CN108831504A - Determination method, apparatus, computer equipment and the storage medium of pitch period - Google Patents
Determination method, apparatus, computer equipment and the storage medium of pitch period Download PDFInfo
- Publication number
- CN108831504A CN108831504A CN201810608431.4A CN201810608431A CN108831504A CN 108831504 A CN108831504 A CN 108831504A CN 201810608431 A CN201810608431 A CN 201810608431A CN 108831504 A CN108831504 A CN 108831504A
- Authority
- CN
- China
- Prior art keywords
- audio signal
- pitch period
- auto
- measured
- under
- 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
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000005236 sound signal Effects 0.000 claims abstract description 257
- 238000005311 autocorrelation function Methods 0.000 claims abstract description 127
- 238000010606 normalization Methods 0.000 claims abstract description 82
- 238000006243 chemical reaction Methods 0.000 claims abstract description 37
- 238000011156 evaluation Methods 0.000 claims description 66
- 230000006870 function Effects 0.000 claims description 37
- 238000004590 computer program Methods 0.000 claims description 33
- 230000008859 change Effects 0.000 claims description 8
- 238000012937 correction Methods 0.000 claims description 6
- 238000005314 correlation function Methods 0.000 claims description 3
- 230000011664 signaling Effects 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 claims 4
- 230000002596 correlated effect Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 6
- 238000001514 detection method Methods 0.000 description 4
- 238000012986 modification Methods 0.000 description 4
- 238000006073 displacement reaction Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 210000001260 vocal cord Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/03—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters
- G10L25/06—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 characterised by the type of extracted parameters the extracted parameters being correlation coefficients
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/90—Pitch determination of speech signals
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
- Auxiliary Devices For Music (AREA)
Abstract
This application involves determination method, apparatus, computer equipment and the storage mediums of a kind of pitch period.The method includes:Obtain under preset each candidate pitch period, audio signal to be measured and it is each offset audio signal between normalization amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to the signal obtained after candidate pitch period offset;To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains the auto-correlated error under each candidate pitch period;The extreme value conversion operation is used to the maximum of the auto-correlation function sum being converted to minimum;According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target pitch period is determined.It can be improved the accuracy of pitch period using this method.
Description
Technical field
The present invention relates to fields of communication technology, set more particularly to a kind of determination method, apparatus of pitch period, computer
Standby and storage medium.
Background technique
Pitch period is the every unlatching of vocal cords and is closed the primary time, and pitch period exists as a kind of feature of voice signal
The fields such as voice coding, identification are widely used.
Currently, common Periodical pitch detection method is mainly Autocorrelation Detection method (auto correlation
Function, ACF), the principle of ACF is the original audio signal and each offset audio letter calculated under each candidate pitch period
Auto-correlation function between number and, maximum auto-correlation function and corresponding candidate pitch period are determined as the original audio signal
Pitch period.
But the pitch period accuracy that the above method determines is lower.
Summary of the invention
Based on this, it is necessary in view of the above technical problems, provide a kind of fundamental tone week that can be improved pitch period accuracy
Determination method, apparatus, computer equipment and the storage medium of phase.
A kind of determination method of pitch period, the method includes:
It obtains under preset each candidate pitch period, returning between audio signal to be measured and each offset audio signal
One change amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to described
The signal obtained after candidate pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each time
Select the auto-correlated error under pitch period;The extreme value conversion operation is for being converted to the maximum of the auto-correlation function sum
Minimum;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone is determined
Period.
The auto-correlation function under each candidate pitch period and execution extreme value in one of the embodiments,
Conversion operation obtains the auto-correlated error under each candidate pitch period, including:
According to the audio signal to be measured, each auto-correlation function and, obtain under each candidate pitch period
Auto-correlated error.
In one of the embodiments, it is described according to the audio signal to be measured, each auto-correlation function and, obtain
Auto-correlated error under each candidate pitch period, including:
According to formulaCalculate each candidate pitch
Auto-correlated error E under periodn(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, RnIt (k) is the candidate pitch week
The auto-correlation function of phase k and, w (n) is signal weighting window, and N is window function length, and α is the noise correction factor.
In one of the embodiments, the normalization amplitude difference energy according under each candidate pitch period and
Auto-correlated error determines target pitch period, including:
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, pitch evaluation is determined
Set;It include the pitch evaluation value under each candidate pitch period in the pitch evaluation set;
Determine that the corresponding candidate pitch period of minimum fundamental tone estimated value in the pitch evaluation set is the target base
The sound period.
In one of the embodiments, the normalization amplitude difference energy according under each candidate pitch period and
Auto-correlated error determines pitch evaluation set, including:
According to formula Pn(k)=β * Dn(k)+γ*En(k), the pitch evaluation value under each candidate pitch period is calculated
Pn(k),
Wherein, k is the candidate pitch period, DnIt (k) is the normalization amplitude difference energy under the candidate pitch period k
Amount, EnIt (k) is the auto-correlated error under the candidate pitch period k, β and γ are weighted factor ,+γ=1.0 β;
According to the pitch evaluation value under each candidate pitch period, the pitch evaluation set is determined.
In one of the embodiments, it is described acquisition under preset each candidate pitch period, audio signal to be measured with
Normalization amplitude difference energy between each offset audio signal, including:
According to the audio signal to be measured and each offset audio signal, the audio signal to be measured and each is determined
Normalization amplitude difference energy between the offset audio signal.
It is described according to the audio signal to be measured and each inclined frequency-shift signaling in one of the embodiments, it determines
Normalization amplitude difference energy between the audio signal to be measured and each offset audio signal, including:
According to formulaCalculate the audio signal to be measured and each institute
State the normalization amplitude difference energy D between offset audio signaln(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, and s (n+k) is the offset audio
Signal, N are window function length.
In one of the embodiments, it is described acquisition under preset each candidate pitch period, audio signal to be measured with
Auto-correlation function between each offset audio signal and, including:
According to the audio signal to be measured and each first offset audio signal, the audio signal to be measured and each is determined
It is described first offset audio signal between the first auto-correlation function and;The first offset audio signal is the audio to be measured
The signal that signal obtains after deviating according to the candidate pitch period;
According to the audio signal to be measured and each second offset audio signal, the audio signal to be measured and each is determined
It is described second offset audio signal between the second auto-correlation function and;The second offset audio signal is the audio to be measured
Signal is according to the signal obtained after the frequency multiplication offset of the candidate pitch period;
And, determined in each candidate according to each first auto-correlation function and each second auto-correlation function
Under pitch period, the corresponding auto-correlation function of the audio signal to be measured and.
