CN104157296A - Audio frequency evaluative method and device - Google Patents
Audio frequency evaluative method and device Download PDFInfo
- Publication number
- CN104157296A CN104157296A CN201410362595.5A CN201410362595A CN104157296A CN 104157296 A CN104157296 A CN 104157296A CN 201410362595 A CN201410362595 A CN 201410362595A CN 104157296 A CN104157296 A CN 104157296A
- Authority
- CN
- China
- Prior art keywords
- sequence
- simple sentence
- measured
- related coefficient
- fixed reference
- 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.)
- Granted
Links
Landscapes
- Machine Translation (AREA)
Abstract
The embodiment of the invention provides an audio frequency evaluative method and device. The method comprises the following steps: acquiring the order of a to-be-evaluated singing simple sentence in a target audio file and the to-be-evaluated characteristic sequence of the singing simple sentence; positioning a referential single sentence in a source audio file and acquiring the referential characteristic sequence of the referential single sentence according to the order of the singing single sentence in the target audio file; conducting related calculation on the referential characteristic sequence and the to-be-evaluated characteristic sequence to acquire related coefficient sequence; confirming the evaluation grade on the singing single sentence according to the related coefficient sequence. According to the invention, evaluation can be conducted on the singing single sentence comprised in the target audio file; application requirements to audio file are satisfied; the application intellectuality of the audio file is improved.
Description
Technical field
The present invention relates to Internet technical field, be specifically related to audio signal processing technique field, relate in particular to a kind of audio frequency assessment method and device.
Background technology
Along with the development of Internet technology, in internet audio storehouse, include a large amount of such as the audio file such as song, snatch of song, application about internet audio also day by day increases, for example: KTV (Karaoke Television, Karaoke) system, K song system etc.In the process of application audio file, most of user wishes to know in real time deduction level in audio frequency deductive procedure, for example: user is in the time singing certain song, wish to obtain the test and appraisal mark of each of song, to understand the performance situation of every, etc., therefore, how the audio frequency simple sentence in the audio file such as such as song is tested and assessed and become a technical matters urgently to be resolved hurrily.
Summary of the invention
The embodiment of the present invention provides a kind of audio frequency assessment method and device, can test and assess to the performance simple sentence of target audio file including, meets the application demand to audio file, and the application that promotes audio file is intelligent.
Embodiment of the present invention first aspect provides a kind of audio frequency assessment method, can comprise:
Obtain performance simple sentence the to be tested and assessed order in target audio file and the characteristic sequence to be measured of described performance simple sentence;
Order according to described performance simple sentence in described target audio file, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence;
Described fixed reference feature sequence and described characteristic sequence to be measured are carried out to related operation, obtain related coefficient sequence;
According to described related coefficient sequence, determine the test and appraisal mark of described performance simple sentence.
Embodiment of the present invention second aspect provides a kind of audio frequency assessment device, can comprise:
Acquisition module to be measured, for obtaining performance simple sentence to be tested and assessed at the order of target audio file and the characteristic sequence to be measured of described performance simple sentence;
With reference to acquisition module, for the order at described target audio file according to described performance simple sentence, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence;
Related operation module, for described fixed reference feature sequence and described characteristic sequence to be measured are carried out to related operation, obtains related coefficient sequence;
Test and appraisal module, for according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence.
Implement the embodiment of the present invention, there is following beneficial effect:
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Brief description of the drawings
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
The process flow diagram of a kind of audio frequency assessment method that Fig. 1 provides for the embodiment of the present invention;
Fig. 2 is the process flow diagram of an embodiment of the step S103 shown in Fig. 1;
Fig. 3 is the process flow diagram of an embodiment of the step S104 shown in Fig. 1;
Fig. 4 is the process flow diagram of another embodiment of the step S104 shown in Fig. 1;
The structural representation of a kind of audio frequency assessment device that Fig. 5 provides for the embodiment of the present invention;
Fig. 6 is the structural representation of the embodiment of the related operation module shown in Fig. 5;
Fig. 7 is the structural representation of an embodiment of the test and appraisal module shown in Fig. 5;
Fig. 8 is the structural representation of another embodiment of the test and appraisal module shown in Fig. 5.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiment.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
In the embodiment of the present invention, audio file can include but not limited to: the files such as song, snatch of song.Source audio file refers to can be used for the file as the reference data of audio frequency test and appraisal, for example: and original singer's song, or the snatch of song intercepting from original singer's song etc.; Target audio file refers to carry out the file of audio frequency test and appraisal, for example: and the song that user sings again for original singer's song, or the snatch of song intercepting the song of again singing from user etc.
