CN105808719B - Audio-frequency information recommended method and device - Google Patents

Audio-frequency information recommended method and device Download PDF

Info

Publication number
CN105808719B
CN105808719B CN201610128170.7A CN201610128170A CN105808719B CN 105808719 B CN105808719 B CN 105808719B CN 201610128170 A CN201610128170 A CN 201610128170A CN 105808719 B CN105808719 B CN 105808719B
Authority
CN
China
Prior art keywords
audio file
target audio
recommendation
quality measurements
target
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
Application number
CN201610128170.7A
Other languages
Chinese (zh)
Other versions
CN105808719A (en
Inventor
雷波
张超论
丁亮
陈容
陈传艺
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangzhou Kugou Computer Technology Co Ltd
Original Assignee
Guangzhou Kugou Computer Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangzhou Kugou Computer Technology Co Ltd filed Critical Guangzhou Kugou Computer Technology Co Ltd
Priority to CN201610128170.7A priority Critical patent/CN105808719B/en
Publication of CN105808719A publication Critical patent/CN105808719A/en
Application granted granted Critical
Publication of CN105808719B publication Critical patent/CN105808719B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/63—Querying
    • G06F16/635—Filtering based on additional data, e.g. user or group profiles
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
    • G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/683—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content

Abstract

The invention discloses a kind of audio-frequency information recommended method and devices, belong to network technique field.This method comprises: carrying out audio quality detection to target audio file, obtain the first quality measurements, first quality measurements are that there is no defect or existing defects, and when the first quality measurements are existing defects, it further include the defect type of target audio file in first quality measurements, target audio file is the audio file for having uploaded and not carried out audio quality detection and recommendation；It is searched from the recommendation of audio library of storage and the matched recommendation of audio file of target audio file and the second quality measurements, the second quality measurements is the quality measurements of recommendation of audio file；If found, the first quality measurements and the second quality measurements are based on, recommendation of audio library is updated；Updated recommendation of audio library is recommended.The present invention improves the accuracy rate of audio quality detection, and then improves the accuracy rate of audio-frequency information recommendation.