It is described according to each first auto-correlation function and each second auto-correlation letter in one of the embodiments,
Number and, determine under each candidate pitch period, the corresponding auto-correlation function of the audio signal to be measured and, including:
According to formulaIt calculates in each candidate
Under pitch period, the corresponding auto-correlation function of the audio signal to be measured and Rn(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, and t is the candidate pitch period
Frequency, when t=1, s (n+t*k) is the first offset audio signal, t>When 1, s (n+t*k) is the second offset audio
Signal, w (n) and w (n+t*k) are signal weighting window, and T is maximum candidate pitch period, and M is that weighting windows are long.
A kind of determining device of pitch period, including:
Module is obtained, for obtaining under preset each candidate pitch period, audio signal to be measured and each offset sound
Normalization amplitude difference energy and auto-correlation function between frequency signal and;Wherein, the offset audio signal is described to acoustic
The signal that frequency signal obtains after deviating according to the candidate pitch period;
Conversion module, under each candidate pitch period auto-correlation function and execute extreme value conversion operation,
Obtain the auto-correlated error under each candidate pitch period;The extreme value conversion operation be used for the auto-correlation function and
Maximum be converted to minimum;
Determining module, for according to the normalization amplitude difference energy and auto-correlation mistake under each candidate pitch period
Difference determines target pitch period.
A kind of computer equipment, including memory and processor, the memory are stored with computer program, the processing
Device realizes following steps when executing the computer program:
It obtains under preset each candidate pitch period, returning between audio signal to be measured and each offset audio signal
One change amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to described
The signal obtained after candidate pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each time
Select the auto-correlated error under pitch period;The extreme value conversion operation is for being converted to the maximum of the auto-correlation function sum
Minimum;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone is determined
Period.
A kind of computer readable storage medium, is stored thereon with computer program, and the computer program is held by processor
Following steps are realized when row:
It obtains under preset each candidate pitch period, returning between audio signal to be measured and each offset audio signal
One change amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to described
The signal obtained after candidate pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each time
Select the auto-correlated error under pitch period;The extreme value conversion operation is for being converted to the maximum of the auto-correlation function sum
Minimum;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone is determined
Period.
Determination method, apparatus, computer equipment and the storage medium of above-mentioned pitch period are obtained in preset each candidate
Normalization amplitude difference energy and auto-correlation function under pitch period, between audio signal to be measured and each offset audio signal
With, under each candidate pitch period auto-correlation function and execute extreme value conversion operation, obtain under each candidate pitch period
Auto-correlated error target base is determined according to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period
The sound period, since extreme value conversion is for being converted to minimum for the maximum of auto-correlation function sum, so that auto-correlated error is to return
One change amplitude difference energy is presented as minimum at pitch period, avoids pitch period in normalization magnitude difference energy function
Peak feature difference bring error between auto-correlation function, accordingly, it is determined that pitch period accuracy it is relatively high.
Detailed description of the invention
Fig. 1 is a kind of determination method flow diagram for pitch period that one embodiment provides;
Fig. 2 is a kind of method flow diagram for calculating auto-correlation function sum that one embodiment provides;
Fig. 3 is a kind of flow chart of implementation of the step 103 in embodiment illustrated in fig. 1;
Fig. 4 is a kind of flow chart of implementation of the step 201 in embodiment illustrated in fig. 2;
Fig. 5 is a kind of block diagram of the determining device for pitch period that one embodiment provides;
Fig. 6 is a kind of structure chart for computer equipment that one embodiment provides.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, not
For limiting the application.
The determination method of pitch period provided by the present application, can be applied in Acoustic detection environment, detect voice signal
Pitch period.The executing subject of this method can be terminal, server etc..Wherein, terminal can be, but not limited to be various
People's computer, laptop, smart phone, tablet computer and portable wearable device, server can use independent clothes
The server cluster of business device either multiple servers composition is realized.
Fig. 1 is a kind of determination method flow diagram for pitch period that one embodiment provides.As shown in Figure 1, this method packet
Include following steps:
Step 101 obtains under preset each candidate pitch period, audio signal to be measured and each offset audio signal
Between normalization amplitude difference energy and auto-correlation function and;Wherein, offset audio signal is audio signal to be measured according to candidate
The signal obtained after pitch period offset.
In the present embodiment, audio signal to be measured is sample to original audio signal using preset sample frequency
The signal arrived.Search range can be predefined for each audio signal to be measured according to the statistical property of voice signal,
It include the frequency multiplication of multiple preset candidate pitch periods and each candidate pitch period in the search range.For example, preset adopt
Sample frequency is 8KHz, and the fundamental frequency of voice signal is usually 55hz-400hz, so searching for model under the sample frequency of 8KHz
20-144 can be determined as by enclosing.
After search range has been determined, according to each candidate pitch period in search range treat survey audio signal into
Line displacement obtains the corresponding offset audio signal of each candidate pitch period.Normalization amplitude difference predetermined can be used
Energy function calculates the normalization amplitude difference energy between audio signal to be measured and each offset audio signal, using pre-defined
Auto-correlation function calculate audio signal to be measured and it is each offset audio signal between auto-correlation function and, normalization amplitude difference
Energy function has different peak features from auto-correlation function, and pitch period is presented as on normalization magnitude difference energy function
Minimum, pitch period are presented as maximum on auto-correlation function.
It should be noted that the present embodiment for search range, calculate normalization amplitude difference energy and auto-correlation function and
Method it is without restriction.
Step 102, under each candidate pitch period auto-correlation function and execute extreme value conversion operation, obtain each time
Select the auto-correlated error under pitch period;Extreme value conversion operation is used to the maximum of auto-correlation function sum being converted to minimum.
In the present embodiment, to the corresponding auto-correlation function of each candidate pitch period and extreme value conversion behaviour can be executed
Make, obtains the auto-correlated error under each candidate pitch period, which is equivalent to pitch period in auto-correlation letter
Maximum on number is converted to minimum.
For example, auto-correlated error function can be pre-defined, it should be certainly by audio signal to be measured and auto-correlation function and input
Dependent error functions obtain auto-correlated error.Pitch period is presented as minimum on auto-correlated error function.It can also use
Maximum can be converted to the method for minimum to execute extreme value conversion operation by others, and the present embodiment is simultaneously without restriction.
Step 103, according to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, determine target
Pitch period.