In the embodiment of the present invention, audio file can be arranged and be formed by least one audio frequency simple sentence order, the set description of this at least one audio frequency simple sentence in audio file, need the part of singing.Taking song A as example, the description of song A can be expressed as follows:
[661，860]aaaaaaaa
[1541，320]bbbbbbbb
[1871，245]cccccccc
……
In the description of above-mentioned song A, such as " aaaaaaaa ", " bbbbbbbb ", " cccccccc " can be respectively used to represent an audio frequency simple sentence, " [] " before each audio frequency simple sentence is for describing the time attribute of corresponding audio frequency simple sentence, its unit interval is generally ms, for example: above-mentioned [661, 860] for the time attribute of description audio simple sentence " aaaaaaaa ", " 661 " wherein represent the start time of audio frequency simple sentence " aaaaaaaa ", " 860 " represent the duration of audio frequency simple sentence " aaaaaaaa ", suppose song A totally 5 minutes, audio frequency simple sentence " aaaaaaaa " is sung since 661ms, lasting 860ms finishes to sing.According to the sequencing of start time, can determine the order of each audio frequency simple sentence that audio file comprises, for example: according to the description of above-mentioned song A, audio frequency simple sentence " aaaaaaaa " is first audio frequency simple sentence, and its order in song A is 1; Audio frequency simple sentence " bbbbbbbb " is second audio frequency simple sentence, and its order in song A is 2; By that analogy.Be understandable that, before each audio frequency simple sentence starts or after finishing, for example also can comprise, without the part of singing: the 0-661ms time period of above-mentioned song A is that this part can comprise prelude information without the part of singing.
In the embodiment of the present invention, described source audio file is arranged and is formed by least one audio frequency simple sentence order, and this audio frequency simple sentence can be described as performance simple sentence.Described target audio file is arranged and is formed by least one audio frequency simple sentence order, and this audio frequency simple sentence can be described as performance simple sentence.Except special instruction, in the follow-up each embodiment of the present invention, the described reference simple sentence of locating from described source audio file that refers in particular to reference to simple sentence, described performance simple sentence refers in particular to performance simple sentence to be tested and assessed in described target audio file.
Below in conjunction with Fig. 1-Fig. 4, the audio frequency assessment method that the embodiment of the present invention is provided describes in detail.
Refer to Fig. 1, the process flow diagram of a kind of audio frequency assessment method providing for the embodiment of the present invention; The method can comprise the following steps S101-step S104.
S101, obtains performance simple sentence the to be tested and assessed order in target audio file and the characteristic sequence to be measured of described performance simple sentence.
Note is again note, refers to the symbol of the carrying out of the sound for recording different length, can comprise whole note, minim, crotchet, quaver etc. kind.Audio frequency simple sentence can be expressed as a frame sequence of multiple audio frames compositions, and each audio frame all carries note, and according to each audio frame, the order of the time order and function in this audio frequency simple sentence forms the melody of this audio frequency simple sentence to each note.Pitch is again pitch, refers to the height of sound.Audio frequency simple sentence can be expressed as a frame sequence of multiple audio frames compositions, and each audio frame all carries pitch, and according to each audio frame, the order of the time order and function in this audio frequency simple sentence forms the melody of this audio frequency simple sentence to each pitch.To sum up, the sequence of notes of audio frequency simple sentence or pitch sequence all can reflect the melody characteristics of this audio frequency simple sentence.
In the embodiment of the present invention, preferably, performance simple sentence to be tested and assessed can be current performance simple sentence, and now audio frequency test and appraisal process can be real-time audio test and appraisal process.Wherein, current performance simple sentence refers to corresponding performance simple sentence of current time of playing in described target audio file.Target audio file is song A, the example that is described as with above-mentioned song A: suppose song A totally 5 minutes, the current time of playing is 1895ms, according to the description of song A, 1895ms belonged in the described time period of time attribute of audio frequency simple sentence " cccccccc ", can determine that thus audio frequency simple sentence " cccccccc " is current performance simple sentence, further, can determine that audio frequency simple sentence " cccccccc " is performance simple sentence to be tested and assessed, described in performance simple sentence the to be tested and assessed order in target audio file be 3.
In this step, can obtain the characteristic sequence to be measured of this performance simple sentence to be tested and assessed, sequence of notes or the pitch sequence of described characteristic sequence to be measured performance simple sentence to be tested and assessed described in being.Preferably, after the performance simple sentence performance of testing and assessing finishes, obtain the characteristic sequence to be measured of this performance simple sentence to be tested and assessed described.
S102, the order according to described performance simple sentence in described target audio file, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence.
Described source audio file can be arranged and form with reference to simple sentence order by least one.In this step, the order in described target audio file is identical with performance simple sentence described to be tested and assessed for the order of the reference simple sentence of locating in described source audio file.According to the above-mentioned example of the present embodiment, if taking song A as target audio file, the original singer song B of song A in the time publishing and distributing is source audio file, the order of performance simple sentence described to be tested and assessed is 3, the order of the reference simple sentence that song B locates is also 3, from song B, choose the 3rd with reference to simple sentence the test and appraisal benchmark as performance simple sentence described to be tested and assessed.
In a kind of feasible embodiment of the embodiment of the present invention, described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, and described fixed reference feature sequence is the described sequence of notes with reference to simple sentence.In the feasible embodiment of the another kind of the embodiment of the present invention, described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, and institute's fixed reference feature sequence is the described pitch sequence with reference to simple sentence.
S103, carries out related operation to described fixed reference feature sequence and described characteristic sequence to be measured, obtains related coefficient sequence.
Because described fixed reference feature sequence can be used for characterizing the melody characteristics of the reference simple sentence of locating in source audio file, described characteristic sequence to be measured can be used for characterizing the melody characteristics of performance simple sentence to be tested and assessed in target audio file, this step can, to the related operation between described fixed reference feature sequence and described characteristic sequence to be measured, obtain related coefficient sequence.