Description

Audio-frequency information recommended method and device
Technical field
The present invention relates to network technique field, in particular to a kind of audio-frequency information recommended method and device.
Background technique
With the fast development of network technology, there are more and more audio playing platforms.It can in audio playing platform Uploaded to the audio-frequency information of the audio playing platform to show, the audio file of these audio-frequency informations can be divided into it is multiple not With sound quality type may have multiple audios in the same sound quality type of the audio-frequency information and for same audio-frequency information File.For example, may have multiple audios for belonging to same sound quality type in the audio playing platform for same song File.When user searches for a certain audio-frequency information, it is difficult to rapidly select one from multiple audio files of same sound quality type Therefore a preferable audio file of sound quality needs a kind of audio-frequency information recommended method.
Currently, providing a kind of audio-frequency information recommended method, comprising: when receiving target audio file, judge target Whether the sound quality type of audio file is lossless type, and target audio file is to have uploaded and do not carried out defect dipoles and recommended Audio file.When the sound quality type of target audio file is not lossless type, the code rate of target audio file is obtained, and by mesh The code rate of mark with phonetic symbols frequency file is compared with specified range of code rates.When the code rate of target audio file be located at specified range of code rates it When interior, determining target audio file, there is no defects.It obtains from the recommendation of audio library of storage and belongs to together with target audio file One sound quality type and the recommendation of audio file for belonging to same audio-frequency information, if the recommendation of audio file existing defects obtained, The recommendation of audio file that will acquire replaces with target audio file, and target audio file is recommended.
In the implementation of the present invention, the inventor finds that the existing technology has at least the following problems:
Above-mentioned technology only judges target audio file with the presence or absence of defect simply by code rate, and accuracy rate is lower, And for the audio file of lossless type may also can existing defects, therefore, above-mentioned technology reduce audio-frequency information recommendation Accuracy rate.
Summary of the invention
In order to solve problems in the prior art, the embodiment of the invention provides a kind of audio-frequency information recommended method and devices. The technical solution is as follows:
On the one hand, a kind of audio-frequency information recommended method is provided, which comprises
Audio quality detection is carried out to target audio file, obtains the first quality measurements, first quality testing As a result for there is no defect or existing defects, and when first quality measurements are existing defects, first matter It further include the defect type of the target audio file in amount testing result, the target audio file is to have uploaded and do not carried out Audio file audio quality detection and recommended；
It is searched and the matched recommendation of audio file of the target audio file and the second matter from the recommendation of audio library of storage Testing result is measured, second quality measurements are the quality measurements of the recommendation of audio file；
If found, first quality measurements and second quality measurements are based on, are pushed away to described Audio repository is recommended to be updated；
Updated recommendation of audio library is recommended.
Optionally, described that audio quality detection is carried out to target audio file, the first quality measurements are obtained, including such as At least one of under type:
The target audio file is decoded by specified decoder, if target audio file decoding is lost It loses, it is determined that the target audio file existing defects and defect type are processing failure；
The target audio file is decoded by the specified decoder, if the target audio file decodes Success, then the target audio data obtained based on decoding determine the practical duration of the target audio file, if the reality Duration is less than or equal to specified duration, it is determined that the target audio file existing defects and length when defect type is；
Based on the tonepulse of the target audio file, judge in the target audio file with the presence or absence of instantaneous arteries and veins Punching, if there are transient pulses in the target audio file, it is determined that the target audio file existing defects and defect class Type is current sound；
It determines the practical cut frequency of the target audio file and refers to cut frequency, if the practical cut frequency It is greater than or equal to specified cut frequency with the difference with reference between cut frequency, it is determined that the target audio file exists Defect and defect type are sound quality code stream mismatch；
Based on the tonepulse of the target audio file, judge in the target audio file with the presence or absence of instantaneous quiet Sound, if existed in the target audio file instantaneous mute, it is determined that the target audio file existing defects and defect class Type is cassette tape；
The correlation and capacity volume variance for determining target audio file left and right acoustic channels when playing, if the correlation It is greater than or equal to specified difference value less than or equal to specified correlation or the capacity volume variance, it is determined that the target audio text Part existing defects and defect type are left and right acoustic channels Type-Inconsistencies.
Optionally, the tonepulse based on the target audio file, judge in the target audio file whether There are transient pulses, comprising:
The first pulse amplitude, the second pulse amplitude and third pulse are obtained from the tonepulse of the target audio file Amplitude, first pulse amplitude are the pulse amplitude in the tonepulse at the first sampled point, second pulse amplitude For the pulse amplitude at the second sampled point in the tonepulse, the third pulse amplitude is that third is adopted in the tonepulse Pulse amplitude at sampling point, first sample are and described between second sampled point and the third sampled point Second sampled point and the third sampled point are adjacent with first sampled point；
First pulse amplitude is subtracted into second pulse amplitude, obtains the first difference, by the first pulse width Value subtracts the third pulse amplitude, obtains the second difference；
When first difference is more than or equal to the first specified amplitude and second difference is greater than or equal to described first When specified amplitude, determine that there are transient pulses in the target audio file.
Optionally, the practical cut frequency of the determination target audio file and refer to cut frequency, comprising:
The cut frequency of multiframe audio data in the target audio file, and the statistics target audio are counted respectively The code rate of file；
The practical cut frequency of the target audio file is determined based on the cut frequency of the multiframe audio data；
The reference cut frequency of the target audio file is determined based on the code rate of the target audio file.
Optionally, the tonepulse based on the target audio file, judge in the target audio file whether In the presence of instantaneous mute, comprising:
The 4th pulse amplitude is obtained from the tonepulse of the target audio file, the 4th pulse amplitude is described Pulse amplitude in tonepulse at any sampled point；
When the 4th pulse amplitude is less than or equal to the second specified amplitude, determine in the target audio file In the presence of instantaneous mute.
Optionally, described to be based on first quality measurements and second quality measurements, to the recommendation Audio repository is updated, comprising:
Based on first quality measurements and second quality measurements, determine the target audio file with Priority relationship between the recommendation of audio file；
When the priority of the target audio file is higher than the priority of the recommendation of audio file, by the recommendation sound Frequency file replaces with the target audio file, to be updated to the recommendation of audio library.
Optionally, described to be based on first quality measurements and second quality measurements, determine the mesh Priority relationship between mark with phonetic symbols frequency file and the recommendation of audio file, comprising:
When first quality measurements are there is no when defect, if second quality measurements are to exist to lack It falls into, it is determined that the priority of the target audio file is higher than the priority of the recommendation of audio file, if second matter Measuring testing result is that there is no defects, it is determined that the priority of the target audio file is preferential with the recommendation of audio file Grade is identical；
When first quality measurements are existing defects, if second quality measurements are that there is no lack It falls into, it is determined that the priority of the target audio file is lower than the priority of the recommendation of audio file, if second matter Amount testing result is existing defects, then based on the defects of first quality measurements type and second quality testing The defects of as a result type, determines the priority relationship of the target audio file and the recommendation of audio file.
Optionally, the defect type for including based on first quality measurements and the second quality testing knot The defect type that fruit includes determines the priority relationship of the target audio file and the recommendation of audio file, comprising:
Determine that first number and second number, first number are defect type in first quality measurements Number, second number are defect type number in second quality measurements；
Judge whether first number is greater than second number；
If it is, determine that the priority of the target audio file is lower than the priority of the recommendation of audio file, it is no Then, determine that the priority of the target audio file is higher than the priority of the recommendation of audio file.
On the other hand, a kind of audio-frequency information recommendation apparatus is provided, described device includes:
Audio quality detection module obtains the first quality testing for carrying out audio quality detection to target audio file As a result, first quality measurements are that there is no defect or existing defects, and work as first quality measurements and be It further include the defect type of the target audio file, the target sound when existing defects, in first quality measurements Frequency file is the audio file for having uploaded and not carried out audio quality detection and recommendation；
Searching module, for being searched and the matched recommendation of audio of the target audio file from the recommendation of audio library of storage File and the second quality measurements, second quality measurements are the quality measurements of the recommendation of audio file；
Update module, if being based on first quality measurements and second quality testing for finding As a result, being updated to the recommendation of audio library；
Recommending module, for recommending updated recommendation of audio library.
Optionally, the audio quality detection module includes such as at least one of lower unit:
First determination unit, for being decoded by specified decoder to the target audio file, if the mesh Mark with phonetic symbols frequency file decoding failure, it is determined that the target audio file existing defects and defect type are processing failure；
Second determination unit, for being decoded by the specified decoder to the target audio file, if institute Target audio file successfully decoded is stated, then the target audio data obtained based on decoding determine the reality of the target audio file Duration, if the practical duration is less than or equal to specified duration, it is determined that the target audio file existing defects and defect Length when type is；
Third determination unit judges the target audio file for the tonepulse based on the target audio file In whether there is transient pulse, if there are transient pulses in the target audio file, it is determined that the target audio file Existing defects and defect type are current sound；
4th determination unit, for determining the practical cut frequency of the target audio file and with reference to cut frequency, such as Practical cut frequency described in fruit and the difference with reference between cut frequency are greater than or equal to specified cut frequency, it is determined that institute Target audio file existing defects and defect type are stated as sound quality code stream mismatch；
5th determination unit judges the target audio file for the tonepulse based on the target audio file In with the presence or absence of instantaneous mute, if existed in the target audio file instantaneous mute, it is determined that the target audio file Existing defects and defect type are cassette tape；