In the present embodiment, according to the corresponding normalization amplitude difference energy of each candidate pitch period and auto-correlated error,
Determine target pitch period.For example, for normalization amplitude difference energy and auto-correlated error difference under each candidate pitch period
Weight is assigned, the weight of weight and auto-correlated error further according to the normalization amplitude difference energy under each candidate pitch period is next
Determine target pitch period;Alternatively, using preset screening rule, to the normalization amplitude difference energy under each candidate pitch period
Amount and auto-correlated error are screened, and determine target pitch period;Alternatively, according to certain rule, to each candidate pitch week
Normalization amplitude difference energy and auto-correlated error under phase are quantified (quantization is equivalent to scoring) respectively, by each candidate pitch
The quantized value of normalization amplitude difference energy under period is added with the quantized value of auto-correlated error, determines the highest candidate of quantized value
Pitch period is target pitch period.
The determination method of pitch period provided by the embodiments of the present application obtains under preset each candidate pitch period,
Audio signal to be measured and it is each offset audio signal between normalization amplitude difference energy and auto-correlation function and, to each candidate
Auto-correlation function and execution extreme value conversion operation under pitch period, obtain the auto-correlated error under each candidate pitch period,
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target pitch period is determined, due to pole
Value conversion is for being converted to minimum for the maximum of auto-correlation function sum, so that auto-correlated error is normalization amplitude difference energy
Minimum is presented as at pitch period, avoid pitch period normalization magnitude difference energy function and auto-correlation function it
Between peak feature difference bring error, accordingly, it is determined that pitch period accuracy it is relatively high.
On the basis of embodiment shown in Fig. 1, introduces calculate normalization amplitude difference energy and auto-correlation function separately below
The specific method of sum.
In one embodiment, on the basis of embodiment shown in Fig. 1, step " is obtained in preset each candidate pitch
Under period, audio signal to be measured and each normalization amplitude difference energy deviated between audio signal ", including:According to acoustic
Frequency signal and each offset audio signal determine the normalization amplitude difference between audio signal to be measured and each offset audio signal
Energy.
Further, according to audio signal to be measured and each inclined frequency-shift signaling, audio signal to be measured and each offset are determined
Normalization amplitude difference energy between audio signal, including:According to formula
Calculate the normalization amplitude difference energy D between audio signal to be measured and each offset audio signaln(k), wherein k is candidate base
Sound period, s (n) are audio signal to be measured, and s (n+k) is offset audio signal, and N is window function length.
In the present embodiment, a normalization magnitude difference energy function can be pre-establishedCalculate the normalizing between audio signal to be measured and each offset audio signal
Change amplitude difference energy Dn(k), when k is close or equal to actual pitch period, magnitude difference energy function is normalizedCalculated result level off to 0, that is, making Dn(k) the smallest k it is closest or
Equal to actual pitch period.
It should be noted thatFor one kind provided by the embodiments of the present application
Possible implementation, for including formulaFunction and to formulaModification or the obtained function of Parameters variation, be all contained in the scheme of the application.
In some scenes, calculate under candidate pitch period, between audio signal to be measured and offset audio signal from phase
Close function and including:Auto-correlation function at candidate pitch period between audio signal to be measured and offset audio signal and, and
Auto-correlation function at the frequency multiplication of candidate pitch period between audio signal to be measured and offset audio signal and.Reality shown in Fig. 1
On the basis of applying example, " obtain under preset each candidate pitch period, audio signal to be measured and each as shown in Fig. 2, step
Deviate audio signal between auto-correlation function and ", including:
Step 201, according to audio signal to be measured and each first offset audio signal, determine audio signal to be measured and each
First offset audio signal between the first auto-correlation function and;First offset audio signal is audio signal to be measured according to candidate
The signal obtained after pitch period offset.
In the present embodiment, for each candidate pitch period k, treated according to candidate pitch period k survey audio signal into
Line displacement is obtained the corresponding first offset audio signal of each candidate pitch period k and is calculated using preset auto-correlation algorithm
Under each candidate pitch period k, survey audio signal and it is each first offset audio signal between the first auto-correlation function and.
Step 202, according to audio signal to be measured and each second offset audio signal, determine audio signal to be measured and each
Second offset audio signal between the second auto-correlation function and;Second offset audio signal is audio signal to be measured according to candidate
The signal obtained after the frequency multiplication offset of pitch period.
In the present embodiment, each candidate pitch period k can correspond to multiple frequencys multiplication, for example, times of candidate pitch period
Frequency can be expressed as t*k, and t is greater than 1, using preset auto-correlation algorithm, calculates under each candidate pitch period k, audio to be measured
Signal and it is each second offset audio signal between the second auto-correlation function and.
Step 203, according to each first auto-correlation function and each second auto-correlation function and, determine in each candidate pitch
Under period, the corresponding auto-correlation function of audio signal to be measured and.
Further, according to each first auto-correlation function and each second auto-correlation function and, determine in each candidate pitch
Under period, the corresponding auto-correlation function of audio signal to be measured and, including:According to formula
It calculates under each candidate pitch period, the corresponding auto-correlation function of audio signal to be measured and Rn(k), wherein k is candidate pitch
Period, s (n) are audio signal to be measured, and t is the frequency of the candidate pitch period, and when t=1, s (n+t*k) is described first
Deviate audio signal, t>When 1, s (n+t*k) is the second offset audio signal, and w (n) and w (n+t*k) are signal weighting window,
T is maximum candidate pitch period, and M is that weighting windows are long.
In the present embodiment, an auto-correlation function can be pre-establishedMeter
Calculate the auto-correlation function and R between audio signal to be measured and each offset audio signaln(k).When audio signal to be measured is the period
When signal, when k levels off to or is equal to actual pitch period, Rn(k) it is presented as maximum.Wherein, w (n) and M can be basis
What the characteristic of audio signal to be measured determined, it is not limited in the present embodiment.
It should be noted that formulaIt is the present embodiment
One optional scheme, for includingFunction, and to formulaModification or the obtained function of Parameters variation, be all contained in the application's
In scheme.
Optionally, on the basis of embodiment shown in Fig. 1, step is " to the auto-correlation function under each candidate pitch period
With execution extreme value conversion operation, the auto-correlated error under each candidate pitch period is obtained ", including:According to audio signal to be measured,
Each auto-correlation function and, obtain the auto-correlated error under each candidate pitch period.
In the present embodiment, the auto-correlation function between audio signal to be measured and each offset audio signal can first be calculated
With, further according between audio signal to be measured, audio signal to be measured and each offset audio signal auto-correlation function and, calculate every
Auto-correlated error under a candidate pitch period.