S104, according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence.
In this step, described in the test and appraisal mark of performance simple sentence to be tested and assessed higher, described in showing, the singing effect of performance simple sentence to be tested and assessed is better, the singing effect of its more approaching located reference simple sentence.Otherwise, described in the test and appraisal mark of performance simple sentence to be tested and assessed lower, described in showing, the singing effect of performance simple sentence to be tested and assessed is poorer, it more departs from the singing effect of located reference simple sentence.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Below in conjunction with accompanying drawing 2-accompanying drawing 4, the each step in the audio frequency assessment method shown in accompanying drawing 1 is described in detail.It should be noted that, during subsequent figure 2-is embodiment illustrated in fig. 4, the reference simple sentence of locating comprises N audio frame, and described fixed reference feature sequence can be expressed as p (i); Wherein, i is integer, and 0≤i≤N-1.Particularly, if described fixed reference feature sequence is the described sequence of notes with reference to simple sentence, the note of first audio frame in the reference simple sentence that p (0) represents to locate, the note of second audio frame in the reference simple sentence that p (1) represents to locate, the note of N audio frame in the reference simple sentence that by that analogy, p (N-1) represents to locate.If described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, the pitch of first audio frame in the reference simple sentence that p (0) represents to locate, the note of second audio frame in the reference simple sentence that p (1) represents to locate, the note of N audio frame in the reference simple sentence that by that analogy, p (N-1) represents to locate.
Set performance simple sentence to be tested and assessed and comprise N audio frame, described characteristic sequence to be measured can be expressed as s (i), and wherein, i is integer, and 0≤i≤N-1.Particularly, if described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, s (0) represent described in the note of first audio frame in performance simple sentence to be tested and assessed, the note of second audio frame described in s (1) represents in performance simple sentence to be tested and assessed, by that analogy, s (N-1) represent described in the note of N audio frame in performance simple sentence to be tested and assessed.If described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, s (0) represent described in the pitch of first audio frame in performance simple sentence to be tested and assessed, the pitch of second audio frame described in s (1) represents in performance simple sentence to be tested and assessed, by that analogy, s (N-1) represent described in the pitch of N audio frame in performance simple sentence to be tested and assessed.
Referring to Fig. 2, is the process flow diagram of an embodiment of the step S103 shown in Fig. 1; This step S103 can comprise the following steps s1001-step s1004.
S1001, calculates respectively the average of described fixed reference feature sequence and the average of described characteristic sequence to be measured.
In this step, can adopt following formula (1) to calculate respectively the average of described fixed reference feature sequence p (i) and the average of described characteristic sequence s to be measured (i), this formula (1) is as follows:
MP＝mean(p(i))
MS＝mean(s(i)) (1)
In above-mentioned formula (1), MP represents the average of described fixed reference feature sequence p (i), and MS represents the average of described fixed reference feature sequence p (i), and mean () is the operation of averaging.
S1002, adopts the average of described fixed reference feature sequence, and described fixed reference feature sequence is carried out to regular processing, adopts the average of described characteristic sequence to be measured, and described characteristic sequence to be measured is carried out to regular processing.
The object of regular processing is: described fixed reference feature sequence and described characteristic sequence to be measured are adjusted to same benchmark, to eliminate described fixed reference feature sequence and described characteristic sequence to be measured because average is asked for the inconsistent calculation deviation impact being brought of standard.
In this step, can adopt formula (2) to carry out regular processing to described fixed reference feature sequence, this formula (2) can be expressed as follows:
p2(i)＝p(i)-MP (2)
In above-mentioned formula (2), p2 (i) represents the fixed reference feature sequence obtaining after regular processing.
In this step, can adopt formula (3) to carry out regular processing to described characteristic sequence to be measured, this formula (3) can be expressed as follows:
s2(i)＝s(i)-MS (3)
In above-mentioned formula (3), s2 (i) represents the characteristic sequence to be measured obtaining after regular processing.
S1003, adopts default slicing threshold value, and regular described fixed reference feature sequence after treatment is converted to referential data sequence, and regular described characteristic sequence to be measured after treatment is converted to sequence of values to be measured.
Wherein, described default slicing threshold value can be set according to actual needs, and preferably, described slicing threshold value can adopt formula (4) to set, and this formula (4) can be expressed as follows:
Th_xue＝max(max(abs(p2(i)),max(abs(s2(i))) (4)
In above-mentioned formula (4), Th_xue represents described default slicing threshold value, and max () asks for operation for maximal value, and abs () asks for operation for absolute value.
In this step, can adopt formula (5) that regular described fixed reference feature sequence after treatment is converted to referential data sequence, this formula (5) can be expressed as follows:
In above-mentioned formula (5), p3 (i) represents referential data sequence.
In this step, can adopt formula (6) that regular described characteristic sequence to be measured after treatment is converted to sequence of values to be measured, this formula (6) can be expressed as follows:
In above-mentioned formula (6), s3 (i) represents sequence of values to be measured.
S1004, adopts cross correlation function to carry out related operation to described referential data sequence and described sequence of values to be measured, obtains related coefficient sequence.