6th determination unit, for determining the correlation and energy difference of target audio file left and right acoustic channels when playing It is different, if the correlation is less than or equal to specified correlation or the capacity volume variance is greater than or equal to specified difference value, It determines the target audio file existing defects and defect type is left and right acoustic channels Type-Inconsistencies.
Optionally, the third determination unit includes:
First obtains subelement, for obtaining the first pulse amplitude, the from the tonepulse of the target audio file Two pulse amplitudes and third pulse amplitude, first pulse amplitude are the pulse width in the tonepulse at the first sampled point Value, second pulse amplitude are the pulse amplitude in the tonepulse at the second sampled point, and the third pulse amplitude is Pulse amplitude in the tonepulse at third sampled point, first sample is in second sampled point and described Between three sampled points, and second sampled point and the third sampled point are adjacent with first sampled point；
Computation subunit obtains the first difference, incites somebody to action for first pulse amplitude to be subtracted second pulse amplitude First pulse amplitude subtracts the third pulse amplitude, obtains the second difference；
First determines subelement, for being greater than or equal to the first specified amplitude and second difference when first difference When more than or equal to the first specified amplitude, determine that there are transient pulses in the target audio file；
Second determines subelement, if for there are transient pulses in the target audio file, it is determined that the target Audio file existing defects and defect type are current sound.
Optionally, the 4th determination unit includes:
Subelement is counted, for counting the cut frequency of multiframe audio data in the target audio file respectively, and Count the code rate of the target audio file；
Third determines subelement, for determining the target audio file based on the cut frequency of the multiframe audio data Practical cut frequency；
4th determines subelement, for determining the ginseng of the target audio file based on the code rate of the target audio file Examine cut frequency；
5th determines subelement, if for the practical cut frequency and described big with reference to the difference between cut frequency In or be equal to specified cut frequency, it is determined that the target audio file existing defects and defect type be sound quality code stream not Match.
Optionally, the 5th determination unit includes:
Second obtains subelement, for obtaining the 4th pulse amplitude, institute from the tonepulse of the target audio file Stating the 4th pulse amplitude is the pulse amplitude in the tonepulse at any sampled point；
6th determines subelement, is used for when the 4th pulse amplitude is less than or equal to the second specified amplitude, really Exist in the fixed target audio file instantaneous mute；
7th determines subelement, if instantaneous mute for existing in the target audio file, it is determined that the target Audio file existing defects and defect type are cassette tape.
Optionally, the update module includes:
7th determination unit is determined for being based on first quality measurements and second quality measurements Priority relationship between the target audio file and the recommendation of audio file；
Replacement unit is higher than the priority of the recommendation of audio file for the priority when the target audio file When, the recommendation of audio file is replaced with into the target audio file, to be updated to the recommendation of audio library.
Optionally, the 7th determination unit includes:
8th determines subelement, for when first quality measurements be there is no when defect, if described second Quality measurements are existing defects, it is determined that the priority of the target audio file is higher than the excellent of the recommendation of audio file First grade, if second quality measurements are that there is no defects, it is determined that the priority of the target audio file and institute The priority for stating recommendation of audio file is identical；
9th determines subelement, is used for when first quality measurements are existing defects, if second matter Measuring testing result is that there is no defects, it is determined that the priority of the target audio file is excellent lower than the recommendation of audio file First grade, if second quality measurements are existing defects, based on the defects of first quality measurements class The defects of type and second quality measurements type determine the target audio file and the recommendation of audio file Priority relationship.
Optionally, it the described 9th determines subelement, is specifically used for:
When first quality measurements are existing defects, if second quality measurements are that there is no lack It falls into, it is determined that the priority of the target audio file is lower than the priority of the recommendation of audio file；
If second quality measurements are existing defects, it is determined that first number and second number, described first Number is defect type number in first quality measurements, and second number is in second quality measurements Defect type number；
Judge whether first number is greater than second number；
If it is, determine that the priority of the target audio file is lower than the priority of the recommendation of audio file, it is no Then, determine that the priority of the target audio file is higher than the priority of the recommendation of audio file.
Technical solution provided in an embodiment of the present invention has the benefit that in embodiments of the present invention, to target sound Frequency file carry out audio quality detection, obtain the first quality measurements, due to the first quality measurements can for there is no Defect or there are disadvantages, and when the first quality measurements are existing defects, can also include in the first quality measurements The defect type of target audio file, that is to say, when the embodiment of the present invention carries out audio quality detection to target audio file, no It can only determine that target audio file with the presence or absence of defect, can also determine target sound in target audio file existing defects The defect type of frequency file, therefore the accuracy rate for carrying out quality testing to target audio file is higher.In addition, the embodiment of the present invention It is that audio quality detection is carried out to all audio files for having uploaded and not carried out audio quality detection and recommendation, wherein just including The audio file of lossless type, therefore, the embodiment of the present invention can be further improved the accuracy rate of audio quality detection, Jin Erti The accuracy rate of high audio information recommendation.
Detailed description of the invention
To describe the technical solutions in the embodiments of the present invention more clearly, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other Attached drawing.
Fig. 1 is a kind of audio-frequency information recommended method flow chart provided in an embodiment of the present invention；
Fig. 2 is another audio-frequency information recommended method flow chart provided in an embodiment of the present invention；
Fig. 3 is a kind of structural schematic diagram of audio-frequency information recommendation apparatus provided in an embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with attached drawing to embodiment party of the present invention Formula is described in further detail.
Before to the embodiment of the present invention carrying out that explanation is explained in detail, first the application scenarios of the embodiment of the present invention are given It introduces.Due to that can show the audio-frequency information uploaded in audio playing platform, and then user can be from the audio playing platform The audio file of audio-frequency information needed for upper search, to carry out audio broadcasting, and for same audio-frequency information, audio playing platform In may include the audio file of multiple and different sound quality types, for example, can wrap in audio playing platform for same song Include the audio file of multiple sound quality types such as lossless, smooth, standard.But the sound of the audio file uploaded in audio playing platform Matter type may be by the lower audio file of acoustical quality and is converted to, and so works as and is broadcast by the higher speaker of quality Corresponding effect may be not achieved when putting, for example, the song of the standard sound quality uploaded in audio playing platform is by smooth sound The song of matter is converted to, and when playing the song by the higher speaker of quality, the effect of standard sound quality may be not achieved Fruit.Therefore, in audio file there may be in the case where drawbacks described above, when passing through audio playing platform recommendation of audio information, It needs to judge that audio file whether there is defect, and then determines the higher audio of quality from the audio file of same sound quality type File is recommended.The prior art only judges target audio file with the presence or absence of defect, accuracy rate simply by code rate It is lower, and for the audio file of lossless type may also can existing defects, and the prior art is not to the sound of lossless type Frequency file carries out defects detection, and therefore, the embodiment of the invention provides a kind of audio-frequency information recommended methods, and audio text can be improved The defects detection accuracy rate of part, and then improve the accuracy rate that audio-frequency information is recommended.
Fig. 1 is a kind of audio-frequency information recommended method flow chart provided in an embodiment of the present invention.Referring to Fig. 1, this method comprises:
Step 101: audio quality detection being carried out to target audio file, obtains the first quality measurements, the first mass Testing result is that there is no defect or existing defects, and when the first quality measurements are existing defects, the first mass is examined It further include the defect type of target audio file in survey result, target audio file is to have uploaded and do not carried out audio quality detection With the audio file of recommendation.
Step 102: being searched from the recommendation of audio library of storage and the matched recommendation of audio file of target audio file and the Two quality measurements, the second quality measurements are the quality measurements of recommendation of audio file.
Step 103: if found, the first quality measurements and the second quality measurements are based on, to recommendation sound Frequency library is updated.
Step 104: updated recommendation of audio library is recommended.
In embodiments of the present invention, audio quality detection is carried out to target audio file, obtains the first quality measurements, Since the first quality measurements can be for there is no defect or there are disadvantages, and when the first quality measurements are existing defects When, it can also include the defect type of target audio file in the first quality measurements, that is to say, the embodiment of the present invention is to mesh When mark with phonetic symbols frequency file carries out audio quality detection, it can not only determine that target audio file, can also be in mesh with the presence or absence of defect When mark with phonetic symbols frequency file existing defects, the defect type of target audio file is determined, therefore quality inspection is carried out to target audio file The accuracy rate of survey is higher.In addition, the embodiment of the present invention is to all sounds uploaded and do not carry out audio quality detection and recommendation Frequency file carries out audio quality detection, wherein just including the audio file of lossless type, therefore, the embodiment of the present invention can be into one Step improves the accuracy rate of audio quality detection, and then improves the accuracy rate that audio-frequency information is recommended.
Optionally, audio quality detection is carried out to target audio file, obtains the first quality measurements, including such as lower section At least one of formula:
Target audio file is decoded by specified decoder, if target audio file decoding failure, it is determined that Target audio file existing defects and defect type are processing failure；
Target audio file is decoded by specified decoder, if target audio file successfully decoded, is based on The practical duration that obtained target audio data determine target audio file is decoded, if the practical duration of target audio file is small In or be equal to specified duration, it is determined that target audio file existing defects and length when defect type is；
Tonepulse based on target audio file judges with the presence or absence of transient pulse in target audio file, if mesh There are transient pulses in mark with phonetic symbols frequency file, it is determined that target audio file existing defects and defect type are current sound；
It determines the practical cut frequency of target audio file and refers to cut frequency, if the reality of target audio file is cut Difference between disconnected frequency and reference cut frequency is greater than or equal to specified cut frequency, it is determined that target audio file, which exists, to be lacked It falls into and defect type is sound quality code stream mismatch；
