CN108630228A - Sound quality recognition methods, device, system and vehicle - Google Patents
Sound quality recognition methods, device, system and vehicle Download PDFInfo
- Publication number
- CN108630228A CN108630228A CN201710165978.7A CN201710165978A CN108630228A CN 108630228 A CN108630228 A CN 108630228A CN 201710165978 A CN201710165978 A CN 201710165978A CN 108630228 A CN108630228 A CN 108630228A
- Authority
- CN
- China
- Prior art keywords
- sound quality
- sound
- sample
- information
- audio
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/27—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique
- G10L25/30—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique using neural networks
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/27—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique
- G10L25/39—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the analysis technique using genetic algorithms
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The present invention relates to simulation technical field, disclosing a kind of sound quality recognition methods, device, system and vehicle, this method includes：Obtain the acoustic information in environment；And according to the acoustic information and sound quality identification model, determine the sound quality of the acoustic information.The present invention can carry out sound quality evaluation or identification in real time, rapidly, accurately, effectively to the acoustic information in the environment of acquisition, realization objectifies the result of subjective assessment, working time and the job costs for greatly reducing related personnel have saved time cost and labour cost while improving sound quality and evaluating accuracy.
Description
Technical field
The present invention relates to simulation technical fields, more particularly to a kind of sound quality recognition methods, a kind of sound quality identification dress
It sets, a kind of sound quality identifying system and a kind of vehicle.
Background technology
With the development of science and technology with progress, people’s lives level becomes better and better, to the pursuit of quality of the life also gradual
It improves.For example, for the sound in specific environment (such as vehicle interior etc.), people are not only concerned about the " noise of acoustic environment now
It is big less ", and more concerned with acoustic environment " sound relaxes uncomfortable ", and the latter is exactly sound quality problem.And traditional sound measurement is only
" noise size " can be measured in acoustic environment, and " the sound comfort level " of acoustic environment cannot be evaluated.
It is general in the prior art to use the objective analysis of related software (Artemis sound quality analysis softwares etc.) and personnel
The mode of combination of subjective assessment the sound quality of sample sound is evaluated.Although this method is very effectively, quite
It is time-consuming and laborious, such as subjective assessment not only will carry out relevant audition training to evaluation personnel, but also also need to according to sound
The difference of signal selects different evaluation methods, and many homework burdens are brought to relevant practitioner.I.e. in the prior art
Lack a kind of scheme that sound quality evaluation or identification can in real time, rapidly, accurately, be effectively carried out to sample sound.
Invention content
The purpose of the invention is to overcome, of the existing technology lack can in real time, rapidly, accurately, effectively
The problem of ground carries out sound quality evaluation or identification to sample sound, provide a kind of sound quality recognition methods, device, system and
Vehicle.
To achieve the goals above, one aspect of the present invention provides a kind of sound quality recognition methods, and this method includes：Obtain ring
Acoustic information in border；And according to the acoustic information and sound quality identification model, determine the sound product of the acoustic information
Matter.
Preferably, the sound quality identification model is established according to following steps：Obtain multiple sample sound in the environment
Message ceases and the corresponding sample sound quality information of the multiple sample audio information；And by the multiple sample audio information
As the input of neural network model, corresponding to described in each sample audio information in the multiple sample audio information
Sample sound quality information is trained the neural network model as the corresponding output of neural network model, described in generation
Sound quality identification model.
Preferably, the neural network model is established using BP neural network algorithm.
Preferably, the neural network model is established using the improved BP neural network algorithm of Genetic Algorithms.
Preferably, the sound quality that the acoustic information is determined according to the acoustic information and sound quality identification model
Including：Using the acoustic information as the input of the sound quality identification model, the sound quality identification model is run；And really
The recognition result of sound quality of the output of the fixed sound quality identification model as the acoustic information.
Preferably, the acoustic information include it is below at least one：Wave audio data, wave audio data are corresponding
The loudness of sound, the sharpness of the corresponding sound of wave audio data, the corresponding sound of wave audio data shake degree and
The roughness of the corresponding sound of wave audio data；The sample audio information include it is below at least one：Sample audio sound
Sharpness, the sample of frequency evidence, the loudness of the corresponding sound of sample audio audio data, the corresponding sound of sample audio audio data
The shake degree of the corresponding sound of this wave audio data and the roughness of the corresponding sound of sample audio audio data；And
The sound quality of the acoustic information includes at least one of sound quality grade harmony quality evaluation value；The sample audio information
Sound quality include at least one of sample sound quality grade and sample sound quality evaluation of estimate.