In a kind of feasible embodiment of this step, can adopt formula (7) to carry out related operation to described referential data sequence and described sequence of values to be measured, this formula (7) can be expressed as follows:
In above-mentioned formula (7), R (n) represents related coefficient sequence; " " represents multiplication operations; S3 (i-n) represents the sequence that s3 (i) loopy moving n forms, wherein, and 0≤n≤N-1.
In the feasible embodiment of the another kind of this step, can adopt formula (8) to carry out related operation to described referential data sequence and described sequence of values to be measured, this formula (8) can be expressed as follows:
In above-mentioned formula (8), R (n) represents related coefficient sequence; " " represents multiplication operations; P3 (i-n) represents the sequence that p3 (i) loopy moving n forms, wherein, and 0≤n≤N-1.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Referring to Fig. 3, is the process flow diagram of an embodiment of the step S104 shown in Fig. 1; This step S104 can comprise the following steps s2001-step s2002.
S2001, calculates the maximal value of described related coefficient sequence.
In this step, can adopt following formula (9) to calculate the maximal value of described related coefficient sequence, this formula (9) can be expressed as follows:
RMAX＝max(R(n)) (9)
In above-mentioned formula (9), R (n) represents related coefficient sequence; Max () asks for operation for maximal value; RMAX represents the maximal value of described related coefficient sequence.
S2002, by the maximal value of described related coefficient sequence, is defined as the test and appraisal mark of described performance simple sentence.
This step can directly be defined as the maximal value of described related coefficient sequence the test and appraisal mark of described performance simple sentence to be tested and assessed, the value that the test and appraisal mark of described performance simple sentence is RMAX.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Referring to Fig. 4, is the process flow diagram of another embodiment of the step S104 shown in Fig. 1; This step S104 can comprise the following steps s3001-step s3003.
S3001, calculates the maximal value of described related coefficient sequence.
This step can embodiment shown in Figure 3 step s2001, be not repeated herein.
S3002, maps to preset fraction interval by the maximal value of described related coefficient sequence, obtains the peaked mapping value of described related coefficient sequence.
Described preset fraction interval can be set according to actual needs, for example: described preset fraction interval can be set as [0,10]; Or described preset fraction interval can be set as [0,100].In this step, can adopt [score_min, score_max] represent described preset fraction interval, by linearity or nonlinear method, the maximal value RMAX of described related coefficient sequence is mapped to described preset fraction interval, the mapping value obtaining can represent score, this score is positioned in the preset fraction interval shown in [score_min, score_max].
S3003, is defined as described mapping value the scoring of described performance simple sentence.
This step can be defined as described mapping value score the test and appraisal mark of described performance simple sentence to be tested and assessed, the value that the test and appraisal mark of described performance simple sentence is score.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Below in conjunction with Fig. 5-Fig. 8, a kind of audio frequency assessment device that the embodiment of the present invention is provided describes in detail.It should be noted that, the audio frequency assessment device described in Fig. 5-Fig. 8 can be applicable to carry out the method shown in above-mentioned accompanying drawing 1-accompanying drawing 4.In practical application, described audio frequency assessment device can run on server end, or runs on such as in notebook computer, mobile phone, PAD (panel computer), intelligent wearable device etc. terminal.
Refer to Fig. 5, the structural representation of a kind of audio frequency assessment device providing for the embodiment of the present invention; This device can comprise: acquisition module 101 to be measured, with reference to acquisition module 102, related operation module 103 and test and appraisal module 104.
Acquisition module 101 to be measured, for obtaining performance simple sentence to be tested and assessed at the order of target audio file and the characteristic sequence to be measured of described performance simple sentence.
Note is again note, refers to the symbol of the carrying out of the sound for recording different length, can comprise whole note, minim, crotchet, quaver etc. kind.Audio frequency simple sentence can be expressed as a frame sequence of multiple audio frames compositions, and each audio frame all carries note, and according to each audio frame, the order of the time order and function in this audio frequency simple sentence forms the melody of this audio frequency simple sentence to each note.Pitch is again pitch, refers to the height of sound.Audio frequency simple sentence can be expressed as a frame sequence of multiple audio frames compositions, and each audio frame all carries pitch, and according to each audio frame, the order of the time order and function in this audio frequency simple sentence forms the melody of this audio frequency simple sentence to each pitch.To sum up, the sequence of notes of audio frequency simple sentence or pitch sequence all can reflect the melody characteristics of this audio frequency simple sentence.
In the embodiment of the present invention, preferably, performance simple sentence to be tested and assessed can be current performance simple sentence, and now audio frequency test and appraisal process can be real-time audio test and appraisal process.Wherein, current performance simple sentence refers to corresponding performance simple sentence of current time of playing in described target audio file.Target audio file is song A, the example that is described as with above-mentioned song A: suppose song A totally 5 minutes, the current time of playing is 1895ms, according to the description of song A, 1895ms belonged in the described time period of time attribute of audio frequency simple sentence " cccccccc ", can determine that thus audio frequency simple sentence " cccccccc " is current performance simple sentence, further, can determine that audio frequency simple sentence " cccccccc " is performance simple sentence to be tested and assessed, described in performance simple sentence the to be tested and assessed order in target audio file be 3.