Tonepulse based on target audio file judges in target audio file with the presence or absence of instantaneous mute, if mesh Exist in mark with phonetic symbols frequency file instantaneous mute, it is determined that target audio file existing defects and defect type is cassette tape；
Determine the correlation and capacity volume variance of target audio file left and right acoustic channels when playing, if target audio file Correlation is less than or equal to specified correlation or capacity volume variance is greater than or equal to specified difference value, it is determined that target audio file Existing defects and defect type are left and right acoustic channels Type-Inconsistencies.
Optionally, based on the tonepulse of target audio file, judge to whether there is transient pulse in target audio file, Include:
The first pulse amplitude, the second pulse amplitude and third pulse width are obtained from the tonepulse of target audio file Value, the first pulse amplitude are the pulse amplitude in the tonepulse of target audio file at the first sampled point, the second pulse amplitude For the pulse amplitude at the second sampled point in the tonepulse of target audio file, third pulse amplitude is target audio file Pulse amplitude in tonepulse at third sampled point, the first sample between the second sampled point and third sampled point, and Second sampled point and third sampled point are adjacent with the first sampled point；
First pulse amplitude is subtracted into the second pulse amplitude, obtains the first difference, the first pulse amplitude is subtracted into third arteries and veins Amplitude is rushed, the second difference is obtained；
When the first difference is greater than or equal to the first specified amplitude and the second difference is greater than or equal to the first specified amplitude, really Setting the goal in audio file, there are transient pulses.
Optionally it is determined that the practical cut frequency of target audio file and refer to cut frequency, comprising:
The cut frequency of multiframe audio data in target audio file, and the code of statistics target audio file are counted respectively Rate；
The practical cut frequency of target audio file is determined based on the cut frequency of the multiframe audio data；
The reference cut frequency of target audio file is determined based on the code rate of target audio file.
Optionally, based on the tonepulse of target audio file, judge in target audio file with the presence or absence of instantaneous mute, Include:
The 4th pulse amplitude is obtained from the tonepulse of target audio file, the 4th pulse amplitude is target audio file Tonepulse in pulse amplitude at any sampled point；
When the 4th pulse amplitude is less than or equal to the second specified amplitude, determine in target audio file in the presence of instantaneous quiet Sound.
Optionally, the first quality measurements and the second quality measurements are based on, recommendation of audio library is updated, are wrapped It includes:
Based on the first quality measurements and the second quality measurements, target audio file and recommendation of audio file are determined Between priority relationship；
When the priority of target audio file is higher than the priority of recommendation of audio file, recommendation of audio file is replaced with Target audio file, to be updated to recommendation of audio library.
Optionally, the first quality measurements and the second quality measurements are based on, determine target audio file and are recommended Priority relationship between audio file, comprising:
When the first quality measurements are there is no when defect, if the second quality measurements are existing defects, really Set the goal audio file priority be higher than recommendation of audio file priority, if the second quality measurements be there is no lack It falls into, it is determined that the priority of target audio file is identical as the priority of recommendation of audio file；
When the first quality measurements are existing defects, if the second quality measurements are that there is no defects, really Set the goal audio file priority be lower than recommendation of audio file priority, if the second quality measurements be exist lack It falls into, then based on the defects of the first quality measurements the defects of type and the second quality measurements type, determines target The priority relationship of audio file and recommendation of audio file.
Optionally, the defect that the defect type and the second quality measurements for including based on the first quality measurements include Type determines the priority relationship of target audio file and recommendation of audio file, comprising:
Determining first number and second number, first number is defect type number in the first quality measurements, second Number is defect type number in the second quality measurements；
Judge whether first number is greater than second number；
If it is, determining that the priority of target audio file is lower than the priority of recommendation of audio file, otherwise, it determines mesh The priority of mark with phonetic symbols frequency file is higher than the priority of recommendation of audio file.
All the above alternatives, can form alternative embodiment of the invention according to any combination, and the present invention is real It applies example and this is no longer repeated one by one.
Fig. 2 is a kind of audio-frequency information recommended method flow chart provided in an embodiment of the present invention.Referring to fig. 2, this method application In audio playing platform, this method comprises:
Step 201: audio quality detection being carried out to target audio file, obtains the first quality measurements, the first mass Testing result is that there is no defect or existing defects, and when the first quality measurements are existing defects, the first mass is examined It further include the defect type of target audio file in survey result, target audio file is to have uploaded and do not carried out audio quality detection With the audio file of recommendation.
Since the audio file that the audio file of upload may be by bass matter is converted to, and the audio file uploaded In may existing defects, therefore, in order to improve recommendation audio-frequency information quality, need to uploading to audio playing platform Audio file carries out audio quality detection.
Wherein, audio quality detection is carried out to target audio file, obtains the first quality measurements, including such as under type (1) at least one of-(6):
(1), target audio file is decoded by specified decoder, if target audio file decoding failure, It determines target audio file existing defects and defect type is processing failure.
It should be noted that specified decoder can be arranged in advance, which is common for decoding audio The decoder of file, and when some audio file can not specify decoder to be decoded by this, determine the audio file It can not be decoded by other decoders, for example, the specified decoder can be FFmpeg (for recording, converting digital sound Frequently, video, and the open source computer program of stream can be translated into) common decoder, the embodiment of the present invention, which does not do this, to be had Body limits.
Further, when carrying out audio quality detection to target audio file only by mode (1), in target audio In the case that file is by specified decoder successfully decoded, can determining target audio file, there is no defects.
(2), target audio file is decoded by specified decoder, if target audio file successfully decoded, The practical duration of target audio file is determined based on the obtained target audio data of decoding, if target audio file it is practical when It is long to be less than or equal to specified duration, it is determined that target audio file existing defects and length when defect type is.
Specifically, target audio file is decoded by specified decoder, if target audio file successfully decoded, The sampling number in target audio data that then statistical decoder obtains, and obtain and adopt from the target audio data that decoding obtains Sample frequency obtains the practical duration of target audio file by the obtained sampling number of statistics divided by the sample frequency, and by target The practical duration of audio file is compared with specified duration, is specified if the practical duration of target audio file is less than or equal to Duration, it is determined that target audio file existing defects and length when defect type is.
For example, by specified decoder to target audio file successfully decoded, at this point it is possible to the target that statistical decoder obtains Sampling number in audio data is 150, and it is 15Hz that sample frequency is obtained from the target audio data that decoding obtains, and will be counted Obtained sampling number 150 divided by the sample frequency, obtain target audio file it is practical when it is 10 seconds a length of, if specified duration Be 15 seconds, determine that the practical duration of target audio file is less than specified duration, so determine target audio file existing defects and Length when defect type is.
Further, when carrying out audio quality detection to target audio file only by mode (2), if target sound Frequency file decoding failure, display decoding failure prompt information, decoding failure prompt information is for prompting target audio file solution Code failure, and if the practical duration of target audio file is greater than specified duration, it is determined that defect is not present in target audio file. But when pass-through mode (1) and (2) carry out audio quality detection to target audio file, only when target audio file solution Code success, and when the practical duration of target audio file is greater than specified duration, determining target audio file, there is no defects.
It should be noted that target audio file it is practical when a length of target audio file actually broadcasting when playing out Duration is put, which can be arranged in advance, for example, the specified duration can be 10 seconds, 15 seconds etc., the embodiment of the present invention It is not specifically limited in this embodiment.
In addition, can be PCM (Pulse Code Modulation, pulse volume based on the target audio data that decoding obtains Code modulation) data.
(3), based on the tonepulse of target audio file, judge with the presence or absence of transient pulse in target audio file, such as There are transient pulses in fruit target audio file, it is determined that target audio file existing defects and defect type are current sound.
Wherein, based on the tonepulse of target audio file, judge in target audio file with the presence or absence of transient pulse, packet It includes: obtaining the first pulse amplitude, the second pulse amplitude and third pulse amplitude from the tonepulse of target audio file, first Pulse amplitude is the pulse amplitude in the tonepulse of the target audio file at the first sampled point, and the second pulse amplitude is the mesh Pulse amplitude in the tonepulse of mark with phonetic symbols frequency file at the second sampled point, third pulse amplitude are the sound of the target audio file Pulse amplitude in frequency pulse at third sampled point, the first sample is between the second sampled point and third sampled point, and Two sampled points and third sampled point are adjacent with the first sampled point；First pulse amplitude is subtracted into the second pulse amplitude, obtains First pulse amplitude is subtracted third pulse amplitude, obtains the second difference by one difference；Refer to when the first difference is greater than or equal to first When tentering value and the second difference are greater than or equal to the first specified amplitude, determine that there are transient pulses in target audio file.
When the first difference is greater than or equal to the first specified amplitude and the second difference is greater than or equal to the first specified amplitude, really The gap of fixed first pulse amplitude and the second pulse amplitude is larger, and the gap of the first pulse amplitude and third pulse amplitude compared with Greatly, and then in the tonepulse of determining target audio file there are pulse amplitudes to compare the biggish arteries and veins of front and back pulse amplitude gap Punching, that is to say, there are transient pulses in target audio file.For there are the audio files of transient pulse, when progress audio letter Breath is played and is played when the position of transient pulse, it may appear that the sound of " whistling ".
For example, it is 3 that the first pulse amplitude obtained from the tonepulse of target audio file, which is the 25, second pulse amplitude, It is 5 with third pulse amplitude, the first pulse amplitude 25 is subtracted into the second pulse amplitude 3, obtaining the first difference is 22, by the first arteries and veins It rushes amplitude 25 and subtracts third pulse amplitude 5, obtaining the second difference is 20, if the first specified amplitude is 15, determines the first difference Greater than the first specified amplitude and the second difference is greater than or equal to the first specified amplitude, and then determines that there are winks in target audio file When pulse.