Second aspect of the present invention provides a kind of sound quality identification device, which includes：Acquisition module, for obtaining environment
In acoustic information；And identification module, for determining the sound according to the acoustic information and sound quality identification model
The sound quality of information.
Preferably, the sound quality identification model is established according to following steps：Obtain multiple sample sound in the environment
Message ceases and the corresponding sample sound quality information of the multiple sample audio information；And by the multiple sample audio information
As the input of neural network model, corresponding to described in each sample audio information in the multiple sample audio information
Sample sound quality information is trained the neural network model as the corresponding output of neural network model, described in generation
Sound quality identification model.
Preferably, the neural network model uses BP neural network algorithm or the improved BP nerve nets of Genetic Algorithms
Network algorithm is established.
Preferably, the identification module is additionally operable to：Using the acoustic information as the input of the sound quality identification model,
Run the sound quality identification model；And determine sound product of the output of the sound quality identification model as the acoustic information
The recognition result of matter.
Preferably, which further includes：Display module, the recognition result of the sound quality for showing the acoustic information.
Preferably, the acoustic information include it is below at least one：Wave audio data, wave audio data are corresponding
The loudness of sound, the sharpness of the corresponding sound of wave audio data, the corresponding sound of wave audio data shake degree and
The roughness of the corresponding sound of wave audio data；The sample audio information include it is below at least one：Sample audio sound
Sharpness, the sample of frequency evidence, the loudness of the corresponding sound of sample audio audio data, the corresponding sound of sample audio audio data
The shake degree of the corresponding sound of this wave audio data and the roughness of the corresponding sound of sample audio audio data；And
The sound quality of the acoustic information includes at least one of sound quality grade harmony quality evaluation value；The sample audio information
Sound quality include at least one of sample sound quality grade and sample sound quality evaluation of estimate.
Third aspect present invention provides a kind of sound quality identifying system, which includes：Voice collection device, for acquiring
Acoustic information in environment；And sound quality identification device provided by the invention, the sound quality identification device and the sound
Harvester connects.
Preferably, the voice collection device is ears microphone.
Fourth aspect present invention provides a kind of vehicle including above-mentioned sound quality identifying system.
Through the above technical solutions, can in real time, rapidly, accurately, effectively to the sound in the environment of acquisition
Information carries out sound quality evaluation or identification, and realization objectifies the result of subjective assessment, greatly reduces the work of related personnel
Time and job costs have saved time cost and labour cost while improving sound quality and evaluating accuracy.
Description of the drawings
Fig. 1 is a kind of structural schematic diagram of the example sound quality identifying system of embodiment according to the present invention；
Fig. 2 is a kind of structural schematic diagram of the example sound quality identification device of embodiment according to the present invention；
Fig. 3 is a kind of structural schematic diagram of the example sound quality identification model of embodiment according to the present invention；
Fig. 4 is a kind of structural schematic diagram of the example sound quality identification device of embodiment according to the present invention；
Fig. 5 is a kind of structural schematic diagram of the example sound quality identifying system of embodiment according to the present invention；And
Fig. 6 is a kind of flow chart of the example sound quality recognition methods of embodiment according to the present invention.
Specific implementation mode
The specific implementation mode of the present invention is described in detail below in conjunction with attached drawing.It should be understood that this place is retouched
The specific implementation mode stated is merely to illustrate and explain the present invention, and is not intended to restrict the invention.
In order to realize in real time, rapidly, accurately, effectively in the environment of acquisition acoustic information carry out sound quality
Evaluation or identification, the present invention consider various embodiments, will hereafter be described in detail one by one：
Embodiment 1
Fig. 1 is a kind of structural schematic diagram of the example sound quality identifying system 1000 of embodiment according to the present invention, such as
Shown in Fig. 1, which may include：Voice collection device 200, for acquiring the acoustic information in environment；It is provided by the invention
Sound quality identification device 100, the sound quality identification device 100 are connect with the voice collection device 200.