Described acquisition module to be measured 101 can obtain the characteristic sequence to be measured of this performance simple sentence to be tested and assessed, sequence of notes or the pitch sequence of described characteristic sequence to be measured performance simple sentence to be tested and assessed described in being.Preferably, described acquisition module 101 to be measured after the performance simple sentence performance of testing and assessing finishes, obtains the characteristic sequence to be measured of this performance simple sentence to be tested and assessed described.
With reference to acquisition module 102, for the order at described target audio file according to described performance simple sentence, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence.
Described source audio file can be arranged and form with reference to simple sentence order by least one.The order in described target audio file is identical with performance simple sentence described to be tested and assessed for the order of described reference simple sentence of locating with reference to acquisition module 102 in described source audio file.According to the above-mentioned example of the present embodiment, if taking song A as target audio file, the original singer song B of song A in the time publishing and distributing is source audio file, the order of performance simple sentence described to be tested and assessed is 3, the order of the reference simple sentence that song B locates is also 3, from song B, choose the 3rd with reference to simple sentence the test and appraisal benchmark as performance simple sentence described to be tested and assessed.
In a kind of feasible embodiment of the embodiment of the present invention, described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, and described fixed reference feature sequence is the described sequence of notes with reference to simple sentence.In the feasible embodiment of the another kind of the embodiment of the present invention, described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, and institute's fixed reference feature sequence is the described pitch sequence with reference to simple sentence.
Related operation module 103, for described fixed reference feature sequence and described characteristic sequence to be measured are carried out to related operation, obtains related coefficient sequence.
Because described fixed reference feature sequence can be used for characterizing the melody characteristics of the reference simple sentence of locating in source audio file, described characteristic sequence to be measured can be used for characterizing the melody characteristics of performance simple sentence to be tested and assessed in target audio file, described related operation module 103 can, to the related operation between described fixed reference feature sequence and described characteristic sequence to be measured, obtain related coefficient sequence.
Test and appraisal module 104, for according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence.
The test and appraisal mark of performance simple sentence described to be tested and assessed is higher, and described in showing, the singing effect of performance simple sentence to be tested and assessed is better, the singing effect of its more approaching located reference simple sentence.Otherwise, described in the test and appraisal mark of performance simple sentence to be tested and assessed lower, described in showing, the singing effect of performance simple sentence to be tested and assessed is poorer, it more departs from the singing effect of located reference simple sentence.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Below in conjunction with accompanying drawing 6-accompanying drawing 8, the 26S Proteasome Structure and Function of the each module in the audio frequency assessment device shown in accompanying drawing 5 is described in detail.It should be noted that, during subsequent figure 6-is embodiment illustrated in fig. 8, the reference simple sentence of locating comprises N audio frame, and described fixed reference feature sequence can be expressed as p (i); Wherein, i is integer, and 0≤i≤N-1.Particularly, if described fixed reference feature sequence is the described sequence of notes with reference to simple sentence, the note of first audio frame in the reference simple sentence that p (0) represents to locate, the note of second audio frame in the reference simple sentence that p (1) represents to locate, the note of N audio frame in the reference simple sentence that by that analogy, p (N-1) represents to locate.If described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, the pitch of first audio frame in the reference simple sentence that p (0) represents to locate, the note of second audio frame in the reference simple sentence that p (1) represents to locate, the note of N audio frame in the reference simple sentence that by that analogy, p (N-1) represents to locate.
Set performance simple sentence to be tested and assessed and comprise N audio frame, described characteristic sequence to be measured can be expressed as s (i), and wherein, i is integer, and 0≤i≤N-1.Particularly, if described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, s (0) represent described in the note of first audio frame in performance simple sentence to be tested and assessed, the note of second audio frame described in s (1) represents in performance simple sentence to be tested and assessed, by that analogy, s (N-1) represent described in the note of N audio frame in performance simple sentence to be tested and assessed.If described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, s (0) represent described in the pitch of first audio frame in performance simple sentence to be tested and assessed, the pitch of second audio frame described in s (1) represents in performance simple sentence to be tested and assessed, by that analogy, s (N-1) represent described in the pitch of N audio frame in performance simple sentence to be tested and assessed.
Referring to Fig. 6, is the structural representation of the embodiment of the related operation module shown in Fig. 5; Described related operation module 103 can comprise: mean value computation unit 1301, regular processing unit 1302, sequence converting unit 1303 and related operation unit 1304.
Mean value computation unit 1301, for calculating respectively the average of described fixed reference feature sequence and the average of described characteristic sequence to be measured.
Described mean value computation unit 1301 can adopt the formula (1) in embodiment illustrated in fig. 2 to calculate respectively the average MP of described fixed reference feature sequence p (i) and the average MS of described characteristic sequence s to be measured (i).
Regular processing unit 1302, for adopting the average of described fixed reference feature sequence, carries out regular processing to described fixed reference feature sequence, adopts the average of described characteristic sequence to be measured, and described characteristic sequence to be measured is carried out to regular processing.