It should be noted that since the first sample is between the second sampled point and third sampled point, and the second sampling Point and third sampled point are adjacent with the first sampled point, and therefore, the first sampled point can be located at the second sampled point later and with the It is before two sampled points are adjacent, and the first sampled point can be located at third sampled point and adjacent with third sampled point or first adopts Sampling point can be located at the second sampled point before and it is adjacent with the second sampled point, and the first sampled point can be located at third sampled point it It is afterwards and adjacent with third sampled point.It that is to say, when the second sample is before the first sampled point, third sample is in After one sampled point, when the second sample is after the first sampled point, third sample is before the first sampled point, originally Inventive embodiments are not specifically limited in this embodiment.
In addition, the first specified amplitude can be arranged in advance, the embodiment of the present invention is not specifically limited in this embodiment.
Further, when carrying out audio quality detection to target audio file only by mode (3), if target sound Transient pulse is not present in frequency file, it is determined that defect is not present in target audio file.But when through the above way (1)- At least one of (3) when mode carries out audio quality detection, when being determined by way of at least one there is no when defect, really Defect is not present in the fixed target audio file.
(4), it determines the practical cut frequency of target audio file and refers to cut frequency, if the reality of target audio file Difference between border cut frequency and reference cut frequency is greater than or equal to specified cut frequency, it is determined that target audio file is deposited In defect and defect type is that sound quality code stream mismatches.
Specifically, the cut frequency of multiframe audio data in target audio file, and statistics target audio are counted respectively The code rate of file；The practical cut frequency of target audio file is determined based on the cut frequency of the multiframe audio data；Based on mesh The code rate of mark with phonetic symbols frequency file determines the reference cut frequency of target audio file.The practical cut frequency of target audio file is subtracted It goes to obtain practical cut frequency with reference to cut frequency and with reference to the difference between cut frequency, by the difference and specifies truncation frequency Rate is compared, if the difference is greater than or equal to specified cut frequency, it is determined that target audio file existing defects and defect Type is sound quality code stream mismatch.
It wherein, can when the cut frequency based on the multiframe audio data determines the practical cut frequency of target audio file To determine the average value of the cut frequency of the multiframe audio data, the reality which is determined as the target audio file is cut Disconnected frequency；Alternatively, counting the frequency of occurrence of identical cut frequency, and will occur from the cut frequency of the multiframe audio data The most cut frequency of number is determined as the practical cut frequency of target audio file.Certainly, in practical application, can also pass through Other modes are determined, and the embodiment of the present invention is not specifically limited in this embodiment.
In addition, can be based on when the code rate based on target audio file determines the reference cut frequency of target audio file The code rate of target audio file obtains corresponding cut frequency from the corresponding relationship between the code rate and cut frequency of storage, And the cut frequency that will acquire is determined as the reference cut frequency of target audio file.
Wherein, the cut frequency in the corresponding relationship between code rate and cut frequency is based on corresponding code rate, according to one Fixed algorithm is determined in advance to obtain.Therefore, the reference truncation frequency of target audio file is determined based on the code rate of target audio file When rate, it can be obtained based on the code rate of target audio file according to certain algorithm determination.
Due to having previously been based on after code rate determines cut frequency and be stored in the corresponding relationship between code rate and cut frequency, When determining the reference cut frequency of target audio file, it can be directly based upon the code rate of target audio file, closed from the correspondence It is acquired in system, it is simple, convenient, it is higher to obtain efficiency.
It should be noted that specified cut frequency can be arranged in advance, for example specified cut frequency is 1KHz, and the present invention is real Example is applied to be not specifically limited in this embodiment.
Further, when carrying out audio quality detection to target audio file only by mode (4), if target sound Difference between the practical cut frequency and reference cut frequency of frequency file is less than specified cut frequency, it is determined that target audio text Defect is not present in part.But when at least one of (1)-(4) mode carries out audio quality detection through the above way, when logical It crosses at least one mode and determines that there is no defects there is no the target audio file when defect, is determined.
(5), based on the tonepulse of target audio file, judge in target audio file with the presence or absence of instantaneous mute, such as Exist in fruit target audio file instantaneous mute, it is determined that target audio file existing defects and defect type is cassette tape.
Wherein, based on the tonepulse of target audio file, judge in target audio file with the presence or absence of instantaneous mute, packet It includes: obtaining the 4th pulse amplitude from the tonepulse of target audio file, the 4th pulse amplitude is the audio arteries and veins of target audio Pulse amplitude in punching at any sampled point；When the 4th pulse amplitude is less than or equal to the second specified amplitude, target sound is determined Exist in frequency file instantaneous mute.
When the 4th pulse amplitude is less than or equal to the second specified amplitude, determines in the tonepulse of target audio file and deposit Pulse amplitude be 0 or close to 0 pulse, that is to say exist in target audio file instantaneous mute.It is instantaneous quiet for existing The audio file of sound can be recognized suddenly without sound when carrying out audio-frequency information broadcasting and play when instantaneous mute position To be cassette tape.
For example, the 4th pulse amplitude obtained from the tonepulse of target audio file is 0.3, if the second specified width Value is 0.5, and it is instantaneous mute to determine that the 4th pulse amplitude exists less than the second specified amplitude, and then in determining target audio file.
It should be noted that the second specified amplitude can be arranged in advance, the embodiment of the present invention is not specifically limited in this embodiment.
Further, when carrying out audio quality detection to target audio file only by mode (5), if target sound There is no instantaneous mute in frequency file, it is determined that defect is not present in target audio file.But when through the above way (1)- At least one of (5) when mode carries out audio quality detection, when being determined by way of at least one there is no when defect, really Defect is not present in the fixed target audio file.
(6), the correlation and capacity volume variance for determining target audio file left and right acoustic channels when playing, if the correlation is small In or equal to specified correlation or the capacity volume variance more than or equal to specified difference value, it is determined that target audio file exists scarce It falls into and defect type is left and right acoustic channels Type-Inconsistencies.
It should be noted that specified correlation and specified difference value can be arranged in advance, the embodiment of the present invention to this not It is specifically limited.
In addition, determining that the target audio file correlation of left and right acoustic channels and method of capacity volume variance when playing can refer to This is not set forth in detail in the relevant technologies, the embodiment of the present invention.
Further, when carrying out audio quality detection to target audio file only by mode (6), if target sound The correlation of frequency file is greater than specified correlation and the capacity volume variance is less than specified difference value, it is determined that target audio file is not deposited In defect.But when at least one of (1)-(6) mode carries out audio quality detection through the above way, when by this extremely A kind of few mode determines that there is no defects there is no the target audio file when defect, is determined.
It in embodiments of the present invention, not only can be by least one of above-mentioned six kinds of modes mode to target audio text Part carry out audio quality detection, it is, of course, also possible to by other means to target audio file carry out audio quality detection, with The defect type of target audio file is determined when target audio file existing defects, for example, defect type can also include endless It is whole, have promotion message, noise reduction rubbish etc., the embodiment of the present invention is not specifically limited in this embodiment.
Step 202: being searched from the recommendation of audio library of storage and the matched recommendation of audio file of target audio file and the Two quality measurements, the second quality measurements are the quality measurements of the recommendation of audio file.
It should be noted that being searched and the matched recommendation of audio file of target audio file from the recommendation of audio library of storage When, can be searched according to assignment algorithm, and the matching at this refer to target audio file belong to same sound quality type or Person belongs to the recommendation of audio file of close sound quality type, this is not set forth in detail in the embodiment of the present invention.
In addition, since audio quality detection can be carried out when each audio file uploads to audio playing platform, and work as certain When one audio file can be recommended, the quality measurements of the audio file can be stored in recommendation of audio library, therefore, It can be searched from recommendation of audio library and the quality measurements of the matched recommendation of audio file of target audio file.
Wherein, the audio file stored in recommendation of audio library is the audio file of audio-frequency information to be recommended, and recommends sound It not only can store the audio file of audio-frequency information to be recommended and the quality measurements of the audio file in frequency library, it can be with The other information in audio-frequency information to be recommended is stored, the embodiment of the present invention is not specifically limited in this embodiment.
Step 203: if found, being based on the first quality measurements and the second quality measurements, determine target Priority relationship between audio file and recommendation of audio file.
Specifically, if found from recommendation of audio library and the matched recommendation of audio file of target audio file and second Quality measurements, and when the first quality measurements are there is no when defect, if the second quality measurements are to exist to lack It falls into, it is determined that the priority of target audio file is higher than the priority of the recommendation of audio file, if the second quality measurements For there is no defects, it is determined that the priority of target audio file is identical as the priority of recommendation of audio file；When the first mass Testing result be existing defects when, if the second quality measurements be there is no defects, it is determined that target audio file it is excellent First grade is lower than the priority of recommendation of audio file, if the second quality measurements are existing defects, is examined based on the first mass The defects of result the defects of type and the second quality measurements type is surveyed, determines target audio file and recommendation of audio text The priority relationship of part.
Wherein, the defect class that the defect type and the second quality measurements for including based on the first quality measurements include Type determines the priority relationship of target audio file and recommendation of audio file, comprising: determine first number and second number, the One number is defect type number in the first quality measurements, and second number is defect type in the second quality measurements Number；Judge whether first number is greater than second number；If it is, determining that the priority of target audio file is lower than recommendation of audio The priority of file, otherwise, it determines the priority of target audio file is higher than the priority of recommendation of audio file.
It optionally, is to exist in the first quality measurements when target audio file is the audio file of lossless type In the case where defect, the priority that can directly determine target audio file is lower than the priority of recommendation of audio file.
Further, if do not found from the recommendation of audio library of storage and the matched recommendation sound of target audio file Target audio file and the first quality measurements then directly can be added to recommendation by frequency file and the second quality measurements In audio repository, to be updated to recommendation of audio file.
Step 204: when the priority of target audio file is higher than the priority of recommendation of audio file, by recommendation of audio library In the recommendation of audio file replace with target audio file, to be updated to recommendation of audio library.