Specifically, the environment may include various application environments, such as vehicle interior, studio hall etc..The sound is adopted
Acquisition means 200 can be the device of any acoustic information (such as voice signal) that can be acquired in environment.For example, the sound
Harvester be (wear-type) ears microphone 200a, can restore to the greatest extent the signal that human ear is heard authenticity and
The sound quality that simulation human ear is heard in true environment, to improve the accuracy of data source.
Then, voice collection device 200 can send collected acoustic information to sound quality identification device 100.Sound
It can configuration acquisition the above sound information according to the present invention after quality identification device 100；And according to the acoustic information
And sound quality identification model determines the sound quality of the acoustic information.The sound quality can be sound quality grade and/or sound
Quality evaluation value.For example, sound quality grade (such as basic, normal, high grade) or evaluation of estimate (such as fractional value).For sound quality
Identification device 100 will in more detail be illustrated in following examples.
Using system described in embodiment 1, can in real time, rapidly, accurately, effectively in the environment of acquisition
Acoustic information carries out sound quality evaluation or identification, and realization objectifies the result of subjective assessment, greatly reduces related personnel's
Working time and job costs have saved time cost and labour cost, significantly while improving sound quality and evaluating accuracy
The flow evaluated in the past Sound Environment Quality is simplified, allows the intuitive Real-time Feedback of sound quality evaluation result to measurement
Personnel can greatly improve working efficiency.The system can be used in the acoustic environment in need evaluated, not by space
With the limitation of environment.
Embodiment 2
Fig. 2 is a kind of structural schematic diagram of the example sound quality identification device 100 of embodiment according to the present invention, is such as schemed
Shown in 2, which may include：Acquisition module 101, for obtaining the acoustic information in environment；And identification module 102, it uses
In the sound quality for determining the acoustic information according to the acoustic information and sound quality identification model.
Wherein, acquisition module 101 may be configured to obtain the acoustic information in environment, and the acoustic information may include
One or more of below：
(1) wave audio data, such as 200 collected voice signal of the above sound harvester；
(2) the loudness N of the corresponding sound of wave audio data, is calculated for example, by using Zwicker methods, reflects human ear pair
The subjective feeling degree of sound intensity；
(3) the sharpness S of the corresponding sound of wave audio data, such as Zwicker are calculated, and reflect voice signal
Ear-piercing degree；
(4) the shake degree F of the corresponding sound of wave audio data, such as Zwicker are calculated, and reflection human ear is to sound
The supervisor of the bright variation degree of sound equipment experiences；
(5) the roughness R of the corresponding sound of wave audio data, such as Zwicker are calculated, and reflect voice signal
The degree of modulation.
Since above-mentioned parameter can influence evaluation or identification of the related personnel to sound quality, sound quality identification device
100 may be configured to obtain evaluation condition of at least one of the above-mentioned parameter as identification sound quality.In order to obtain more
Identification accuracy well can also obtain above-mentioned all parameters.
Then, identification module 102 can determine the sound letter according to the acoustic information and sound quality identification model
The sound quality of breath.For example, the identification module 102 can be using the acoustic information as the defeated of the sound quality identification model
Enter, runs the sound quality identification model；And determine that the output of the sound quality identification model is believed as the sound later
The recognition result of the sound quality of breath.
Using the device described in embodiment 2, can in real time, rapidly, accurately, effectively in the environment of acquisition
Acoustic information carries out sound quality evaluation or identification, and realization objectifies the result of subjective assessment, greatly reduces related personnel's
Working time and job costs have saved time cost and labour cost, significantly while improving sound quality and evaluating accuracy
The flow evaluated in the past Sound Environment Quality is simplified, allows the intuitive Real-time Feedback of sound quality evaluation result to measurement
Personnel can greatly improve working efficiency.The device can be used in the acoustic environment in need evaluated, not by space
With the limitation of environment.
Embodiment 3
In the embodiment 3, in order to realize the purpose of the embodiment of the present invention 2, a kind of example sound quality identification model is provided
300, sound quality identification device 100 can pre-establish sound quality identification model 300.Fig. 3 is a kind of implementation according to the present invention
The structural schematic diagram of the example sound quality identification model of mode, as shown in figure 3, the sound quality identification model 300 can basis
Following steps are established：
(1) the step of establishing training sample set 301 obtains multiple sample audio information in the environment in this step
And the corresponding sample sound quality information of the multiple sample audio information.For example, the sample audio information may include with
Under at least one：Sample audio audio data, the loudness N of the corresponding sound of sample audio audio data, sample audio audio
The sharpness S of the corresponding sound of data, the shake degree F and sample audio audio of the corresponding sound of sample audio audio data
The roughness R of the corresponding sound of data；And the sound quality of the sample audio information may include sample sound quality grade and
At least one of sample sound quality evaluation of estimate.