The object of regular processing is: described fixed reference feature sequence and described characteristic sequence to be measured are adjusted to same benchmark, to eliminate described fixed reference feature sequence and described characteristic sequence to be measured because average is asked for the inconsistent calculation deviation impact being brought of standard.Described regular processing unit 1302 can adopt the formula (2) in embodiment illustrated in fig. 2 to carry out regular processing to described fixed reference feature sequence, obtains the fixed reference feature sequence p2 (i) obtaining after regular processing; And can adopt the formula (3) in embodiment illustrated in fig. 2 to carry out regular processing to described characteristic sequence to be measured, obtain the characteristic sequence s2 to be measured (i) obtaining after regular processing.
Sequence converting unit 1303, for adopting default slicing threshold value, is converted to referential data sequence by regular described fixed reference feature sequence after treatment, and regular described characteristic sequence to be measured after treatment is converted to sequence of values to be measured.
Wherein, described default slicing threshold value can be set according to actual needs, and preferably, described slicing threshold value Th_xue can adopt the formula (4) in embodiment illustrated in fig. 2 to set.Described sequence converting unit 1303 can adopt the formula (5) in embodiment illustrated in fig. 2 that regular described fixed reference feature sequence after treatment is converted to referential data sequence p3 (i); And can adopt the formula (6) in embodiment illustrated in fig. 2 that regular described characteristic sequence to be measured after treatment is converted to sequence of values s3 to be measured (i).
Related operation unit 1304, for adopting cross correlation function to carry out related operation to described referential data sequence and described sequence of values to be measured, obtains related coefficient sequence.
In a kind of feasible embodiment of the present embodiment, described related operation unit 1304 can adopt the formula (7) in embodiment illustrated in fig. 2 to carry out related operation to described referential data sequence and described sequence of values to be measured, obtains related coefficient sequence R (n).In the feasible embodiment of the another kind of the present embodiment, described related operation unit 1304 can adopt the formula (8) in embodiment illustrated in fig. 2 to carry out related operation to described referential data sequence and described sequence of values to be measured, obtains related coefficient sequence R (n).
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Referring to Fig. 7, is the structural representation of an embodiment of the test and appraisal module shown in Fig. 5; Described test and appraisal module 104 can comprise: the first computing unit 1401 and the first test and appraisal unit 1402.
The first computing unit 1401, for calculating the maximal value of described related coefficient sequence.
Described the first computing unit 1401 can adopt the formula (9) in embodiment illustrated in fig. 3 to calculate the maximal value RMAX of described related coefficient sequence.
The first test and appraisal unit 1402, for by the maximal value of described related coefficient sequence, is defined as the test and appraisal mark of described performance simple sentence.
Described the first test and appraisal unit 1402 can directly be defined as the maximal value of described related coefficient sequence the test and appraisal mark of described performance simple sentence to be tested and assessed, the value that the test and appraisal mark of described performance simple sentence is RMAX.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
Referring to Fig. 8, is the structural representation of another embodiment of the test and appraisal module shown in Fig. 5; Described test and appraisal module 104 can comprise: the second computing unit 1411, map unit 1412 and the second test and appraisal unit 1413.
The second computing unit 1411, for calculating the maximal value of described related coefficient sequence.
The computation process of described first computing unit 1401 of the computation process of described the second computing unit 1411 in can embodiment shown in Figure 7, is not repeated herein.
Map unit 1412, for the maximal value of described related coefficient sequence is mapped to preset fraction interval, obtains the peaked mapping value of described related coefficient sequence.
Described preset fraction interval can be set according to actual needs, for example: described preset fraction interval can be set as [0,10]; Or described preset fraction interval can be set as [0,100].Described map unit 1412 can adopt [score_min, score_max] represent described preset fraction interval, by linearity or nonlinear method, the maximal value RMAX of described related coefficient sequence is mapped to described preset fraction interval, the mapping value obtaining can represent score, this score is positioned in the preset fraction interval shown in [score_min, score_max].
The second test and appraisal unit 1413, for being defined as described mapping value the test and appraisal mark of described performance simple sentence.
Described the second test and appraisal unit 1413 can be defined as described mapping value score the test and appraisal mark of described performance simple sentence to be tested and assessed, the value that the test and appraisal mark of described performance simple sentence is score.
In the embodiment of the present invention, can obtain the characteristic sequence to be measured of performance simple sentence to be tested and assessed in target audio file, and obtain the fixed reference feature sequence of the reference simple sentence in source audio file, obtain related coefficient sequence by the related operation to characteristic sequence to be measured and fixed reference feature sequence, the test and appraisal of the performance simple sentence to target audio file have been realized based on this related coefficient sequence, both met user for the actual demand in the application process of audio file, the application that has promoted again audio file is intelligent.
One of ordinary skill in the art will appreciate that all or part of flow process realizing in above-described embodiment method, can carry out the hardware that instruction is relevant by computer program to complete, described program can be stored in a computer read/write memory medium, this program, in the time carrying out, can comprise as the flow process of the embodiment of above-mentioned each side method.Wherein, described storage medium can be magnetic disc, CD, read-only store-memory body (Read-Only Memory, ROM) or random store-memory body (Random Access Memory, RAM) etc.
Above disclosed is only preferred embodiment of the present invention, certainly can not limit with this interest field of the present invention, and the equivalent variations of therefore doing according to the claims in the present invention, still belongs to the scope that the present invention is contained.