It is alternatively possible to according to formulaCalculate each institute
State the auto-correlated error E under candidate pitch periodn(k), wherein k is the candidate pitch period, and s (n) is the audio to be measured
Signal, Rn(k) for the candidate pitch period k auto-correlation function and, w (n) be signal weighting window, N is window function length, α
For the noise correction factor.
In the present embodiment, the auto-correlated error function that minimum is presented as at a pitch period can be constructed in advanceEach candidate pitch week is calculated using auto-correlated error function
Phase corresponding auto-correlated error, wherein w (n) and N can determine that α is for promoting noise according to the characteristic of audio signal to be measured
Accuracy in detection under environment.
It should be noted thatIt is provided in this embodiment
Optinal plan, for including formulaFunction and to formulaModification or the obtained function of Parameters variation, be all contained in the side of the application
In case.
In the embodiment of the present application, using auto-correlated error functionMeter
The corresponding auto-correlated error of each candidate pitch period is calculated, the maximum of each auto-correlation function sum is converted into minimum, is kept away
Exempt from peak feature difference bring error of the pitch period between normalization magnitude difference energy function and auto-correlation function, because
This, determining pitch period accuracy is relatively high.
Fig. 3 is a kind of flow chart of implementation of the step 103 in embodiment illustrated in fig. 1, as shown in figure 3, step 103
" according to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, determining target pitch period " can wrap
It includes:
Step 301, according to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, determine fundamental tone
Estimation set;It include the pitch evaluation value under each candidate pitch period in pitch evaluation set.
It in the present embodiment, can be according to the normalization amplitude difference energy and auto-correlation mistake under each candidate pitch period
Difference calculates the pitch evaluation value under each candidate pitch period, and the pitch evaluation value under all candidate pitch periods is made structure
At pitch evaluation set.For example, the normalization amplitude difference energy under candidate pitch period is added with auto-correlated error, it is somebody's turn to do
Pitch evaluation value under candidate pitch period, alternatively, to the normalization amplitude difference energy under each candidate pitch period and from phase
It closes error and is weighted summation, obtain the pitch evaluation value under each candidate pitch period.
Further, as shown in figure 4, a kind of implementation of step 301 may comprise steps of:
Step 401, according to formula Pn(k)=β * Dn(k)+γ*En(k), the fundamental tone calculated under each candidate pitch period is estimated
Evaluation Pn(k)。
Wherein, k is candidate pitch period, DnIt (k) is the normalization amplitude difference energy under candidate pitch period k, En(k) it is
Auto-correlated error under candidate pitch period k, β and γ are weighted factor ,+γ=1.0 β.
It in the present embodiment, can be the normalization amplitude difference energy and auto-correlated error point under each candidate pitch period
Certain weight is not assigned, that is, weighted factor β and γ is determined, by the normalization amplitude difference energy β * D after weightingn(k) and after weighting
Auto-correlated error γ * En(k) it is added, obtains pitch evaluation value Pn(k)。
Step 402, according to the pitch evaluation value under each candidate pitch period, determine pitch evaluation set.
In the present embodiment, after calculating the pitch evaluation value under each candidate pitch period, by all candidate pitch
Pitch evaluation value under period constitutes pitch evaluation set.Pitch evaluation value in pitch evaluation set can be according to search model
The sequence of interior candidate pitch period is enclosed to sort, and can also be pitch evaluation value according to sequence from big to small or from small to large
Arrangement.
In the present embodiment, according to formula Pn(k)=β * Dn(k)+γ*En(k), the base under each candidate pitch period is calculated
Sound estimated value Pn(k), according to the pitch evaluation value under each candidate pitch period, determine pitch evaluation set, using weighting because
Sub- β and γ is weighted amendment to normalization amplitude difference energy and auto-correlated error respectively, so that the base in pitch evaluation set
Sound estimated value is more accurate, reliable.
Step 302 determines that the corresponding candidate pitch period of minimum fundamental tone estimated value in pitch evaluation set is target base
The sound period.
In the present embodiment, due to pitch period normalization magnitude difference energy function and auto-correlated error function on equal body
It is now minimum, then pitch evaluation value is smaller, more levels off to actual pitch period, therefore, can will be in pitch evaluation set
The corresponding candidate pitch period of minimum fundamental tone estimated value be target pitch period.For example, can will be in pitch evaluation set
Pitch evaluation value is arranged according to sequence from small to large, is by the corresponding candidate pitch period of pitch evaluation value to make number one
Target pitch period;Alternatively, minimum fundamental tone estimated value can also be determined from pitch evaluation set using Bubble Sort Algorithm, from
And determining the corresponding candidate pitch period of minimum fundamental tone estimated value is target pitch period.
Pitch period provided by the embodiments of the present application determines method, according to the normalization amplitude under each candidate pitch period
Poor energy and auto-correlated error determine pitch evaluation set, determine that the minimum fundamental tone estimated value in pitch evaluation set is corresponding
Candidate pitch period is target pitch period, since pitch period is in normalization magnitude difference energy function and auto-correlated error function
On be presented as minimum, then pitch evaluation value is smaller, and more leveling off to actual pitch period therefore can be by pitch evaluation
The corresponding candidate pitch period of minimum fundamental tone estimated value in set is target pitch period, and pitch period estimation can be improved
Accuracy.
It should be understood that although each step in the flow chart of Fig. 1-4 is successively shown according to the instruction of arrow,
These steps are not that the inevitable sequence according to arrow instruction successively executes.Unless expressly stating otherwise herein, these steps
Execution there is no stringent sequences to limit, these steps can execute in other order.Moreover, at least one in Fig. 1-4
Part steps may include that perhaps these sub-steps of multiple stages or stage are not necessarily in synchronization to multiple sub-steps
Completion is executed, but can be executed at different times, the execution sequence in these sub-steps or stage is also not necessarily successively
It carries out, but can be at least part of the sub-step or stage of other steps or other steps in turn or alternately
It executes.
Fig. 5 is a kind of block diagram of the determining device for pitch period that one embodiment provides, as shown in figure 5, the device packet
It includes and obtains module 11, conversion module 12 and determining module 13.