It further, can be by mesh when the priority of target audio file is identical as the priority of recommendation of audio file Mark with phonetic symbols frequency file is stored in recommendation of audio library, to be updated to recommendation of audio library, alternatively, can also be by recommendation of audio file Target audio file is replaced with, to be updated to recommendation of audio library, certainly, recommendation of audio library is not updated also, this hair Bright embodiment is not specifically limited in this embodiment.
It should be noted that the embodiment of the present invention not only can be in through the above steps 203 and 204 method to recommendation of audio Library is updated, and certainly, in practical application, can also be updated by other methods to recommendation of audio library, and the present invention is real Example is applied to be not specifically limited in this embodiment.
Step 205: updated recommendation of audio library is recommended.
In embodiments of the present invention, audio quality detection is carried out to target audio file, obtains the first quality measurements, Since the first quality measurements can be for there is no defect or there are disadvantages, and when the first quality measurements are existing defects When, it can also include the defect type of target audio file in the first quality measurements, that is to say, the embodiment of the present invention is to mesh When mark with phonetic symbols frequency file carries out audio quality detection, it can not only determine that target audio file, can also be in mesh with the presence or absence of defect When mark with phonetic symbols frequency file existing defects, the defect type of target audio file is determined, therefore quality inspection is carried out to target audio file The accuracy rate of survey is higher.In addition, the embodiment of the present invention is to all sounds uploaded and do not carry out audio quality detection and recommendation Frequency file carries out audio quality detection, wherein just including the audio file of lossless type, therefore, the embodiment of the present invention can be into one Step improves the accuracy rate of audio quality detection, and then improves the accuracy rate that audio-frequency information is recommended.
Fig. 3 is a kind of structural schematic diagram of audio-frequency information recommendation apparatus provided in an embodiment of the present invention.Referring to Fig. 3, the dress It sets and includes:
Audio quality detection module 301 obtains the inspection of the first mass for carrying out audio quality detection to target audio file It surveys as a result, the first quality measurements are that there is no defect or existing defects, and working as the first quality measurements is to exist to lack It further include the defect type of target audio file when falling into, in the first quality measurements, target audio file is to have uploaded and not Carry out the audio file of audio quality detection and recommendation；
Searching module 302, for being searched and the matched recommendation of audio of target audio file from the recommendation of audio library of storage File and the second quality measurements, the second quality measurements are the quality measurements of recommendation of audio file；
Update module 303, if being based on the first quality measurements and the second quality measurements for finding, Recommendation of audio library is updated；
Recommending module 304, for recommending updated recommendation of audio library.
Optionally, audio quality detection module 301 includes such as at least one of lower unit:
First determination unit, for being decoded by specified decoder to target audio file, if target audio is literary Part decoding failure, it is determined that target audio file existing defects and defect type are processing failure；
Second determination unit, for being decoded by specified decoder to target audio file, if target audio is literary Part successfully decoded, then the target audio data obtained based on decoding determine the practical duration of target audio file, if target sound The practical duration of frequency file is less than or equal to specified duration, it is determined that target audio file existing defects and defect type is duration It is short；
Third determination unit judges whether deposit in target audio file for the tonepulse based on target audio file In transient pulse, if there are transient pulses in target audio file, it is determined that target audio file existing defects and defect class Type is current sound；
4th determination unit, for determining the practical cut frequency of target audio file and with reference to cut frequency, if mesh Difference between the practical cut frequency and reference cut frequency of mark with phonetic symbols frequency file is greater than or equal to specified cut frequency, it is determined that Target audio file existing defects and defect type are sound quality code stream mismatch；
5th determination unit judges whether deposit in target audio file for the tonepulse based on target audio file Instantaneous mute, if existed in target audio file instantaneous mute, it is determined that target audio file existing defects and defect class Type is cassette tape；
6th determination unit, for determining the correlation and capacity volume variance of target audio file left and right acoustic channels when playing, If the correlation of target audio file is less than or equal to specified correlation or capacity volume variance is greater than or equal to specified difference value, It then determines target audio file existing defects and defect type is left and right acoustic channels Type-Inconsistencies.
Optionally, third determination unit includes:
First obtains subelement, for obtaining the first pulse amplitude, the second arteries and veins from the tonepulse of target audio file Amplitude and third pulse amplitude are rushed, the first pulse amplitude is the pulse in the tonepulse of target audio file at the first sampled point Amplitude, the second pulse amplitude are the pulse amplitude in the tonepulse of target audio file at the second sampled point, third pulse width Value is the pulse amplitude in the tonepulse of target audio file at third sampled point, the first sample in the second sampled point and Between third sampled point, and the second sampled point and third sampled point are adjacent with the first sampled point；
Computation subunit obtains the first difference, by the first pulse for the first pulse amplitude to be subtracted the second pulse amplitude Amplitude subtracts third pulse amplitude, obtains the second difference；
First determines subelement, for being greater than or waiting when the first difference is greater than or equal to the first specified amplitude and the second difference When the first specified amplitude, determine that there are transient pulses in target audio file；
Second determines subelement, if for there are transient pulses in target audio file, it is determined that target audio file Existing defects and defect type are current sound.
Optionally, the 4th determination unit includes:
Subelement is counted, for counting the cut frequency of multiframe audio data in target audio file, and statistics respectively The code rate of target audio file；
Third determines subelement, for determining the reality of target audio file based on the cut frequency of the multiframe audio data Cut frequency；
4th determines subelement, and the reference truncation frequency of target audio file is determined for the code rate based on target audio file Rate；
5th determines subelement, if for the practical cut frequency of target audio file and with reference between cut frequency Difference is greater than or equal to specified cut frequency, it is determined that target audio file existing defects and defect type be sound quality code stream not Match.
Optionally, the 5th determination unit includes:
Second obtains subelement, for obtaining the 4th pulse amplitude, the 4th arteries and veins from the tonepulse of target audio file Rush the pulse amplitude in the tonepulse that amplitude is target audio file at any sampled point；
6th determines subelement, for determining target sound when the 4th pulse amplitude is less than or equal to the second specified amplitude Exist in frequency file instantaneous mute；
7th determines subelement, if instantaneous mute for existing in target audio file, it is determined that target audio file Existing defects and defect type are cassette tape.
Optionally, update module 303 includes:
7th determination unit determines target audio for being based on the first quality measurements and the second quality measurements Priority relationship between file and recommendation of audio file；
Replacement unit will be recommended when for being higher than the priority of recommendation of audio file when the priority of target audio file Audio file replaces with target audio file, to be updated to recommendation of audio library.
Optionally, the 7th determination unit includes:
8th determines subelement, for when the first quality measurements are there is no when defect, if the second quality testing It as a result is existing defects, it is determined that the priority of target audio file is higher than the priority of recommendation of audio file, if the second matter Measuring testing result is that there is no defects, it is determined that the priority of target audio file is identical as the priority of recommendation of audio file；
9th determines subelement, is used for when the first quality measurements are existing defects, if the second quality testing knot Fruit is that there is no defects, it is determined that the priority of target audio file is lower than the priority of recommendation of audio file, if the second matter Amount testing result is existing defects, then based in the defects of the first quality measurements type and the second quality measurements Defect type determines the priority relationship of target audio file and recommendation of audio file.
Optionally, it the 9th determines subelement, is specifically used for:
When the first quality measurements are existing defects, if the second quality measurements are that there is no defects, really Set the goal audio file priority be lower than recommendation of audio file priority；
If the second quality measurements are existing defects, it is determined that first number and second number, first number are the Defect type number in one quality measurements, second number are defect type number in the second quality measurements；
Judge whether first number is greater than second number；
If it is, determining that the priority of target audio file is lower than the priority of recommendation of audio file, otherwise, it determines mesh The priority of mark with phonetic symbols frequency file is higher than the priority of recommendation of audio file.
In embodiments of the present invention, audio quality detection is carried out to target audio file, obtains the first quality measurements, Since the first quality measurements can be for there is no defect or there are disadvantages, and when the first quality measurements are existing defects When, it can also include the defect type of target audio file in the first quality measurements, that is to say, the embodiment of the present invention is to mesh When mark with phonetic symbols frequency file carries out audio quality detection, it can not only determine that target audio file, can also be in mesh with the presence or absence of defect When mark with phonetic symbols frequency file existing defects, the defect type of target audio file is determined, therefore quality inspection is carried out to target audio file The accuracy rate of survey is higher.In addition, the embodiment of the present invention is to all sounds uploaded and do not carry out audio quality detection and recommendation Frequency file carries out audio quality detection, wherein just including the audio file of lossless type, therefore, the embodiment of the present invention can be into one Step improves the accuracy rate of audio quality detection, and then improves the accuracy rate that audio-frequency information is recommended.
It should be understood that audio-frequency information recommendation apparatus provided by the above embodiment is when carrying out audio-frequency information recommendation, only The example of the division of the above functional modules, in practical application, can according to need and by above-mentioned function distribution by Different functional modules is completed, i.e., the internal structure of device is divided into different functional modules, described above complete to complete Portion or partial function.In addition, audio-frequency information recommendation apparatus provided by the above embodiment and audio-frequency information recommended method embodiment Belong to same design, specific implementation process is detailed in embodiment of the method, and which is not described herein again.
Those of ordinary skill in the art will appreciate that realizing that all or part of the steps of above-described embodiment can pass through hardware It completes, relevant hardware can also be instructed to complete by program, the program can store in a kind of computer-readable In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely presently preferred embodiments of the present invention, is not intended to limit the invention, it is all in spirit of the invention and Within principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.