For example, sample audio audio data can be by voice signal carry out sampling processing after, be cut into 5s durations
The sample of signal of (international standard).These sample of signal are subjected to sound quality evaluation, on the one hand bring relevant software for calculation into, into
Row parameter objective computation, i.e. one or more of its corresponding loudness N, sharpness S, shake degree F, roughness R；On the other hand
Related personnel is organized to carry out subjective assessment, such as sample sound quality grade (basic, normal, high to assign different numerical value respectively) and sample
Sound quality evaluation of estimate (is graded), is later preserved the sound quality evaluation result of these samples, the training sample as neural network
This collection.For example, sample audio information Xi1 ... Xin as shown in Figure 3 and corresponding sample sound quality information Yi1 ... Yin,
Middle i=1 ..., m, i, m, n are positive integer.
(2) the step of establishing neural network model 302, in this step, using the multiple sample audio information as god
Input Xi1 ... Xin through network model, corresponding to the institute of each sample audio information in the multiple sample audio information
The corresponding output Yi1 ... Yin that sample sound quality information is stated as neural network model is trained the neural network model,
To generate the sound quality identification model.
Preferably, the neural network model uses BP (Back-propagation) neural network algorithm, BP nerve nets
Network is a kind of Multi-layered Feedforward Networks model, using the network mould of error (difference of reality output and desired output) back-propagation algorithm
Type.BP learning algorithms adjust weights by using gradient steepest descent method, and the final overall error for realizing network is minimum.Process packet
Include the forward-propagating of signal and two processes of backpropagation of error.During the forward-propagating of signal, network structure, weights and
Threshold value has all been provided signal after input layer entrance, is transmitted to each hidden layer and is received, and is transmitted to after treatment each
A output layer then enters the backpropagation of error when the output signal of output layer and idea output are there are when gap；Error
In back-propagation process, error signal successively propagates back to input layer again from output layer by certain form by hidden layer,
Calculate the influence of the weights and threshold value of each layer to whole overall error successively forward by final layer during error is successively propagated
Thus gradient changes the weights and threshold value of preceding layer.Two processes of forward-propagating and backpropagation constantly alternately and repeatedly carry out,
Until being finally reached network convergence, to make BP neural network reality output and desired output gradually approach, final realization is smaller
Error, ideal output.
However, the learning algorithm due to BP neural network comes with some shortcomings, for example the training speed of BP networks is slow,
Global search capability is not rapid, is always absorbed in local minimum value etc..In order to advanced optimize above-mentioned model, as a kind of excellent
The embodiment of choosing, may be used GA (Genetic Algorithm) genetic algorithms to its (such as to network structure, each layer weigh
Value and threshold value) it optimizes, genetic algorithm is mimic biology evolutionism and the optimization algorithm model that genetics principle obtains, it
It is simple with algorithm, easy the advantages of finding global optimal solution and implicit parallel processing.Its main feature is that directly to operation object number
According to being handled, gradient, successional prerequisite are required no knowledge about, thus possesses the energy of preferably overall selection optimal data
Power and good robustness.The GA-BP neural network prediction models that artificial neural network and genetic algorithms are combined are established, something lost is utilized
Propagation algorithm carrys out Optimized BP Neural Network, may comprise steps of：
A. n different structures are randomly generated, each structure is encoded；
B. the structure in population is trained with different initial weights with initial weight；
C. the adaptability of individual is determined；
D. the excellent individual of selection fitness enters next-generation；
E. group is intersected and is made a variation, generate the population of a new generation；
F. the process for repeating step b-e, until network structure is met the requirements.
Later, established neural network model can be verified, if neural network model does not reach expected accurate
Property, then the re -training model, i.e. cycle execute above-mentioned (1) (2) step, until model reaches expected accuracy.
Using the embodiment 3, sound quality identification model 300 can be established, is provided for sound quality exact evaluation or identification
Model basis.