Module 11 is obtained for acquisition under preset each candidate pitch period, audio signal to be measured and each offset sound
Normalization amplitude difference energy and auto-correlation function between frequency signal and;Wherein, the offset audio signal is described to acoustic
The signal that frequency signal obtains after deviating according to the candidate pitch period;
Conversion module 12 is used for the auto-correlation function under each candidate pitch period and executes extreme value conversion operation,
Obtain the auto-correlated error under each candidate pitch period;The extreme value conversion operation be used for the auto-correlation function and
Maximum be converted to minimum;
Determining module 13 is used for according to the normalization amplitude difference energy and auto-correlation mistake under each candidate pitch period
Difference determines target pitch period.
In one of the embodiments, conversion module 12 be specifically used for according to the audio signal to be measured, it is each it is described from
Correlation function and, obtain the auto-correlated error under each candidate pitch period.
Conversion module 12 is specifically used for according to formula in one of the embodiments,
Calculate the auto-correlated error E under each candidate pitch periodn(k), wherein k is the candidate pitch period, and s (n) is institute
State audio signal to be measured, Rn(k) for the candidate pitch period k auto-correlation function and, w (n) be signal weighting window, N is window
Function length, α are the noise correction factor.
The determining module 13 is specifically used for according under each candidate pitch period in one of the embodiments,
Amplitude difference energy and auto-correlated error are normalized, determines pitch evaluation set;It include each described in the pitch evaluation set
Pitch evaluation value under candidate pitch period;Determine the corresponding candidate base of minimum fundamental tone estimated value in the pitch evaluation set
The sound period is the target pitch period.
The determining module 13 is specifically used for according under each candidate pitch period in one of the embodiments,
Amplitude difference energy and auto-correlated error are normalized, determines pitch evaluation set, including:The determining module 13 is according to formula Pn
(k)=β * Dn(k)+γ*En(k), the pitch evaluation value P under each candidate pitch period is calculatedn(k), wherein k is described
Candidate pitch period, DnIt (k) is the normalization amplitude difference energy under the candidate pitch period k, EnIt (k) is the candidate pitch
Auto-correlated error under period k, β and γ are weighted factor ,+γ=1.0 β;According to the base under each candidate pitch period
Sound estimated value determines the pitch evaluation set.
The acquisition module 11 obtains under preset each candidate pitch period in one of the embodiments, to be measured
Normalization amplitude difference energy between audio signal and each offset audio signal, including:The acquisition module 11 is according to described
Audio signal to be measured and each offset audio signal, determine the audio signal to be measured and each offset audio signal
Between normalization amplitude difference energy.
The acquisition module 11 is according to the audio signal to be measured and each offset sound in one of the embodiments,
Frequency signal determines the normalization amplitude difference energy between the audio signal to be measured and each offset audio signal, including:
The acquisition module 11 is according to formulaCalculate the audio signal to be measured with
Normalization amplitude difference energy D between each offset audio signaln(k), wherein k is the candidate pitch period, s (n)
For the audio signal to be measured, s (n+k) is the offset audio signal, and N is window function length.
The acquisition module 11 obtains under preset each candidate pitch period in one of the embodiments, to be measured
Audio signal and it is each offset audio signal between auto-correlation function and, including:The acquisition module 11 is according to described to be measured
Audio signal and each first offset audio signal determine the audio signal to be measured and each first offset audio signal
Between the first auto-correlation function and;The first offset audio signal is the audio signal to be measured according to the candidate pitch
The signal obtained after period migration;According to the audio signal to be measured and each second offset audio signal, determine described to be measured
Audio signal and it is each it is described second offset audio signal between the second auto-correlation function and;The second offset audio signal
It is the audio signal to be measured according to the signal obtained after the frequency multiplication offset of the candidate pitch period;Certainly according to each described first
Correlation function and each second auto-correlation function and, determine under each candidate pitch period, the audio letter to be measured
Number corresponding auto-correlation function and.
The acquisition module 11 is according to each first auto-correlation function and each described the in one of the embodiments,
Two auto-correlation functions and, determine under each candidate pitch period, the corresponding auto-correlation function of the audio signal to be measured
With, including:The acquisition module 11 is according to formulaIt calculates
Under each candidate pitch period, the corresponding auto-correlation function of the audio signal to be measured and Rn(k), wherein k is described
Candidate pitch period, s (n) they are the audio signal to be measured, and t is the frequency of the candidate pitch period, when t=1, s (n+t*
It k) is the first offset audio signal, t>When 1, s (n+t*k) is the second offset audio signal, w (n) and w (n+t*k)
For signal weighting window, T is maximum candidate pitch period, and M is that weighting windows are long.
In Installation practice, other than the embodiment of device exclusive rights, also need to write out all method Xiang Congquan one-to-one correspondence
Device item embodiment.
The specific of determining device about pitch period limits the determination method that may refer to above for pitch period
Restriction, details are not described herein.Modules in the determining device of above-mentioned pitch period can be fully or partially through software, hard
Part and combinations thereof is realized.Above-mentioned each module can be embedded in the form of hardware or independently of in the processor in computer equipment,
It can also be stored in a software form in the memory in computer equipment, execute the above modules in order to which processor calls
Corresponding operation.
In one embodiment, a kind of computer equipment is provided, which can be server, internal junction
Composition can be as shown in Figure 6.The computer equipment include by system bus connect processor, memory, network interface and
Database.Wherein, the processor of the computer equipment is for providing calculating and control ability.The memory packet of the computer equipment
Include non-volatile memory medium, built-in storage.The non-volatile memory medium is stored with operating system, computer program and data
Library.The built-in storage provides environment for the operation of operating system and computer program in non-volatile memory medium.The calculating
The database of machine equipment is for storing the data such as audio signal to be measured, candidate pitch period, various functions predetermined.The meter
The network interface for calculating machine equipment is used to communicate with external terminal by network connection.When the computer program is executed by processor
To realize a kind of determination method of pitch period.
It will be understood by those skilled in the art that structure shown in Fig. 6, only part relevant to application scheme is tied
The block diagram of structure does not constitute the restriction for the computer equipment being applied thereon to application scheme, specific computer equipment
It may include perhaps combining certain components or with different component layouts than more or fewer components as shown in the figure.
In one embodiment, a kind of computer equipment, including memory and processor are provided, is stored in memory
Computer program, the processor realize following steps when executing computer program:
It obtains under preset each candidate pitch period, returning between audio signal to be measured and each offset audio signal
One change amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to described
The signal obtained after candidate pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each time
Select the auto-correlated error under pitch period;The extreme value conversion operation is for being converted to the maximum of the auto-correlation function sum
Minimum;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone is determined
Period.