Claims (10)
1. an audio frequency assessment method, is characterized in that, comprising:
Obtain performance simple sentence the to be tested and assessed order in target audio file and the characteristic sequence to be measured of described performance simple sentence;
Order according to described performance simple sentence in described target audio file, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence;
Described fixed reference feature sequence and described characteristic sequence to be measured are carried out to related operation, obtain related coefficient sequence;
According to described related coefficient sequence, determine the test and appraisal mark of described performance simple sentence.
2. the method for claim 1, is characterized in that, described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, and described fixed reference feature sequence is the described sequence of notes with reference to simple sentence; Or,
Described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, and institute's fixed reference feature sequence is the described pitch sequence with reference to simple sentence.
3. the method for claim 1, is characterized in that, described described fixed reference feature sequence and described characteristic sequence to be measured is carried out to related operation, obtains related coefficient sequence, comprising:
Calculate respectively the average of described fixed reference feature sequence and the average of described characteristic sequence to be measured;
Adopt the average of described fixed reference feature sequence, described fixed reference feature sequence is carried out to regular processing, adopt the average of described characteristic sequence to be measured, described characteristic sequence to be measured is carried out to regular processing;
Adopt default slicing threshold value, regular described fixed reference feature sequence after treatment is converted to referential data sequence, regular described characteristic sequence to be measured after treatment is converted to sequence of values to be measured;
Adopt cross correlation function to carry out related operation to described referential data sequence and described sequence of values to be measured, obtain related coefficient sequence.
4. the method as described in claim 1-3 any one, is characterized in that, described according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence, comprising:
Calculate the maximal value of described related coefficient sequence;
By the maximal value of described related coefficient sequence, be defined as the test and appraisal mark of described performance simple sentence.
5. the method as described in claim 1-3 any one, is characterized in that, described according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence, comprising:
Calculate the maximal value of described related coefficient sequence;
The maximal value of described related coefficient sequence is mapped to preset fraction interval, obtain the peaked mapping value of described related coefficient sequence;
Described mapping value is defined as to the test and appraisal mark of described performance simple sentence.
6. an audio frequency assessment device, is characterized in that, comprising:
Acquisition module to be measured, for obtaining performance simple sentence to be tested and assessed at the order of target audio file and the characteristic sequence to be measured of described performance simple sentence;
With reference to acquisition module, for the order at described target audio file according to described performance simple sentence, position reference simple sentence in source audio file, and obtain the described fixed reference feature sequence with reference to simple sentence;
Related operation module, for described fixed reference feature sequence and described characteristic sequence to be measured are carried out to related operation, obtains related coefficient sequence;
Test and appraisal module, for according to described related coefficient sequence, determines the test and appraisal mark of described performance simple sentence.
7. device as claimed in claim 6, is characterized in that, described characteristic sequence to be measured is the sequence of notes of described performance simple sentence, and described fixed reference feature sequence is the described sequence of notes with reference to simple sentence; Or,
Described characteristic sequence to be measured is the pitch sequence of described performance simple sentence, and institute's fixed reference feature sequence is the described pitch sequence with reference to simple sentence.
8. device as claimed in claim 6, is characterized in that, described related operation module comprises:
Mean value computation unit, for calculating respectively the average of described fixed reference feature sequence and the average of described characteristic sequence to be measured;
Regular processing unit, for adopting the average of described fixed reference feature sequence, carries out regular processing to described fixed reference feature sequence, adopts the average of described characteristic sequence to be measured, and described characteristic sequence to be measured is carried out to regular processing;
Sequence converting unit, for adopting default slicing threshold value, is converted to referential data sequence by regular described fixed reference feature sequence after treatment, and regular described characteristic sequence to be measured after treatment is converted to sequence of values to be measured;
Related operation unit, for adopting cross correlation function to carry out related operation to described referential data sequence and described sequence of values to be measured, obtains related coefficient sequence.
9. the device as described in claim 6-8 any one, is characterized in that, described test and appraisal module comprises:
The first computing unit, for calculating the maximal value of described related coefficient sequence;
The first test and appraisal unit, for by the maximal value of described related coefficient sequence, is defined as the test and appraisal mark of described performance simple sentence.