Embodiment 4
Fig. 4 is a kind of structural schematic diagram of the example sound quality identification device of embodiment according to the present invention, such as Fig. 4 institutes
Show, the device in addition to including in above-described embodiment 2 acquisition module 101 and identification module 102 other than, can also include：Show mould
Block 103, the recognition result of the sound quality for showing the acoustic information, such as display module 103 can be display screen etc..
Specifically, display module 103 can be shown according to the sound quality identified in above-described embodiment 1-3 as a result, for example
Sound quality grade (such as basic, normal, high grade) and/or sound quality evaluation of estimate (such as fractional value).
Using the device described in embodiment 4, it is filled with the blank of existing sound quality subjective assessment real-time visual, is simplified
The flow evaluated in the past Sound Environment Quality allows the intuitive Real-time Feedback of sound quality evaluation result to survey crew,
Working efficiency can be greatly improved.The device can be used in the acoustic environment in need evaluated, not by space and ring
The limitation (ambient temperature has exceeded except the very extreme environment of normal use temperature etc.) in border.
Embodiment 5
Fig. 5 is a kind of structural schematic diagram of the example sound quality identifying system 1000 of embodiment according to the present invention, such as
Shown in Fig. 5, a kind of example system is provided in the embodiment 5, the voice collection device is (wear-type) ears microphone
200a can restore the authenticity for the signal that human ear is heard and simulate the sound that human ear is heard in true environment to the greatest extent
Quality, to improve the accuracy of data source.
Then, voice collection device 200 can send collected acoustic information to sound quality identification device 100, example
As shown in figure 5, the sound quality identification device 100 can be designed as to the equipment 100a in Fig. 5.Sound quality identification device 100
The sound quality of acoustic information can be determined with the configuration of 1-4 according to the abovementioned embodiments of the present invention later.The sound quality can be
Sound quality grade and/or sound quality evaluation of estimate.More intelligently, equipment 100a may include display screen, can on the display screen
To show sound quality grade (such as basic, normal, high grade) and/or evaluation of estimate (such as fractional value).
Using the system described in embodiment 5, it is filled with the blank of existing sound quality subjective assessment real-time visual, is simplified
The flow evaluated in the past Sound Environment Quality allows the intuitive Real-time Feedback of sound quality evaluation result to survey crew,
Working efficiency can be greatly improved.The device can be used in the acoustic environment in need evaluated, not by space and ring
The limitation in border.
Embodiment 6
Fig. 6 is a kind of flow chart of the example sound quality recognition methods of embodiment according to the present invention, as shown in fig. 6,
This method may comprise steps of：
Step S11 obtains the acoustic information in environment；And
Step S12 determines the sound quality of the acoustic information according to the acoustic information and sound quality identification model.
Preferably, the sound quality identification model is established according to following steps：Obtain multiple sample sound in the environment
Message ceases and the corresponding sample sound quality information of the multiple sample audio information；And by the multiple sample audio information
As the input of neural network model, corresponding to described in each sample audio information in the multiple sample audio information
Sample sound quality information is trained the neural network model as the corresponding output of neural network model, described in generation
Sound quality identification model.
Preferably, the neural network model is established using BP neural network algorithm.
Preferably, the neural network model is established using the improved BP neural network algorithm of Genetic Algorithms.
Preferably, the sound quality that the acoustic information is determined according to the acoustic information and sound quality identification model
Including：Using the acoustic information as the input of the sound quality identification model, the sound quality identification model is run；And really
The recognition result of sound quality of the output of the fixed sound quality identification model as the acoustic information.
Preferably, the acoustic information include it is below at least one：Wave audio data, wave audio data are corresponding
The loudness of sound, the sharpness of the corresponding sound of wave audio data, the corresponding sound of wave audio data shake degree and
The roughness of the corresponding sound of wave audio data；The sample audio information include it is below at least one：Sample audio sound
Sharpness, the sample of frequency evidence, the loudness of the corresponding sound of sample audio audio data, the corresponding sound of sample audio audio data
The shake degree of the corresponding sound of this wave audio data and the roughness of the corresponding sound of sample audio audio data；And
The sound quality of the acoustic information includes at least one of sound quality grade harmony quality evaluation value；The sample audio information
Sound quality include at least one of sample sound quality grade and sample sound quality evaluation of estimate.
It should be understood that each specific embodiment of above-mentioned sound quality recognition methods, identifies in example sound quality
Device, system embodiment in done and explain (as described above) in detail, details are not described herein.
Sound quality identification device provided by the invention can realize in the form of hardware or software, such as can be with software
Form is applied in any scene appropriate that sound quality is evaluated or identified, can also in the form of hardware, this
Invention is to this without limiting.Those skilled in the art can be with the open above-mentioned various embodiments of selection according to the ... of the embodiment of the present invention
Any one of, or the combination of above-mentioned various embodiments is selected to configure sound quality identification device and system, and it is other
Alternative embodiment also falls into the protection domain of the embodiment of the present invention.
In addition, the present invention also provides a kind of vehicle including the sound quality identifying system described in above-described embodiment 1-6, the vehicle
Above-mentioned sound quality identifying system or device can be utilized to carry out sound quality evaluation or identification.
The preferred embodiment of the present invention is described in detail above in association with attached drawing, still, the present invention is not limited thereto.At this
In the range of the technology design of invention, a variety of simple variants can be carried out to technical scheme of the present invention, for example, each particular technique
Feature is combined in any suitable manner, and in order to avoid unnecessary repetition, the present invention is to various combinations of possible ways
No longer separately illustrate.But it should also be regarded as the disclosure of the present invention for these simple variants and combination, belongs to the present invention
Protection domain.
Claims (15)
1. a kind of sound quality recognition methods, which is characterized in that this method includes：
Obtain the acoustic information in environment；And
According to the acoustic information and sound quality identification model, the sound quality of the acoustic information is determined.
2. according to the method described in claim 1, it is characterized in that, the sound quality identification model is established according to following steps：
Obtain the multiple sample audio information and the corresponding sample sound quality of the multiple sample audio information in the environment
Information；And
Using the multiple sample audio information as the input of neural network model, correspond in the multiple sample audio information
Each sample audio information the sample sound quality information as neural network model corresponding output to the nerve
Network model is trained, to generate the sound quality identification model.
3. according to the method described in claim 2, it is characterized in that, the neural network model is built using BP neural network algorithm
It is vertical.
4. according to the method described in claim 2, it is characterized in that, the neural network model is improved using Genetic Algorithms
BP neural network algorithm is established.
5. method according to claim 3 or 4, which is characterized in that described to be known according to the acoustic information and sound quality
Other model determines that the sound quality of the acoustic information includes：
Using the acoustic information as the input of the sound quality identification model, the sound quality identification model is run；And
Determine the recognition result for exporting the sound quality as the acoustic information of the sound quality identification model.
6. according to the method described in claim 5, it is characterized in that, the acoustic information include it is below at least one：Sound
Sharpness, the wave audio of audio data, the loudness of the corresponding sound of wave audio data, the corresponding sound of wave audio data
The shake degree of the corresponding sound of data and the roughness of the corresponding sound of wave audio data；The sample audio packet
Include it is below at least one：Sample audio audio data, the loudness of the corresponding sound of sample audio audio data, sample audio sound
Frequency is according to the sharpness of corresponding sound, the shake degree and sample audio audio of the corresponding sound of sample audio audio data
The roughness of the corresponding sound of data；And
The sound quality of the acoustic information includes at least one of sound quality grade harmony quality evaluation value；The sample audio
The sound quality of information includes at least one of sample sound quality grade and sample sound quality evaluation of estimate.
7. a kind of sound quality identification device, which is characterized in that the device includes：
Acquisition module, for obtaining the acoustic information in environment；And
Identification module, the sound quality for determining the acoustic information according to the acoustic information and sound quality identification model.
8. device according to claim 7, which is characterized in that the sound quality identification model is established according to following steps：
Obtain the multiple sample audio information and the corresponding sample sound quality of the multiple sample audio information in the environment
Information；And
Using the multiple sample audio information as the input of neural network model, correspond in the multiple sample audio information
Each sample audio information the sample sound quality information as neural network model corresponding output to the nerve
Network model is trained, to generate the sound quality identification model.
9. device according to claim 8, which is characterized in that the neural network model using BP neural network algorithm or
The improved BP neural network algorithm of person's Genetic Algorithms is established.
10. device according to claim 9, which is characterized in that the identification module is additionally operable to：The acoustic information is made
For the input of the sound quality identification model, the sound quality identification model is run；And determine the sound quality identification model
Output as the acoustic information sound quality recognition result.
11. device according to claim 10, which is characterized in that the device further includes：Display module, it is described for showing
The recognition result of the sound quality of acoustic information.
12. according to the devices described in claim 11, which is characterized in that the acoustic information include it is below at least one：Sound
Sound audio data, the loudness of the corresponding sound of wave audio data, the sharpness of the corresponding sound of wave audio data, sound sound
Frequency is according to the shake degree of corresponding sound and the roughness of the corresponding sound of wave audio data；The sample audio information
Including it is below at least one：Sample audio audio data, the loudness of the corresponding sound of sample audio audio data, sample audio
The sharpness of the corresponding sound of audio data, the shake degree and sample audio sound of the corresponding sound of sample audio audio data
Frequency according to corresponding sound roughness；And
The sound quality of the acoustic information includes at least one of sound quality grade harmony quality evaluation value；The sample audio
The sound quality of information includes at least one of sample sound quality grade and sample sound quality evaluation of estimate.
13. a kind of sound quality identifying system, which is characterized in that the system includes：
Voice collection device, for acquiring the acoustic information in environment；
According to the sound quality identification device described in any one of claim 7-12 claims, the sound quality identification device with
The voice collection device connection.
14. system according to claim 13, which is characterized in that the voice collection device is ears microphone.
15. a kind of vehicle including the sound quality identifying system described in claim 13 or 14.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710165978.7A CN108630228A (en) | 2017-03-20 | 2017-03-20 | Sound quality recognition methods, device, system and vehicle |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710165978.7A CN108630228A (en) | 2017-03-20 | 2017-03-20 | Sound quality recognition methods, device, system and vehicle |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108630228A true CN108630228A (en) | 2018-10-09 |
Family
ID=63687616
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710165978.7A Pending CN108630228A (en) | 2017-03-20 | 2017-03-20 | Sound quality recognition methods, device, system and vehicle |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108630228A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110610723A (en) * | 2019-09-20 | 2019-12-24 | 中国第一汽车股份有限公司 | Method, device, equipment and storage medium for evaluating sound quality in vehicle |
| CN111076244A (en) * | 2018-10-19 | 2020-04-28 | 宁波方太厨具有限公司 | Method for diagnosing sound quality of range hood |
| CN111737277A (en) * | 2020-07-28 | 2020-10-02 | 广州汽车集团股份有限公司 | Sample data processing method, device, computer equipment and storage medium |
| CN112798102A (en) * | 2021-04-13 | 2021-05-14 | 盛瑞传动股份有限公司 | Sound quality evaluation method, device, system and medium |
| CN113782053A (en) * | 2021-09-04 | 2021-12-10 | 天津大学 | Urban sound landscape quality automatic monitoring method worthy of protection |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101599138A (en) * | 2009-07-07 | 2009-12-09 | 武汉大学 | Land evaluation method based on artificial neural network |
| CN101915753A (en) * | 2010-07-30 | 2010-12-15 | 浙江师范大学 | Quantitative Analysis Method of Laser Induced Breakdown Spectroscopy Based on Genetic Neural Network |
| CN102475554A (en) * | 2010-11-24 | 2012-05-30 | 比亚迪股份有限公司 | Method for guiding acoustic package in vehicle by using sound quality |
| CN103471709A (en) * | 2013-09-17 | 2013-12-25 | 吉林大学 | Method for predicting noise quality of noise inside passenger vehicle |
| CN103971162A (en) * | 2014-04-04 | 2014-08-06 | 华南理工大学 | Method for improving BP (back propagation) neutral network and based on genetic algorithm |
| CN104102923A (en) * | 2014-07-16 | 2014-10-15 | 西安建筑科技大学 | Nipponia nippon individual recognition method based on MFCC algorithm |
| CN104678988A (en) * | 2014-10-28 | 2015-06-03 | 芜湖杰诺瑞汽车电器系统有限公司 | Engine ECU (electronic control unit) circuit fault diagnosis method |
-
2017
- 2017-03-20 CN CN201710165978.7A patent/CN108630228A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101599138A (en) * | 2009-07-07 | 2009-12-09 | 武汉大学 | Land evaluation method based on artificial neural network |
| CN101915753A (en) * | 2010-07-30 | 2010-12-15 | 浙江师范大学 | Quantitative Analysis Method of Laser Induced Breakdown Spectroscopy Based on Genetic Neural Network |
| CN102475554A (en) * | 2010-11-24 | 2012-05-30 | 比亚迪股份有限公司 | Method for guiding acoustic package in vehicle by using sound quality |
| CN103471709A (en) * | 2013-09-17 | 2013-12-25 | 吉林大学 | Method for predicting noise quality of noise inside passenger vehicle |
| CN103971162A (en) * | 2014-04-04 | 2014-08-06 | 华南理工大学 | Method for improving BP (back propagation) neutral network and based on genetic algorithm |
| CN104102923A (en) * | 2014-07-16 | 2014-10-15 | 西安建筑科技大学 | Nipponia nippon individual recognition method based on MFCC algorithm |
| CN104678988A (en) * | 2014-10-28 | 2015-06-03 | 芜湖杰诺瑞汽车电器系统有限公司 | Engine ECU (electronic control unit) circuit fault diagnosis method |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111076244A (en) * | 2018-10-19 | 2020-04-28 | 宁波方太厨具有限公司 | Method for diagnosing sound quality of range hood |
| CN110610723A (en) * | 2019-09-20 | 2019-12-24 | 中国第一汽车股份有限公司 | Method, device, equipment and storage medium for evaluating sound quality in vehicle |
| CN110610723B (en) * | 2019-09-20 | 2022-02-22 | 中国第一汽车股份有限公司 | Method, device, equipment and storage medium for evaluating sound quality in vehicle |
| CN111737277A (en) * | 2020-07-28 | 2020-10-02 | 广州汽车集团股份有限公司 | Sample data processing method, device, computer equipment and storage medium |
| CN112798102A (en) * | 2021-04-13 | 2021-05-14 | 盛瑞传动股份有限公司 | Sound quality evaluation method, device, system and medium |
| CN113782053A (en) * | 2021-09-04 | 2021-12-10 | 天津大学 | Urban sound landscape quality automatic monitoring method worthy of protection |
| CN113782053B (en) * | 2021-09-04 | 2023-09-22 | 天津大学 | Automatic monitoring method for urban soundscape quality worthy of protection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108630228A (en) | Sound quality recognition methods, device, system and vehicle | |
| JP4005128B2 (en) | Signal quality evaluation | |
| CN112183643B (en) | Method and device for identifying tensile shear fracture of hard rock based on acoustic emission | |
| CN103471709A (en) | Method for predicting noise quality of noise inside passenger vehicle | |
| CN108596016B (en) | Personalized head-related transfer function modeling method based on deep neural network | |
| CN109409308A (en) | A method of the birds species identification based on birdvocalization | |
| CN112161815A (en) | Vehicle road noise subjective evaluation value prediction method | |
| CN109190943A (en) | Dynamic Fire risk assessment method, device and server based on machine learning | |
| CN107527626A (en) | Audio identification system | |
| CN102204319A (en) | Method and device for artificial synthesis of network flow | |
| CN117041847B (en) | Adaptive microphone matching method and system for hearing aid | |
| CN113470475B (en) | Real-operation learning assessment method and system based on scene simulation and Internet of things | |
| CN114760586A (en) | Indoor positioning method, device and system | |
| CN112668105A (en) | Helicopter transmission shaft abnormity judgment method based on SAE and Mahalanobis distance | |
| CN101727896B (en) | An Objective Speech Quality Estimation Method Based on Perceptual Parameters | |
| CN109933933B (en) | Noise treatment method and equipment | |
| CN118566998B (en) | Termite pest monitoring method and monitoring system for protecting buildings | |
| CN112348656A (en) | BA-WNN-based personal loan credit scoring method | |
| Jin et al. | Polyphonic sound event detection using capsule neural network on multi-type-multi-scale time-frequency representation | |
| JP2009525792A (en) | Method, apparatus and computer program for generating control signals for cochlear implants based on audio signals | |
| CN117974736A (en) | Underwater sensor output signal noise reduction method and system based on machine learning | |
| KR102295993B1 (en) | Method of reverberation time estimation using statistic of speech spectral decay rates | |
| CN114091131A (en) | Optimal scheme decision system for cloth mask support design based on cloud platform | |
| CN114358321A (en) | Machine learning detection method for abnormal sound of motor | |
| Sadeghian et al. | The use of artificial neural networks to predict tonal sound annoyance based on noise metrics and psychoacoustics parameters |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181009 |