In one embodiment, following steps are also realized when processor executes computer program:
According to the audio signal to be measured, each auto-correlation function and, obtain under each candidate pitch period
Auto-correlated error.
In one embodiment, following steps are also realized when processor executes computer program:
According to formulaCalculate each candidate pitch
Auto-correlated error E under periodn(k), wherein k is the candidate pitch period, and s (n) is the audio signal to be measured, Rn(k)
For the candidate pitch period k auto-correlation function and, w (n) be signal weighting window, N is window function length, and α is noise correction
The factor.
In one embodiment, following steps are also realized when processor executes computer program:
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, pitch evaluation is determined
Set;It include the pitch evaluation value under each candidate pitch period in the pitch evaluation set;Determine that the fundamental tone is estimated
The corresponding candidate pitch period of minimum fundamental tone estimated value in meter set is the target pitch period.
In one embodiment, following steps are also realized when processor executes computer program:
According to formula Pn(k)=β * Dn(k)+γ*En(k), the pitch evaluation value under each candidate pitch period is calculated
Pn(k), wherein k is the candidate pitch period, DnIt (k) is the normalization amplitude difference energy under the candidate pitch period k, En
It (k) is the auto-correlated error under the candidate pitch period k, β and γ are weighted factor ,+γ=1.0 β;According to each time
The pitch evaluation value under pitch period is selected, determines the pitch evaluation set.
In one embodiment, following steps are also realized when processor executes computer program:
According to the audio signal to be measured and each offset audio signal, the audio signal to be measured and each is determined
Normalization amplitude difference energy between the offset audio signal.
In one embodiment, following steps are also realized when processor executes computer program:
According to formulaCalculate the audio signal to be measured and each institute
State the normalization amplitude difference energy D between offset audio signaln(k), wherein k is the candidate pitch period, and s (n) is described
Audio signal to be measured, s (n+k) are the offset audio signal, and N is window function length.
In one embodiment, following steps are also realized when processor executes computer program:According to the audio to be measured
Signal and each first offset audio signal determine between the audio signal to be measured and each first offset audio signal
The first auto-correlation function and;The first offset audio signal is the audio signal to be measured according to the candidate pitch period
The signal obtained after offset;According to the audio signal to be measured and each second offset audio signal, the audio to be measured is determined
Signal and it is each it is described second offset audio signal between the second auto-correlation function and;The second offset audio signal is institute
Audio signal to be measured is stated according to the signal obtained after the frequency multiplication offset of the candidate pitch period;According to each first auto-correlation
Function and each second auto-correlation function and, determine the audio signal pair to be measured under each candidate pitch period
The auto-correlation function answered and.
In one embodiment, following steps are also realized when processor executes computer program:
According to formulaIt calculates in each candidate
Under pitch period, the corresponding auto-correlation function of the audio signal to be measured and Rn(k), wherein k is the candidate pitch period, s
It (n) is the audio signal to be measured, t is the frequency of the candidate pitch period, and when t=1, s (n+t*k) is described first partially
Move audio signal, t>When 1, s (n+t*k) is the second offset audio signal, and w (n) and w (n+t*k) are signal weighting window, T
For maximum candidate pitch period, M is that weighting windows are long.
In one embodiment, a kind of computer readable storage medium is provided, computer program is stored thereon with, is calculated
Machine program realizes following steps when being executed by processor:
It obtains under preset each candidate pitch period, returning between audio signal to be measured and each offset audio signal
One change amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to described
The signal obtained after candidate pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each time
Select the auto-correlated error under pitch period;The extreme value conversion operation is for being converted to the maximum of the auto-correlation function sum
Minimum;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone is determined
Period.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to the audio signal to be measured, each auto-correlation function and, obtain under each candidate pitch period
Auto-correlated error.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to formulaCalculate each candidate pitch
Auto-correlated error E under periodn(k), wherein k is the candidate pitch period, and s (n) is the audio signal to be measured, Rn(k)
For the candidate pitch period k auto-correlation function and, w (n) be signal weighting window, N is window function length, and α is noise correction
The factor.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, pitch evaluation is determined
Set;It include the pitch evaluation value under each candidate pitch period in the pitch evaluation set;Determine that the fundamental tone is estimated
The corresponding candidate pitch period of minimum fundamental tone estimated value in meter set is the target pitch period.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to formula Pn(k)=β * Dn(k)+γ*En(k), the pitch evaluation value under each candidate pitch period is calculated
Pn(k), wherein k is the candidate pitch period, DnIt (k) is the normalization amplitude difference energy under the candidate pitch period k, En
It (k) is the auto-correlated error under the candidate pitch period k, β and γ are weighted factor ,+γ=1.0 β;According to each time
The pitch evaluation value under pitch period is selected, determines the pitch evaluation set.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to the audio signal to be measured and each offset audio signal, the audio signal to be measured and each is determined
Normalization amplitude difference energy between the offset audio signal.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to formulaCalculate the audio signal to be measured and each institute
State the normalization amplitude difference energy D between offset audio signaln(k), wherein k is the candidate pitch period, and s (n) is described
Audio signal to be measured, s (n+k) are the offset audio signal, and N is window function length.
In one embodiment, following steps are also realized when computer program is executed by processor:According to described to acoustic
Frequency signal and it is each first offset audio signal, determine the audio signal to be measured and it is each it is described first offset audio signal it
Between the first auto-correlation function and;The first offset audio signal is the audio signal to be measured according to the candidate pitch week
The signal obtained after phase offset;According to the audio signal to be measured and each second offset audio signal, determine described to acoustic
Frequency signal and it is each it is described second offset audio signal between the second auto-correlation function and;Described second, which deviates audio signal, is
The audio signal to be measured is according to the signal obtained after the frequency multiplication offset of the candidate pitch period;According to each described first from phase
Close function and each second auto-correlation function and, determine the audio signal to be measured under each candidate pitch period
Corresponding auto-correlation function and.
In one embodiment, following steps are also realized when computer program is executed by processor:
According to formulaIt calculates in each candidate
Under pitch period, the corresponding auto-correlation function of the audio signal to be measured and Rn(k), wherein k is the candidate pitch period, s
It (n) is the audio signal to be measured, t is the frequency of the candidate pitch period, and when t=1, s (n+t*k) is described first partially
Move audio signal, t>When 1, s (n+t*k) is the second offset audio signal, and w (n) and w (n+t*k) are signal weighting window, T
For maximum candidate pitch period, M is that weighting windows are long.