10. the device as described in claim 6-8 any one, is characterized in that, described test and appraisal module comprises:
The second computing unit, for calculating the maximal value of described related coefficient sequence;
Map unit, for the maximal value of described related coefficient sequence is mapped to preset fraction interval, obtains the peaked mapping value of described related coefficient sequence;
The second test and appraisal unit, for being defined as described mapping value the test and appraisal mark of described performance simple sentence.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410362595.5A CN104157296B (en) | 2014-07-28 | 2014-07-28 | A kind of audio frequency assessment method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410362595.5A CN104157296B (en) | 2014-07-28 | 2014-07-28 | A kind of audio frequency assessment method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104157296A true CN104157296A (en) | 2014-11-19 |
| CN104157296B CN104157296B (en) | 2016-04-27 |
Family
ID=51882778
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410362595.5A Active CN104157296B (en) | 2014-07-28 | 2014-07-28 | A kind of audio frequency assessment method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104157296B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104464754A (en) * | 2014-12-11 | 2015-03-25 | 北京中细软移动互联科技有限公司 | Sound brand search method |
| WO2016201959A1 (en) * | 2015-06-15 | 2016-12-22 | 福建星网视易信息系统有限公司 | Method of playing back multimedia file on the basis of singing score and device utilizing same |
| CN109979482A (en) * | 2019-05-21 | 2019-07-05 | 科大讯飞股份有限公司 | A kind of evaluating method and device for audio |
| CN115171729A (en) * | 2022-09-02 | 2022-10-11 | 北京达佳互联信息技术有限公司 | Audio quality determination method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430876A (en) * | 2007-11-08 | 2009-05-13 | 中国科学院声学研究所 | Singing marking system and method |
| CN101441865A (en) * | 2007-11-19 | 2009-05-27 | 盛趣信息技术（上海）有限公司 | Method and system for grading sing genus game |
| WO2014043815A1 (en) * | 2012-09-24 | 2014-03-27 | Hitlab Inc. | A method and system for assessing karaoke users |
-
2014
- 2014-07-28 CN CN201410362595.5A patent/CN104157296B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430876A (en) * | 2007-11-08 | 2009-05-13 | 中国科学院声学研究所 | Singing marking system and method |
| CN101441865A (en) * | 2007-11-19 | 2009-05-27 | 盛趣信息技术（上海）有限公司 | Method and system for grading sing genus game |
| WO2014043815A1 (en) * | 2012-09-24 | 2014-03-27 | Hitlab Inc. | A method and system for assessing karaoke users |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104464754A (en) * | 2014-12-11 | 2015-03-25 | 北京中细软移动互联科技有限公司 | Sound brand search method |
| WO2016201959A1 (en) * | 2015-06-15 | 2016-12-22 | 福建星网视易信息系统有限公司 | Method of playing back multimedia file on the basis of singing score and device utilizing same |
| CN109979482A (en) * | 2019-05-21 | 2019-07-05 | 科大讯飞股份有限公司 | A kind of evaluating method and device for audio |
| CN115171729A (en) * | 2022-09-02 | 2022-10-11 | 北京达佳互联信息技术有限公司 | Audio quality determination method and device, electronic equipment and storage medium |
| CN115171729B (en) * | 2022-09-02 | 2023-01-20 | 北京达佳互联信息技术有限公司 | Audio quality determination method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104157296B (en) | 2016-04-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7812241B2 (en) | Methods and systems for identifying similar songs | |
| JP4322283B2 (en) | Performance determination device and program | |
| CN104157296A (en) | Audio frequency evaluative method and device | |
| CN105023559A (en) | Karaoke processing method and system | |
| CN104620313A (en) | Audio signal analysis | |
| Molina et al. | Fundamental frequency alignment vs. note-based melodic similarity for singing voice assessment | |
| CN104885153A (en) | Apparatus and method for correcting audio data | |
| CN104282316A (en) | Karaoke scoring method based on voice matching, and device thereof | |
| CN102568456A (en) | Notation recording method and a notation recording device based on humming input | |
| CN108206027A (en) | A kind of audio quality evaluation method and system | |
| CN106653037A (en) | Audio data processing method and device | |
| CN102568457A (en) | Music synthesis method and device based on humming input | |
| CN104143340A (en) | Voice frequency evaluation method and device | |
| JP5395399B2 (en) | Mobile terminal, beat position estimating method and beat position estimating program | |
| CN113053365A (en) | Voice separation method, device, equipment and storage medium | |
| CN104143339A (en) | Music signal processing apparatus and method, and program | |
| CN113096689B (en) | Song singing evaluation method, device and medium | |
| CN104217731A (en) | Quick solo music score recognizing method | |
| Sheng et al. | Automatic control of the dynamic range compressor using a regression model and a reference sound | |
| CN105788609A (en) | Method and device of associating a plurality of sound sources, evaluation method, and evaluation system | |
| CN109271501A (en) | A kind of management method and system of audio database | |
| CN104882146A (en) | Method and device for processing audio popularization information | |
| CN103531220B (en) | Lyrics bearing calibration and device | |
| Pang et al. | Automatic detection of vibrato in monophonic music | |
| CN105590633A (en) | Method and device for generation of labeled melody for song scoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20161124 Address after: 510000 Guangzhou, Tianhe District branch Yun Yun Road, No. 16, self built room 2, building 1301 Patentee after: Guangzhou KuGou Networks Co., Ltd. Address before: Shenzhen Futian District City, Guangdong province 518000 Zhenxing Road, SEG Science Park 2 East Room 403 Patentee before: Tencent Technology (Shenzhen) Co., Ltd. |
|
| CP02 | Change in the address of a patent holder | ||
| CP02 | Change in the address of a patent holder |
Address after: No. 315, Huangpu Avenue middle, Tianhe District, Guangzhou City, Guangdong Province Patentee after: GUANGZHOU KUGOU COMPUTER TECHNOLOGY Co.,Ltd. Address before: 510000 room 1301, building 2, No.16 Keyun Road, Tianhe District, Guangzhou City, Guangdong Province Patentee before: GUANGZHOU KUGOU COMPUTER TECHNOLOGY Co.,Ltd. |