least one of the following characteristics: the vehicle control volume characteristic, the vehicle control skylight characteristic and the vehicle control window characteristic.
2. The method of claim 1, wherein the performing feature extraction on the vehicle environment image to obtain vehicle image features comprises:
and adopting a pre-trained image feature extraction model to extract the features of the vehicle environment image to obtain the vehicle image features.
3. The method according to claim 1, wherein the performing feature extraction on the in-vehicle input speech to obtain the in-vehicle speech feature comprises:
respectively extracting the features of the in-vehicle input voice by adopting a plurality of audio feature extraction modes to obtain voice features corresponding to the plurality of audio feature extraction modes;
and performing feature integration on the voice features corresponding to the multiple feature extraction modes to obtain the in-vehicle voice features.
4. The method of claim 3, wherein the plurality of audio feature extraction methods comprises at least two of the following audio feature extraction methods: a Mel frequency cepstrum coefficient feature extraction mode, a zero-crossing rate feature extraction mode and a spectrum centroid feature extraction mode.
5. A volume adjustment device, characterized in that the device comprises:
an acquisition module to acquire a plurality of types of vehicle characteristics, the plurality of types of vehicle characteristics including: vehicle environmental characteristics and vehicle control state characteristics;
the prediction module is used for carrying out prediction processing on the various types of vehicle characteristics by adopting a pre-trained prediction model to obtain a target volume adjustment prediction value;
the acquisition module is also used for carrying out feature extraction on the vehicle environment image to obtain the vehicle image features; carrying out feature extraction on the in-vehicle input voice to obtain in-vehicle voice features; integrating the vehicle image characteristics and the in-vehicle voice characteristics to obtain vehicle environment characteristics; carrying out feature extraction on the vehicle control state to obtain vehicle control state features;
the prediction model includes: each characteristic classification tree corresponds to one type of vehicle characteristic, and each leaf node on each characteristic classification tree is a vehicle characteristic corresponding to a volume adjustment action;
the prediction module is further used for determining a leaf node where each type of vehicle feature is located as a target leaf node from a feature classification tree corresponding to each type of vehicle feature in the prediction model according to the vehicle environment feature and the vehicle control state feature, and determining a volume adjustment parameter value of a volume adjustment action corresponding to the target leaf node as a volume adjustment prediction value corresponding to each type of vehicle feature;
the adjusting module is further used for obtaining a target volume adjusting parameter value according to the volume adjusting predicted value corresponding to the multiple types of vehicle characteristics;
the acquisition module is also used for carrying out feature extraction on the in-vehicle audio playing volume in a continuous value mode to obtain vehicle control volume features;
carrying out feature extraction on the skylight opening state in a discrete value mode to obtain the vehicle control skylight features;
carrying out feature extraction on the opening state of the vehicle window in a discrete value mode to obtain the features of the vehicle control vehicle window;
the vehicle control state characteristic comprises at least one of the following characteristics: the vehicle control volume characteristic, the vehicle control skylight characteristic and the vehicle control window characteristic.
6. A computer device, comprising: a memory storing a computer program executable by the processor, and a processor implementing the volume adjusting method according to any one of claims 1 to 4 when the computer program is executed by the processor.
7. A computer-readable storage medium, having stored thereon a computer program which, when executed by a processor, performs a volume adjustment method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110805665.XA CN113504891B (en) | 2021-07-16 | 2021-07-16 | Volume adjusting method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110805665.XA CN113504891B (en) | 2021-07-16 | 2021-07-16 | Volume adjusting method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113504891A CN113504891A (en) | 2021-10-15 |
| CN113504891B true CN113504891B (en) | 2022-09-02 |
Family
ID=78013558
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110805665.XA Active CN113504891B (en) | 2021-07-16 | 2021-07-16 | Volume adjusting method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113504891B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115107674A (en) * | 2022-01-24 | 2022-09-27 | 长城汽车股份有限公司 | Volume adjusting method and device and automobile |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107800793A (en) * | 2017-10-27 | 2018-03-13 | 江苏大学 | Car networking environment down train music active pushing system |
| CN110366070A (en) * | 2019-06-27 | 2019-10-22 | 广州小鹏汽车科技有限公司 | Method for controlling volume and system, the automobile of onboard system |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9800716B2 (en) * | 2010-09-21 | 2017-10-24 | Cellepathy Inc. | Restricting mobile device usage |
| CN108146360A (en) * | 2017-12-25 | 2018-06-12 | 出门问问信息科技有限公司 | Method, apparatus, mobile unit and the readable storage medium storing program for executing of vehicle control |
| CN110659640B (en) * | 2019-09-27 | 2021-11-30 | 深圳市商汤科技有限公司 | Text sequence recognition method and device, electronic equipment and storage medium |
| CN110995933A (en) * | 2019-12-12 | 2020-04-10 | Oppo广东移动通信有限公司 | Volume adjusting method and device of mobile terminal, mobile terminal and storage medium |
| CN111045642A (en) * | 2019-12-17 | 2020-04-21 | 上海博泰悦臻电子设备制造有限公司 | Volume adjusting method, vehicle-mounted terminal and computer storage medium |
| CN111816199A (en) * | 2020-07-11 | 2020-10-23 | 韶关市启之信息技术有限公司 | Environmental sound control method and system for intelligent cabin of automobile |
| CN111933175B (en) * | 2020-08-06 | 2023-10-24 | 北京中电慧声科技有限公司 | Active voice detection method and system based on noise scene recognition |
| CN112308330B (en) * | 2020-11-09 | 2021-07-09 | 清华大学 | Digital accident database construction method and device and computer equipment |
| CN112669822B (en) * | 2020-12-16 | 2022-11-25 | 爱驰汽车有限公司 | Audio processing method and device, electronic equipment and storage medium |
| CN112785016B (en) * | 2021-02-20 | 2022-06-07 | 南京领行科技股份有限公司 | New energy automobile maintenance and fault monitoring and diagnosis method based on machine learning |
-
2021
- 2021-07-16 CN CN202110805665.XA patent/CN113504891B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107800793A (en) * | 2017-10-27 | 2018-03-13 | 江苏大学 | Car networking environment down train music active pushing system |
| CN110366070A (en) * | 2019-06-27 | 2019-10-22 | 广州小鹏汽车科技有限公司 | Method for controlling volume and system, the automobile of onboard system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113504891A (en) | 2021-10-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11908455B2 (en) | Speech separation model training method and apparatus, storage medium and computer device | |
| DE112017003563B4 (en) | METHOD AND SYSTEM OF AUTOMATIC LANGUAGE RECOGNITION USING POSTERIORI TRUST POINT NUMBERS | |
| JP4166153B2 (en) | Apparatus and method for discriminating emotion of dog based on analysis of voice characteristics | |
| CN110310647B (en) | Voice identity feature extractor, classifier training method and related equipment | |
| CN109243490A (en) | Driver's Emotion identification method and terminal device | |
| US20140114655A1 (en) | Emotion recognition using auditory attention cues extracted from users voice | |
| EP3664084B1 (en) | Electronic device and control method therefor | |
| CN113539242A (en) | Speech recognition method, speech recognition device, computer equipment and storage medium | |
| CN114127849A (en) | Speech emotion recognition method and device | |
| CN111028833B (en) | Interaction method and device for interaction and vehicle interaction | |
| CN113504891B (en) | Volume adjusting method, device, equipment and storage medium | |
| CN113516990A (en) | Voice enhancement method, method for training neural network and related equipment | |
| Hamsa et al. | Speaker identification from emotional and noisy speech using learned voice segregation and speech VGG | |
| Loh et al. | Speech recognition interactive system for vehicle | |
| CN116959464A (en) | Training method of audio generation network, audio generation method and device | |
| CN115547345A (en) | Voiceprint recognition model training and related recognition method, electronic device and storage medium | |
| Chen et al. | CITISEN: A Deep Learning-Based Speech Signal-Processing Mobile Application | |
| CN114299981A (en) | Audio processing method, device, storage medium and equipment | |
| CN115881157A (en) | Audio signal processing method and related equipment | |
| Jagadeeshwar et al. | ASERNet: Automatic speech emotion recognition system using MFCC-based LPC approach with deep learning CNN | |
| CN113345464A (en) | Voice extraction method, system, device and storage medium | |
| CN116705013B (en) | Voice wake-up word detection method and device, storage medium and electronic equipment | |
| Nikitaras et al. | Fine-grained noise control for multispeaker speech synthesis | |
| Long et al. | Offline to online speaker adaptation for real-time deep neural network based LVCSR systems | |
| WO2024055752A1 (en) | Speech synthesis model training method, speech synthesis method, and related apparatuses |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PP01 | Preservation of patent right | ||
| PP01 | Preservation of patent right |
Effective date of registration: 20230131 Granted publication date: 20220902 |
|
| PD01 | Discharge of preservation of patent | ||
| PD01 | Discharge of preservation of patent |
Date of cancellation: 20240108 Granted publication date: 20220902 |
|
| PP01 | Preservation of patent right | ||
| PP01 | Preservation of patent right |
Effective date of registration: 20240227 Granted publication date: 20220902 |