Those of ordinary skill in the art will appreciate that realizing all or part of the process in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the computer program can be stored in a non-volatile computer
In read/write memory medium, the computer program is when being executed, it may include such as the process of the embodiment of above-mentioned each method.Wherein,
To any reference of memory, storage, database or other media used in each embodiment provided herein,
Including non-volatile and/or volatile memory.Nonvolatile memory may include read-only memory (ROM), programming ROM
(PROM), electrically programmable ROM (EPROM), electrically erasable ROM (EEPROM) or flash memory.Volatile memory may include
Random access memory (RAM) or external cache.By way of illustration and not limitation, RAM is available in many forms,
Such as static state RAM (SRAM), dynamic ram (DRAM), synchronous dram (SDRAM), double data rate sdram (DDRSDRAM), enhancing
Type SDRAM (ESDRAM), synchronization link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM
(RDRAM), direct memory bus dynamic ram (DRDRAM) and memory bus dynamic ram (RDRAM) etc..
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The embodiments described above only express several embodiments of the present invention, and the description thereof is more specific and detailed, but simultaneously
It cannot therefore be construed as limiting the scope of the patent.It should be pointed out that coming for those of ordinary skill in the art
It says, without departing from the inventive concept of the premise, various modifications and improvements can be made, these belong to protection of the invention
Range.Therefore, the scope of protection of the patent of the invention shall be subject to the appended claims.
Claims (12)
1. a kind of determination method of pitch period, which is characterized in that the method includes:
It obtains under preset each candidate pitch period, the normalization between audio signal to be measured and each offset audio signal
Amplitude difference energy and auto-correlation function and;Wherein, the offset audio signal is the audio signal to be measured according to the candidate
The signal obtained after pitch period offset;
To the auto-correlation function under each candidate pitch period and extreme value conversion operation is executed, obtains each candidate base
Auto-correlated error under the sound period;The extreme value conversion operation is minimum for being converted to the maximum of the auto-correlation function sum
Value;
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, target fundamental tone week is determined
Phase.
2. the method according to claim 1, wherein the auto-correlation under each candidate pitch period
Function and execution extreme value conversion operation, obtain the auto-correlated error under each candidate pitch period, including:
According to the audio signal to be measured, each auto-correlation function and, obtain under each candidate pitch period from
Correlated error.
3. according to the method described in claim 2, it is characterized in that, it is described according to the audio signal to be measured, it is each it is described from
Correlation function and, obtain the auto-correlated error under each candidate pitch period, including:
According to formulaCalculate each candidate pitch period
Under auto-correlated error En(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, Rn(k) for the candidate pitch period k's
Auto-correlation function and, w (n) is signal weighting window, and N is window function length, and α is the noise correction factor.
4. method according to claim 1-3, which is characterized in that described according to each candidate pitch period
Under normalization amplitude difference energy and auto-correlated error, determine target pitch period, including:
According to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, pitch evaluation collection is determined
It closes;It include the pitch evaluation value under each candidate pitch period in the pitch evaluation set;
Determine that the corresponding candidate pitch period of minimum fundamental tone estimated value in the pitch evaluation set is the target fundamental tone week
Phase.
5. according to the method described in claim 4, it is characterized in that, the normalizing according under each candidate pitch period
Change amplitude difference energy and auto-correlated error, determines pitch evaluation set, including:
According to formula Pn(k)=β * Dn(k)+γ*En(k), the pitch evaluation value P under each candidate pitch period is calculatedn
(k),
Wherein, k is the candidate pitch period, DnIt (k) is the normalization amplitude difference energy under the candidate pitch period k, En
It (k) is the auto-correlated error under the candidate pitch period k, β and γ are weighted factor ,+γ=1.0 β;
According to the pitch evaluation value under each candidate pitch period, the pitch evaluation set is determined.
6. method according to claim 1-3, which is characterized in that the acquisition is in preset each candidate pitch
Under period, audio signal to be measured and each normalization amplitude difference energy deviated between audio signal, including:
According to the audio signal to be measured and each offset audio signal, the audio signal to be measured and each described is determined
Deviate the normalization amplitude difference energy between audio signal.
7. according to the method described in claim 6, it is characterized in that, described according to the audio signal to be measured and each described inclined
Frequency-shift signaling determines the normalization amplitude difference energy between the audio signal to be measured and each offset audio signal, packet
It includes:
According to formulaCalculate the audio signal to be measured and each offset
Normalization amplitude difference energy D between audio signaln(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, and s (n+k) is the offset audio signal,
N is window function length.
8. method according to claim 1-3, which is characterized in that the acquisition is in preset each candidate pitch
Under period, audio signal to be measured and it is each offset audio signal between auto-correlation function and, including:
According to the audio signal to be measured and each first offset audio signal, the audio signal to be measured and each described is determined
First offset audio signal between the first auto-correlation function and;The first offset audio signal is the audio signal to be measured
According to the signal obtained after candidate pitch period offset;
According to the audio signal to be measured and each second offset audio signal, the audio signal to be measured and each described is determined
Second offset audio signal between the second auto-correlation function and;The second offset audio signal is the audio signal to be measured
According to the signal obtained after the frequency multiplication offset of the candidate pitch period;
And, determined in each candidate pitch according to each first auto-correlation function and each second auto-correlation function
Under period, the corresponding auto-correlation function of the audio signal to be measured and.
9. according to the method described in claim 8, it is characterized in that, described according to each first auto-correlation function and each institute
State the second auto-correlation function and, determine under each candidate pitch period, the corresponding auto-correlation of the audio signal to be measured
Function and, including:
According to formulaIt calculates in each candidate pitch week
Under phase, the corresponding auto-correlation function of the audio signal to be measured and Rn(k),
Wherein, k is the candidate pitch period, and s (n) is the audio signal to be measured, and t is the frequency multiplication of the candidate pitch period
Number, when t=1, s (n+t*k) is the first offset audio signal, t>When 1, s (n+t*k) is the second offset audio letter
Number, w (n) and w (n+t*k) they are signal weighting window, and T is maximum candidate pitch period, and M is that weighting windows are long.
10. a kind of determining device of pitch period, which is characterized in that including:
Module is obtained, for obtaining under preset each candidate pitch period, audio signal to be measured and each offset audio are believed
Normalization amplitude difference energy and auto-correlation function between number and;Wherein, the offset audio signal is the audio letter to be measured
Number according to obtained signal after candidate pitch period offset;
Conversion module, for obtaining to the auto-correlation function and execution extreme value conversion operation under each candidate pitch period
Auto-correlated error under each candidate pitch period;The extreme value conversion operation is used for the pole of the auto-correlation function sum
Big value is converted to minimum;
Determining module, for according to the normalization amplitude difference energy and auto-correlated error under each candidate pitch period, really
Set the goal pitch period.