Claims (16)

1. a kind of audio-frequency information recommended method, which is characterized in that the described method includes:
Audio quality detection is carried out to target audio file, obtains the first quality measurements, first quality measurements For there is no defect or existing defects, and when first quality measurements are existing defects, the first mass inspection It further include the defect type of the target audio file in survey result, the target audio file is to have uploaded and do not carried out audio The audio file of quality testing and recommendation；
It searches from the recommendation of audio library of storage and is examined with the matched recommendation of audio file of the target audio file and the second mass It surveys as a result, second quality measurements are the quality measurements of the recommendation of audio file；
If found, first quality measurements and second quality measurements are based on, to the recommendation sound Frequency library is updated；
Updated recommendation of audio library is recommended；
It is described that audio quality detection is carried out to target audio file, the first quality measurements are obtained, including in such as under type At least one:
The target audio file is decoded by specified decoder, if target audio file decoding failure, It determines the target audio file existing defects and defect type is processing failure；
The target audio file is decoded by the specified decoder, if the target audio file is decoded into Function, then the target audio data obtained based on decoding determine the practical duration of the target audio file, if when described practical It is long to be less than or equal to specified duration, it is determined that the target audio file existing defects and length when defect type is；
Based on the tonepulse of the target audio file, judge with the presence or absence of transient pulse in the target audio file, such as There are transient pulses in target audio file described in fruit, it is determined that the target audio file existing defects and defect type are electricity Stream sound；
Based on the tonepulse of the target audio file, judge in the target audio file with the presence or absence of instantaneous mute, such as Exist in target audio file described in fruit instantaneous mute, it is determined that target audio file existing defects and defect type is card Band；
The correlation and capacity volume variance for determining target audio file left and right acoustic channels when playing, if the correlation is less than Or it is equal to specified correlation or the capacity volume variance more than or equal to specified difference value, it is determined that the target audio file is deposited In defect and defect type is left and right acoustic channels Type-Inconsistencies.
2. the method as described in claim 1, which is characterized in that it is described that audio quality detection is carried out to target audio file, it obtains Further include such as under type to the first quality measurements:
It determines the practical cut frequency of the target audio file and refers to cut frequency, if the practical cut frequency and institute It states and is greater than or equal to specified cut frequency with reference to the difference between cut frequency, it is determined that the target audio file existing defects And defect type is sound quality code stream mismatch.
3. the method as described in claim 1, which is characterized in that the tonepulse based on the target audio file is sentenced It whether there is transient pulse in the target audio file of breaking, comprising:
The first pulse amplitude, the second pulse amplitude and third pulse width are obtained from the tonepulse of the target audio file Value, first pulse amplitude are the pulse amplitude in the tonepulse at the first sampled point, and second pulse amplitude is Pulse amplitude in the tonepulse at the second sampled point, the third pulse amplitude are third sampling in the tonepulse Pulse amplitude at point, first sample is between second sampled point and the third sampled point, and described the Two sampled points and the third sampled point are adjacent with first sampled point；
First pulse amplitude is subtracted into second pulse amplitude, obtains the first difference, first pulse amplitude is subtracted The third pulse amplitude is removed, the second difference is obtained；
When first difference is more than or equal to the first specified amplitude and second difference is greater than or equal to described first and specifies When amplitude, determine that there are transient pulses in the target audio file.
4. method according to claim 2, which is characterized in that the practical cut frequency of the determination target audio file With reference cut frequency, comprising:
The cut frequency of multiframe audio data in the target audio file, and the statistics target audio file are counted respectively Code rate；
The practical cut frequency of the target audio file is determined based on the cut frequency of the multiframe audio data；
The reference cut frequency of the target audio file is determined based on the code rate of the target audio file.
5. the method as described in claim 1, which is characterized in that the tonepulse based on the target audio file is sentenced With the presence or absence of instantaneous mute in the target audio file of breaking, comprising:
The 4th pulse amplitude is obtained from the tonepulse of the target audio file, the 4th pulse amplitude is the audio Pulse amplitude in pulse at any sampled point；
When the 4th pulse amplitude is less than or equal to the second specified amplitude, determines in the target audio file and exist instantaneously It is mute.
6. the method as described in claim 1, which is characterized in that described to be based on first quality measurements and described second Quality measurements are updated the recommendation of audio library, comprising:
Based on first quality measurements and second quality measurements, determine the target audio file with it is described Priority relationship between recommendation of audio file；
When the priority of the target audio file is higher than the priority of the recommendation of audio file, by the recommendation of audio text Part replaces with the target audio file, to be updated to the recommendation of audio library.
7. method as claimed in claim 6, which is characterized in that described to be based on first quality measurements and described second Quality measurements determine the priority relationship between the target audio file and the recommendation of audio file, comprising:
When first quality measurements be there is no when defect, if second quality measurements are existing defects, Then determine that the priority of the target audio file is higher than the priority of the recommendation of audio file, if second mass is examined Surveying result is that there is no defects, it is determined that the priority phase of the priority of the target audio file and the recommendation of audio file Together；
When first quality measurements are existing defects, if second quality measurements be there is no defect, Then determine that the priority of the target audio file is lower than the priority of the recommendation of audio file, if second mass is examined Survey result is existing defects, then based on the defects of first quality measurements type and second quality measurements The defects of type, determine the priority relationship of the target audio file and the recommendation of audio file.
8. the method for claim 7, which is characterized in that the defect for including based on first quality measurements The defect type that type and second quality measurements include determines the target audio file and recommendation of audio text The priority relationship of part, comprising:
Determine that first number and second number, first number are defect type number in first quality measurements, Second number is defect type number in second quality measurements；
Judge whether first number is greater than second number；
If it is, determining that the priority of the target audio file is lower than the priority of the recommendation of audio file, otherwise, really The priority of the fixed target audio file is higher than the priority of the recommendation of audio file.
9. a kind of audio-frequency information recommendation apparatus, which is characterized in that described device includes:
Audio quality detection module, for obtaining the first quality measurements to target audio file progress audio quality detection, First quality measurements are that there is no defect or existing defects, and working as first quality measurements is to exist to lack It further include the defect type of the target audio file, the target audio file when falling into, in first quality measurements For the audio file for having uploaded and not carried out audio quality detection and recommendation；
Searching module, for being searched and the matched recommendation of audio file of the target audio file from the recommendation of audio library of storage With the second quality measurements, second quality measurements are the quality measurements of the recommendation of audio file；
Update module, if being based on first quality measurements and second quality measurements for finding, The recommendation of audio library is updated；
Recommending module, for recommending updated recommendation of audio library；
The audio quality detection module includes such as at least one of lower unit:
First determination unit, for being decoded by specified decoder to the target audio file, if the target sound Frequency file decoding failure, it is determined that the target audio file existing defects and defect type are processing failure；
Second determination unit, for being decoded by the specified decoder to the target audio file, if the mesh Mark with phonetic symbols frequency file successfully decoded, then based on the obtained target audio data of decoding determine the target audio file it is practical when It is long, if the practical duration is less than or equal to specified duration, it is determined that the target audio file existing defects and defect class Length when type is；
Third determination unit, for the tonepulse based on the target audio file, judge be in the target audio file No there are transient pulses, if there are transient pulses in the target audio file, it is determined that the target audio file exists Defect and defect type are current sound；
5th determination unit, for the tonepulse based on the target audio file, judge be in the target audio file It is no exist it is instantaneous mute, if existed in the target audio file instantaneous mute, it is determined that the target audio file exists Defect and defect type are cassette tape；
6th determination unit, for determining the correlation and capacity volume variance of target audio file left and right acoustic channels when playing, If the correlation is less than or equal to specified correlation or the capacity volume variance is greater than or equal to specified difference value, it is determined that The target audio file existing defects and defect type are left and right acoustic channels Type-Inconsistencies.
10. device as claimed in claim 9, which is characterized in that the audio quality detection module further includes such as lower unit:
4th determination unit, for determining the practical cut frequency of the target audio file and with reference to cut frequency, if institute It states practical cut frequency and the difference with reference between cut frequency is greater than or equal to specified cut frequency, it is determined that the mesh Mark with phonetic symbols frequency file existing defects and defect type are sound quality code stream mismatch.
11. device as claimed in claim 9, which is characterized in that the third determination unit includes:
First obtains subelement, for obtaining the first pulse amplitude, the second arteries and veins from the tonepulse of the target audio file Amplitude and third pulse amplitude are rushed, first pulse amplitude is the pulse amplitude in the tonepulse at the first sampled point, Second pulse amplitude is the pulse amplitude in the tonepulse at the second sampled point, and the third pulse amplitude is described Pulse amplitude in tonepulse at third sampled point, first sample are adopted in second sampled point and the third Between sampling point, and second sampled point and the third sampled point are adjacent with first sampled point；
Computation subunit obtains the first difference for first pulse amplitude to be subtracted second pulse amplitude, will be described First pulse amplitude subtracts the third pulse amplitude, obtains the second difference；
First determines subelement, for when first difference is more than or equal to the first specified amplitude and second difference is greater than Or when being equal to the first specified amplitude, determine that there are transient pulses in the target audio file；
Second determines subelement, if for there are transient pulses in the target audio file, it is determined that the target audio File existing defects and defect type are current sound.
12. device as claimed in claim 10, which is characterized in that the 4th determination unit includes:
Subelement is counted, for counting the cut frequency of multiframe audio data in the target audio file, and statistics respectively The code rate of the target audio file；
Third determines subelement, for determining the reality of the target audio file based on the cut frequency of the multiframe audio data Border cut frequency；
4th determines subelement, for determining that the reference of the target audio file is cut based on the code rate of the target audio file Disconnected frequency；
5th determine subelement, if be greater than for the practical cut frequency and the difference with reference between cut frequency or Equal to specified cut frequency, it is determined that the target audio file existing defects and defect type are that sound quality code stream mismatches.
13. device as claimed in claim 9, which is characterized in that the 5th determination unit includes:
Second obtains subelement, for obtaining the 4th pulse amplitude from the tonepulse of the target audio file, described the Four pulse amplitudes are the pulse amplitude in the tonepulse at any sampled point；
6th determines subelement, for determining the mesh when the 4th pulse amplitude is less than or equal to the second specified amplitude Exist in mark with phonetic symbols frequency file instantaneous mute；
7th determines subelement, if instantaneous mute for existing in the target audio file, it is determined that the target audio File existing defects and defect type are cassette tape.
14. device as claimed in claim 9, which is characterized in that the update module includes:
7th determination unit, described in determining based on first quality measurements and second quality measurements Priority relationship between target audio file and the recommendation of audio file；
Replacement unit will when for being higher than the priority of the recommendation of audio file when the priority of the target audio file The recommendation of audio file replaces with the target audio file, to be updated to the recommendation of audio library.
15. device as claimed in claim 14, which is characterized in that the 7th determination unit includes:
8th determines subelement, for when first quality measurements are there is no when defect, if second mass Testing result is existing defects, it is determined that the priority of the target audio file is higher than the preferential of the recommendation of audio file Grade, if second quality measurements be there is no defects, it is determined that the priority of the target audio file with it is described The priority of recommendation of audio file is identical；
9th determines subelement, is used for when first quality measurements are existing defects, if second mass is examined Surveying result is that there is no defects, it is determined that the priority of the target audio file is preferential lower than the recommendation of audio file Grade, if second quality measurements are existing defects, based on the defects of first quality measurements type With the defects of second quality measurements type, the excellent of the target audio file and the recommendation of audio file is determined First grade relationship.
16. device as claimed in claim 15, which is characterized in that
Described 9th determines subelement, is specifically used for:
When first quality measurements are existing defects, if second quality measurements be there is no defect, Then determine that the priority of the target audio file is lower than the priority of the recommendation of audio file；
If second quality measurements are existing defects, it is determined that first number and second number, first number For defect type number in first quality measurements, second number is defect in second quality measurements Type number；
Judge whether first number is greater than second number；
If it is, determining that the priority of the target audio file is lower than the priority of the recommendation of audio file, otherwise, really The priority of the fixed target audio file is higher than the priority of the recommendation of audio file.
CN201610128170.7A 2016-03-07 2016-03-07 Audio-frequency information recommended method and device Active CN105808719B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610128170.7A CN105808719B (en) 2016-03-07 2016-03-07 Audio-frequency information recommended method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610128170.7A CN105808719B (en) 2016-03-07 2016-03-07 Audio-frequency information recommended method and device

Publications (2)

Publication Number Publication Date
CN105808719A CN105808719A (en) 2016-07-27
CN105808719B true CN105808719B (en) 2019-09-27

Family

ID=56467777

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610128170.7A Active CN105808719B (en) 2016-03-07 2016-03-07 Audio-frequency information recommended method and device

Country Status (1)

Country Link
CN (1) CN105808719B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106484833A (en) * 2016-09-30 2017-03-08 四川九洲电器集团有限责任公司 A kind of source of sound screening technique and electronic equipment
CN108111908A (en) * 2017-12-25 2018-06-01 深圳Tcl新技术有限公司 Audio quality determines method, equipment and computer readable storage medium
CN108091352B (en) * 2017-12-27 2020-10-13 腾讯音乐娱乐科技（深圳）有限公司 Audio file processing method and device, storage medium and terminal equipment
CN108231091B (en) * 2018-01-24 2021-05-25 广州酷狗计算机科技有限公司 Method and device for detecting whether left and right sound channels of audio are consistent

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102227127A (en) * 2011-06-21 2011-10-26 天津理工大学 Automatic multimedia material defect detection and quality analysis method
CN102568470A (en) * 2012-01-11 2012-07-11 广州酷狗计算机科技有限公司 Acoustic fidelity identification method and system for audio files
CN103455513A (en) * 2012-06-01 2013-12-18 腾讯科技（深圳）有限公司 Updating method and updating device for audio file

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9412129B2 (en) * 2013-01-04 2016-08-09 Skullcandy, Inc. Equalization using user input

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102227127A (en) * 2011-06-21 2011-10-26 天津理工大学 Automatic multimedia material defect detection and quality analysis method
CN102568470A (en) * 2012-01-11 2012-07-11 广州酷狗计算机科技有限公司 Acoustic fidelity identification method and system for audio files
CN103455513A (en) * 2012-06-01 2013-12-18 腾讯科技（深圳）有限公司 Updating method and updating device for audio file

Also Published As

Publication number Publication date
CN105808719A (en) 2016-07-27

Similar Documents

Publication Publication Date Title
Richey et al. Voices obscured in complex environmental settings (voices) corpus
Mesaros et al. A multi-device dataset for urban acoustic scene classification
CN105808719B (en) Audio-frequency information recommended method and device
Mauch et al. The audio degradation toolbox and its application to robustness evaluation
CN106486128B (en) Method and device for processing double-sound-source audio data
Dean et al. The QUT-NOISE-TIMIT corpus for evaluation of voice activity detection algorithms
US7062442B2 (en) Method and arrangement for search and recording of media signals
US8886529B2 (en) Method and device for the objective evaluation of the voice quality of a speech signal taking into account the classification of the background noise contained in the signal
CN102959544A (en) Methods and systems for synchronizing media
JP2005518594A (en) A system that sells products using audio content identification
US10832700B2 (en) Sound file sound quality identification method and apparatus
CN111640411B (en) Audio synthesis method, device and computer readable storage medium
CN108259985A (en) Live audio sound mixing method, device, readable storage medium storing program for executing and equipment
CN109658935B (en) Method and system for generating multi-channel noisy speech
CN110191368A (en) Video data acquiring and alignment schemes, device, electronic equipment and system
CN108091352A (en) A kind of audio file processing method, device and storage medium
WO2023040520A1 (en) Method and apparatus for performing music matching of video, and computer device and storage medium
CN106534980A (en) Anomaly detection method for audio processing system and logging method and device
JP2005522744A (en) How to identify audio content
CN105635418A (en) Method and device for cutting bell
CN106095943B (en) It gives song recitals and knows well range detection method and device
KR102431737B1 (en) Method of searching highlight in multimedia data and apparatus therof
Alonso-Jiménez et al. Automatic detection of audio problems for quality control in digital music distribution
CN111885474A (en) Microphone testing method and device
CN115132183B (en) Training method, device, equipment, medium and program product of audio recognition model

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information

Address after: 510660 Guangzhou City, Guangzhou, Guangdong, Whampoa Avenue, No. 315, self - made 1-17

Applicant after: Guangzhou KuGou Networks Co., Ltd.

Address before: 510000 B1, building, No. 16, rhyme Road, Guangzhou, Guangdong, China 13F

Applicant before: Guangzhou KuGou Networks Co., Ltd.

CB02 Change of applicant information
GR01 Patent grant
GR01 Patent grant