11. a kind of computer equipment, including memory and processor, the memory are stored with computer program, feature exists
In the step of processor realizes any one of claims 1 to 9 the method when executing the computer program.
12. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
The step of method described in any one of claims 1 to 9 is realized when being executed by processor.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810608431.4A CN108831504B (en) | 2018-06-13 | 2018-06-13 | Method and device for determining pitch period, computer equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810608431.4A CN108831504B (en) | 2018-06-13 | 2018-06-13 | Method and device for determining pitch period, computer equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108831504A true CN108831504A (en) | 2018-11-16 |
CN108831504B CN108831504B (en) | 2020-12-04 |
Family
ID=64143869
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810608431.4A Active CN108831504B (en) | 2018-06-13 | 2018-06-13 | Method and device for determining pitch period, computer equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108831504B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111261197A (en) * | 2020-01-13 | 2020-06-09 | 中航华东光电(上海)有限公司 | Real-time voice paragraph tracking method under complex noise scene |
CN112201279A (en) * | 2020-09-02 | 2021-01-08 | 北京佳讯飞鸿电气股份有限公司 | Pitch detection method and device |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4864620A (en) * | 1987-12-21 | 1989-09-05 | The Dsp Group, Inc. | Method for performing time-scale modification of speech information or speech signals |
US6188979B1 (en) * | 1998-05-28 | 2001-02-13 | Motorola, Inc. | Method and apparatus for estimating the fundamental frequency of a signal |
CN101149924A (en) * | 2006-09-18 | 2008-03-26 | 华为技术有限公司 | Method and device for implementing open-loop pitch search |
CN102737645A (en) * | 2012-06-15 | 2012-10-17 | 武汉天喻信息产业股份有限公司 | Algorithm for estimating pitch period of voice signal |
CN103474074B (en) * | 2013-09-09 | 2016-05-11 | 深圳广晟信源技术有限公司 | Pitch estimation method and apparatus |
-
2018
- 2018-06-13 CN CN201810608431.4A patent/CN108831504B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4864620A (en) * | 1987-12-21 | 1989-09-05 | The Dsp Group, Inc. | Method for performing time-scale modification of speech information or speech signals |
US6188979B1 (en) * | 1998-05-28 | 2001-02-13 | Motorola, Inc. | Method and apparatus for estimating the fundamental frequency of a signal |
CN101149924A (en) * | 2006-09-18 | 2008-03-26 | 华为技术有限公司 | Method and device for implementing open-loop pitch search |
CN102737645A (en) * | 2012-06-15 | 2012-10-17 | 武汉天喻信息产业股份有限公司 | Algorithm for estimating pitch period of voice signal |
CN103474074B (en) * | 2013-09-09 | 2016-05-11 | 深圳广晟信源技术有限公司 | Pitch estimation method and apparatus |
Non-Patent Citations (4)
Title |
---|
KEBIN WU 等: "iPEEH: Improving pitch estimation by enhancing harmonics", 《EXPERT SYSTEMS WITH APPLICATIONS》 * |
SAUDI ARABIA: "Noise-Robust Pitch Detection using Auto-correlation Function with Enhancements", 《JOURNAL OF KING SAUD UNIVERSITY - COMPUTER AND INFORMATION SCIENCES》 * |
沈瑜 等: "加权短时自相关函数的基音周期估计算法", 《计算机工程与应用》 * |
潘峥嵘 等: "改进的基音周期检测算法研究", 《计算机工程与应用》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111261197A (en) * | 2020-01-13 | 2020-06-09 | 中航华东光电(上海)有限公司 | Real-time voice paragraph tracking method under complex noise scene |
CN111261197B (en) * | 2020-01-13 | 2022-11-25 | 中航华东光电(上海)有限公司 | Real-time speech paragraph tracking method under complex noise scene |
CN112201279A (en) * | 2020-09-02 | 2021-01-08 | 北京佳讯飞鸿电气股份有限公司 | Pitch detection method and device |
CN112201279B (en) * | 2020-09-02 | 2024-03-29 | 北京佳讯飞鸿电气股份有限公司 | Pitch detection method and device |
Also Published As
Publication number | Publication date |
---|---|
CN108831504B (en) | 2020-12-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108763398B (en) | Database configuration parameter processing method and device, computer equipment and storage medium | |
CN110348562B (en) | Neural network quantization strategy determination method, image identification method and device | |
KR102498093B1 (en) | Method and system for user device identification | |
CN110399839B (en) | Face recognition method, device, equipment and storage medium | |
CN108831504A (en) | Determination method, apparatus, computer equipment and the storage medium of pitch period | |
CN109754135B (en) | Credit behavior data processing method, apparatus, storage medium and computer device | |
CN108804670B (en) | Data recommendation method and device, computer equipment and storage medium | |
CN115081613A (en) | Method and device for generating deep learning model, electronic equipment and storage medium | |
CN111159464A (en) | Audio clip detection method and related equipment | |
CN113360300A (en) | Interface calling link generation method, device, equipment and readable storage medium | |
CN107169045A (en) | A kind of query word method for automatically completing and device based on temporal signatures | |
CN107392220B (en) | Data stream clustering method and device | |
CN116402108A (en) | Model training and graph data processing method, device, medium and equipment | |
CN111461328B (en) | Training method of neural network | |
CN107688948A (en) | Claims Resolution data processing method, device, computer equipment and storage medium | |
CN108831509B (en) | Method and device for determining pitch period, computer equipment and storage medium | |
CN113766405A (en) | Method and device for detecting noise of loudspeaker, electronic equipment and storage medium | |
CN112232417A (en) | Classification method and device, storage medium and terminal | |
CN111835561B (en) | Abnormal user group detection method, device and equipment based on user behavior data | |
CN116561735B (en) | Mutual trust authentication method and system based on multiple authentication sources and electronic equipment | |
CN112800813B (en) | Target identification method and device | |
CN110334260B (en) | Data analysis method, device, computer equipment and storage medium | |
CN110457567B (en) | Method and device for correcting errors of query terms | |
US20230092670A1 (en) | Systems and Methods for Verifying a Device Location | |
CN108834045B (en) | Positioning method and device based on positioning model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |