CN108831504B - Method and device for determining pitch period, computer equipment and storage medium - Google Patents
Method and device for determining pitch period, computer equipment and storage medium Download PDFInfo
- Publication number
- CN108831504B CN108831504B CN201810608431.4A CN201810608431A CN108831504B CN 108831504 B CN108831504 B CN 108831504B CN 201810608431 A CN201810608431 A CN 201810608431A CN 108831504 B CN108831504 B CN 108831504B
- Authority
- CN
- China
- Prior art keywords
- audio signal
- pitch period
- candidate
- offset
- autocorrelation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000005236 sound signal Effects 0.000 claims abstract description 282
- 238000005311 autocorrelation function Methods 0.000 claims abstract description 135
- 238000006243 chemical reaction Methods 0.000 claims abstract description 21
- 230000001131 transforming effect Effects 0.000 claims abstract description 11
- 230000006870 function Effects 0.000 claims description 36
- 238000004590 computer program Methods 0.000 claims description 29
- 238000012937 correction Methods 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 7
- 230000009466 transformation Effects 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000005070 sampling Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000013459 approach Methods 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
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 210000001260 vocal cord Anatomy 0.000 description 1
Images
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
The application relates to a pitch period determination method, a pitch period determination device, a computer device and a storage medium. The method comprises the following steps: acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period; performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum; and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period. The method can improve the accuracy of the pitch period.
Description
Technical Field
The present invention relates to the field of communications technologies, and in particular, to a method, an apparatus, a computer device, and a storage medium for determining a pitch period.
Background
The pitch period is the time when the vocal cords are opened and closed once, and the pitch period is widely applied in the fields of speech coding, recognition and the like as a feature of a speech signal.
Currently, a commonly used pitch period detection method is mainly an auto-correlation function (ACF), and the principle of the ACF is to calculate the sum of auto-correlation functions between an original audio signal and each offset audio signal in each candidate pitch period, and determine the maximum auto-correlation function and the corresponding candidate pitch period as the pitch period of the original audio signal.
However, the pitch period determined by the above method is less accurate.
Disclosure of Invention
In view of the above, it is necessary to provide a pitch period determination method, apparatus, computer device and storage medium capable of improving the accuracy of the pitch period.
A method of pitch period determination, the method comprising:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period.
In one embodiment, the obtaining the autocorrelation error for each of the candidate pitch periods by performing an extremum transformation operation on the sum of the autocorrelation functions for each of the candidate pitch periods comprises:
and acquiring the autocorrelation error under each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum.
In one embodiment, the obtaining an autocorrelation error in each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum includes:
according to the formula
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k)，
Wherein k is the symbolSelecting a pitch period, s (n) as the audio signal to be tested, Rn(k) Is the sum of the autocorrelation functions of the candidate pitch periods k, w (N) is a signal weighting window, N is the window function length, and α is the noise correction factor.
In one embodiment, the determining the target pitch period according to the normalized amplitude difference energy and the autocorrelation error at each of the candidate pitch periods includes:
determining a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period; the pitch estimation set comprises a pitch estimation value under each candidate pitch period;
and determining the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as the target pitch period.
In one embodiment, the determining the set of pitch estimates according to the normalized amplitude difference energy and the autocorrelation error for each of the candidate pitch periods includes:
according to formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k)，
Where k is the candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, of said candidate pitch period kn(k) β and γ are weighting factors, β + γ is 1.0;
and determining the pitch estimation set according to the pitch estimation value under each candidate pitch period.
In one embodiment, the obtaining the normalized amplitude difference energy between the audio signal to be detected and each offset audio signal at each preset candidate pitch period includes:
and determining normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
In one embodiment, the determining, according to the audio signal to be tested and each offset audio signal, normalized amplitude difference energy between the audio signal to be tested and each offset audio signal includes:
according to the formula
Calculating normalized amplitude difference energy D between the audio signal to be detected and each offset audio signaln(k)，
Wherein k is the candidate pitch period, s (N) is the audio signal to be detected, s (N + k) is the offset audio signal, and N is the window function length.
In one embodiment, the obtaining of the sum of autocorrelation functions between the audio signal to be tested and each offset audio signal at each preset candidate pitch period includes:
determining a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
determining a second autocorrelation function sum between the audio signal to be detected and each second offset audio signal according to the audio signal to be detected and each second offset audio signal; the second offset audio signal is a signal obtained after the audio signal to be detected is offset according to the frequency multiplication of the candidate fundamental tone period;
and determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
In one embodiment, the determining, according to each of the first autocorrelation function sums and each of the second autocorrelation function sums, an autocorrelation function sum corresponding to the audio signal to be tested in each of the candidate pitch periods includes:
according to the formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k)，
Wherein k is the candidate pitch period, s (n) is the audio signal to be measured, T is the frequency multiplication factor of the candidate pitch period, when T is 1, s (n + T × k) is the first offset audio signal, when T >1, s (n + T × k) is the second offset audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
An apparatus for pitch period determination, comprising:
the acquiring module is used for acquiring normalized amplitude difference energy and autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
a conversion module, configured to perform extremum conversion on the autocorrelation function and the extremum conversion in each candidate pitch period to obtain an autocorrelation error in each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and the determining module is used for determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period.
The pitch period determining method, the pitch period determining device, the computer equipment and the storage medium acquire the normalized amplitude difference energy and the autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period, performing an extremum transformation operation on the autocorrelation function for each candidate pitch period to obtain an autocorrelation error for each candidate pitch period, determining a target pitch period based on the normalized amplitude difference energy and the autocorrelation error for each candidate pitch period, the extreme value conversion is used for converting the maximum value of the autocorrelation function sum into the minimum value, so that autocorrelation errors are that normalized amplitude difference energy is represented as the minimum value at the pitch period, errors caused by different peak values of the pitch period between the normalized amplitude difference energy function and the autocorrelation function are avoided, and the accuracy of the determined pitch period is higher.
Drawings
FIG. 1 is a flow chart of a method for pitch period determination according to an embodiment;
FIG. 2 is a flow diagram of a method for computing a sum of autocorrelation functions, according to one embodiment;
FIG. 3 is a flow chart of one implementation of step 103 in the embodiment shown in FIG. 1;
FIG. 4 is a flow chart of one implementation of step 201 in the embodiment shown in FIG. 2;
FIG. 5 is a block diagram of an apparatus for pitch period determination according to an embodiment;
FIG. 6 is a block diagram of a computer device, according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The pitch period determining method provided by the application can be applied to an acoustic detection environment to detect the pitch period of a voice signal. The execution subject of the method can be a terminal, a server and the like. The terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices, and the server can be implemented by an independent server or a server cluster formed by a plurality of servers.
Fig. 1 is a flowchart of a method for determining a pitch period according to an embodiment. As shown in fig. 1, the method comprises the steps of:
In this embodiment, the audio signal to be detected is a signal obtained by sampling the original audio signal with a preset sampling frequency. According to the statistical characteristics of the voice signal, a search range can be predetermined for each audio signal to be detected, and the search range comprises a plurality of preset candidate pitch periods and the frequency multiplication of each candidate pitch period. For example, the preset sampling frequency is 8KHz, and the pitch frequency of the voice signal is usually 55hz-400hz, so the search range can be determined to be 20-144 at the sampling frequency of 8 KHz.
After the search range is determined, the audio signal to be detected is shifted according to each candidate pitch period in the search range, and a shifted audio signal corresponding to each candidate pitch period is obtained. The normalized amplitude difference energy between the audio signal to be measured and each offset audio signal can be calculated by adopting a predefined normalized amplitude difference energy function, the sum of the autocorrelation functions between the audio signal to be measured and each offset audio signal can be calculated by adopting a predefined autocorrelation function, the normalized amplitude difference energy function and the autocorrelation function have different peak characteristics, the pitch period is embodied as a minimum value on the normalized amplitude difference energy function, and the pitch period is embodied as a maximum value on the autocorrelation function.
It should be noted that the method for searching the range, calculating the normalized amplitude difference energy and the autocorrelation function sum is not limited in this embodiment.
102, performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain an autocorrelation error under each candidate pitch period; the extremum converting operation is for converting a maximum value of the autocorrelation function sum to a minimum value.
In this embodiment, an extremum transformation operation may be performed on the autocorrelation function corresponding to each candidate pitch period to obtain an autocorrelation error at each candidate pitch period, which is equivalent to transforming a maximum value of the pitch period on the autocorrelation function to a minimum value.
For example, an autocorrelation error function may be predefined, and the audio signal to be measured and the autocorrelation function may be input to the autocorrelation error function to obtain an autocorrelation error. The pitch period is represented as a minimum value on the autocorrelation error function. Other methods for converting the maximum value into the minimum value may be used to perform the extreme value conversion operation, and the embodiment is not limited.
And 103, determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period.
In this embodiment, the target pitch period is determined according to the normalized amplitude difference energy and the autocorrelation error corresponding to each candidate pitch period. For example, weights are respectively given to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period, and then the target pitch period is determined according to the weight of the normalized amplitude difference energy and the weight of the autocorrelation error under each candidate pitch period; or, screening the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period by adopting a preset screening rule to determine a target pitch period; alternatively, the normalized amplitude difference energy and the autocorrelation error in each candidate pitch period are quantized (quantization corresponds to a score) according to a certain rule, and the quantized value of the normalized amplitude difference energy and the quantized value of the autocorrelation error in each candidate pitch period are added to determine the candidate pitch period with the highest quantized value as the target pitch period.
The pitch period determining method provided by the embodiment of the application obtains the normalized amplitude difference energy and the autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period, performing an extremum transformation operation on the autocorrelation function for each candidate pitch period to obtain an autocorrelation error for each candidate pitch period, determining a target pitch period based on the normalized amplitude difference energy and the autocorrelation error for each candidate pitch period, the extreme value conversion is used for converting the maximum value of the autocorrelation function sum into the minimum value, so that autocorrelation errors are that normalized amplitude difference energy is represented as the minimum value at the pitch period, errors caused by different peak values of the pitch period between the normalized amplitude difference energy function and the autocorrelation function are avoided, and the accuracy of the determined pitch period is higher.
Based on the embodiment shown in fig. 1, a specific method for calculating the normalized amplitude-difference energy and the sum of the autocorrelation functions is described below.
In one embodiment, based on the embodiment shown in fig. 1, the step of "obtaining normalized amplitude difference energy between the audio signal to be tested and each offset audio signal at each preset candidate pitch period" includes: and determining the normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
Further, determining normalized amplitude difference energy between the audio signal to be tested and each offset audio signal according to the audio signal to be tested and each offset audio signal, including: according to the formula
Calculating the normalized amplitude difference energy D between the audio signal to be measured and each offset audio signaln(k) Wherein k is a candidate pitch period, s (N) is an audio signal to be tested, s (N + k) is an offset audio signal, and N is a window function length.
In this embodiment, a normalized amplitude-difference energy function may be pre-established
Calculating the normalized amplitude difference energy D between the audio signal to be measured and each offset audio signaln(k) Normalizing the magnitude difference energy function when k is close to or equal to the actual pitch period
Is close to 0, i.e., D is maden(k) The smallest k is closest to or equal to the actual pitch period.
It should be noted that, in the following description,
one possible implementation manner provided for the embodiment of the present application is to include a formula
And a function of
The variations of (2) or the functions resulting from the variation of the parameters are all included in the solution of the present application.
In some scenarios, calculating the sum of autocorrelation functions between the audio signal under test and the offset audio signal at the candidate pitch periods comprises: the sum of the autocorrelation functions between the audio signal to be measured and the offset audio signal at the candidate pitch period, and the sum of the autocorrelation functions between the audio signal to be measured and the offset audio signal at the frequency multiplication of the candidate pitch period. Based on the embodiment shown in fig. 1, as shown in fig. 2, the step "obtaining the sum of autocorrelation functions between the audio signal to be tested and each offset audio signal at each preset candidate pitch period" includes:
step 201, determining a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period.
In this embodiment, for each candidate pitch period k, the audio signal to be measured is shifted according to the candidate pitch period k to obtain a first shifted audio signal corresponding to each candidate pitch period k, and a preset autocorrelation algorithm is adopted to calculate a first autocorrelation function sum between the audio signal to be measured and each first shifted audio signal in each candidate pitch period k.
In this embodiment, each candidate pitch period k may correspond to a plurality of frequency multiples, for example, the frequency multiple of the candidate pitch period may be denoted as t × k, t is greater than 1, and a preset autocorrelation algorithm is adopted to calculate a second autocorrelation function sum between the audio signal to be measured and each second offset audio signal in each candidate pitch period k.
And step 203, determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
Further, determining the autocorrelation function sum corresponding to the audio signal to be detected in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum, including: according to the formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k) Where k is a candidate pitch period, s (n) is an audio signal to be measured, t is a frequency multiplication factor of the candidate pitch period, when t is 1, s (n + t × k) is the first offset audio signal, and t is a frequency multiplication factor of the candidate pitch period, where t is a frequency multiplication factor of the first offset audio signal, and where t is a frequency multiplication factor of the first offset audio signal, t is a frequency multiplication factor of the first offset audio>At time 1, s (n + T × k) is the second shifted audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
In this embodiment, an autocorrelation function may be pre-established
Calculating the autocorrelation function and R between the audio signal to be measured and each offset audio signaln(k) In that respect When the audio signal to be measured is a periodic signal, k approaches or equals to the actual pitch period, Rn(k) Embodied as a maximum. W (n) and M may be determined according to the characteristics of the audio signal to be measured, which is not limited in this embodiment.
It is to be noted that the formula
Is an alternative to the present embodiment for the method comprising
And to formulas
The variations of (2) or the functions resulting from the variation of the parameters are all included in the solution of the present application.
Alternatively, on the basis of the embodiment shown in fig. 1, the step "obtaining the autocorrelation error for each candidate pitch period by performing an extremum transformation operation on the autocorrelation function for each candidate pitch period" includes: and acquiring the autocorrelation error under each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum.
In this embodiment, the autocorrelation function sum between the audio signal to be measured and each offset audio signal may be calculated first, and then the autocorrelation error in each candidate pitch period may be calculated according to the autocorrelation function sum between the audio signal to be measured, and each offset audio signal.
Alternatively, it may be according to a formula
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be tested, Rn(k) Is the sum of the autocorrelation functions of the candidate pitch periods k, w (N) is a signal weighting window, N is the window function length, and α is the noise correction factor.
In this embodiment, an autocorrelation error function embodied as a minimum value at one pitch period may be constructed in advance
And calculating the autocorrelation error corresponding to each candidate pitch period by adopting an autocorrelation error function, wherein w (N) and N can be determined according to the characteristics of the audio signal to be detected, and alpha is used for improving the detection accuracy in a noise environment.
It should be noted that, in the following description,
alternatives provided for the present embodiment to includeFormula (II)
And a function of
The variations of (2) or the functions resulting from the variation of the parameters are all included in the solution of the present application.
In the embodiment of the application, an autocorrelation error function is adopted
And calculating the autocorrelation error corresponding to each candidate pitch period, and converting the maximum value of each autocorrelation function sum into the minimum value, so that the error caused by the difference of peak characteristics of the pitch period between the normalized amplitude difference energy function and the autocorrelation function is avoided, and the determined pitch period has higher accuracy.
Fig. 3 is a flow chart of an implementation of step 103 in the embodiment shown in fig. 1, and as shown in fig. 3, the step 103 "determining the target pitch period according to the normalized amplitude difference energy and the autocorrelation error at each candidate pitch period" may include:
In this embodiment, the pitch estimate value in each candidate pitch period may be calculated according to the normalized amplitude difference energy and the autocorrelation error in each candidate pitch period, and the pitch estimate values in all candidate pitch periods may be used as the pitch estimate set. For example, the normalized amplitude difference energy and the autocorrelation error in the candidate pitch period are added to obtain the pitch estimate in the candidate pitch period, or the normalized amplitude difference energy and the autocorrelation error in each candidate pitch period are weighted and summed to obtain the pitch estimate in each candidate pitch period.
Further, as shown in fig. 4, one implementation of step 301 may include the following steps:
Where k is a candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, for the candidate pitch period kn(k) For the autocorrelation error at the candidate pitch period k, β and γ are weighting factors, and β + γ is 1.0.
In this embodiment, a certain weight may be respectively given to the normalized amplitude difference energy and the autocorrelation error in each candidate pitch period, that is, weighting factors β and γ are determined, and the weighted normalized amplitude difference energy β × D is usedn(k) And weighted autocorrelation error gamma En(k) Adding to obtain pitch estimation value Pn(k)。
In this embodiment, after pitch estimation values in each candidate pitch period are calculated, pitch estimation values in all candidate pitch periods are used to form a pitch estimation set. The pitch estimate values in the pitch estimate set may be sorted in the order of the candidate pitch periods within the search range, or the pitch estimate values may be sorted in the order from large to small or from small to large.
In this embodiment, according to the formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k) And determining a pitch estimation set according to the pitch estimation value in each candidate pitch period, and respectively performing weighted correction on the normalized amplitude difference energy and the autocorrelation error by adopting weighting factors beta and gamma so as to enable the pitch estimation value in the pitch estimation set to be more accurate and reliable.
In this embodiment, since the pitch period is represented as a minimum value in both the normalized amplitude difference energy function and the autocorrelation error function, the smaller the pitch estimate value, the closer the pitch estimate value to the actual pitch period, and therefore, the candidate pitch period corresponding to the minimum pitch estimate value in the pitch estimate set may be the target pitch period. For example, the pitch estimation values in the pitch estimation set may be arranged in descending order, and the candidate pitch period corresponding to the pitch estimation value arranged first may be the target pitch period; alternatively, a bubble sorting algorithm may be used to determine the minimum pitch estimate from the pitch estimate set, so as to determine the candidate pitch period corresponding to the minimum pitch estimate as the target pitch period.
According to the pitch period determining method provided by the embodiment of the application, the pitch estimation set is determined according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period, the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set is determined to be the target pitch period, and as the pitch period is represented as the minimum value on the normalized amplitude difference energy function and the autocorrelation error function, the pitch estimation value is smaller and closer to the actual pitch period, so that the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set can be the target pitch period, and the accuracy of pitch period estimation can be improved.
It should be understood that although the various steps in the flow charts of fig. 1-4 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-4 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
Fig. 5 is a block diagram of an apparatus for determining a pitch period according to an embodiment, as shown in fig. 5, the apparatus includes an obtaining module 11, a converting module 12, and a determining module 13.
The obtaining module 11 is configured to obtain normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
the conversion module 12 is configured to perform extremum conversion on the autocorrelation function sum in each of the candidate pitch periods to obtain an autocorrelation error in each of the candidate pitch periods; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
the determining module 13 is configured to determine a target pitch period according to the normalized amplitude difference energy and the autocorrelation error in each of the candidate pitch periods.
In one embodiment, the conversion module 12 is specifically configured to obtain an autocorrelation error in each candidate pitch period according to the audio signal to be tested and each autocorrelation function sum.
In one embodiment, the conversion module 12 is specifically configured to formulate
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be tested, Rn(k) Is the sum of the autocorrelation functions of the candidate pitch periods k, w (N) is a signal weighting window, N is the window function length, and α is the noise correction factor.
In one embodiment, the determining module 13 is specifically configured to determine a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error in each of the candidate pitch periods; the pitch estimation set comprises a pitch estimation value under each candidate pitch period; and determining the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as the target pitch period.
In one embodiment, the determining module 13 is specifically configured to determine the pitch estimate set according to the normalized amplitude difference energy and the autocorrelation error in each of the candidate pitch periods, and includes: said determination module 13 being according to the formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k) Where k is the candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, of said candidate pitch period kn(k) β and γ are weighting factors, β + γ is 1.0; and determining the pitch estimation set according to the pitch estimation value under each candidate pitch period.
In one embodiment, the obtaining module 11 obtains normalized amplitude difference energy between the audio signal to be detected and each offset audio signal in each preset candidate pitch period, and includes: the obtaining module 11 determines normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
In one embodiment, the determining, by the obtaining module 11, normalized amplitude difference energy between the audio signal to be tested and each offset audio signal according to the audio signal to be tested and each offset audio signal includes: the obtaining module 11 is according to a formula
Calculating normalized amplitude difference energy D between the audio signal to be detected and each offset audio signaln(k) Wherein k is the candidate pitch period, s (N) is the audio signal to be tested, s (N + k) is the offset audio signal, and N is the window function length.
In one embodiment, the obtaining module 11 obtains a sum of autocorrelation functions between the audio signal to be tested and each offset audio signal at each preset candidate pitch period, including: the obtaining module 11 determines a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period; determining a second autocorrelation function sum between the audio signal to be detected and each second offset audio signal according to the audio signal to be detected and each second offset audio signal; the second offset audio signal is a signal obtained after the audio signal to be detected is offset according to the frequency multiplication of the candidate fundamental tone period; and determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
In one embodiment, the determining, by the obtaining module 11, a sum of autocorrelation functions corresponding to the audio signal to be tested in each candidate pitch period according to each first sum of autocorrelation functions and each second sum of autocorrelation functions includes: the obtaining module 11 is according to a formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be detected, t is the frequency multiplication factor of the candidate pitch period, when t is 1, s (n + t x k) is the first offset audio signal, t is the first offset audio signal, and>at time 1, s (n + T × k) is the second shifted audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
In addition to embodiments in which the apparatus is dependent, embodiments in which all method items are written from one apparatus item to one apparatus item are also to be written.
For the specific definition of the pitch period determination device, reference may be made to the above definition of the pitch period determination method, and details are not described here. The modules in the above-described pitch period determining means may be implemented in whole or in part by software, hardware, and combinations thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 6. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing data such as audio signals to be tested, candidate pitch periods, various predefined functions and the like. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a method of pitch period determination.
Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and acquiring the autocorrelation error under each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
according to the formula
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be tested, Rn(k) Is the sum of the autocorrelation functions of the candidate pitch periods k, w (N) is a signal weighting window, N is the window function length, and α is the noise correction factor.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
determining a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period; the pitch estimation set comprises a pitch estimation value under each candidate pitch period; and determining the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as the target pitch period.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
according to formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k) Where k is the candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, of said candidate pitch period kn(k) β and γ are weighting factors, β + γ is 1.0; and determining the pitch estimation set according to the pitch estimation value under each candidate pitch period.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and determining normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
according to the formula
Calculating normalized amplitude difference energy D between the audio signal to be detected and each offset audio signaln(k) Wherein k is the candidate pitch period, s (N) is the audio signal to be tested, s (N + k) is the offset audio signal, and N is the window function length.
In one embodiment, the processor, when executing the computer program, further performs the steps of: determining a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period; determining a second autocorrelation function sum between the audio signal to be detected and each second offset audio signal according to the audio signal to be detected and each second offset audio signal; the second offset audio signal is a signal obtained after the audio signal to be detected is offset according to the frequency multiplication of the candidate fundamental tone period; and determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
according to the formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be detected, t is the frequency multiplication factor of the candidate pitch period, when t is 1, s (n + t x k) is the first offset audio signal, t is the first offset audio signal, and>at time 1, s (n + T × k) is the second shifted audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
and determining a target pitch period according to the normalized amplitude difference energy and the autocorrelation error under each candidate pitch period.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and acquiring the autocorrelation error under each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum.
In one embodiment, the computer program when executed by the processor further performs the steps of:
according toFormula (II)
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be tested, Rn(k) Is the sum of the autocorrelation functions of the candidate pitch periods k, w (N) is a signal weighting window, N is the window function length, and α is the noise correction factor.
In one embodiment, the computer program when executed by the processor further performs the steps of:
determining a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period; the pitch estimation set comprises a pitch estimation value under each candidate pitch period; and determining the candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as the target pitch period.
In one embodiment, the computer program when executed by the processor further performs the steps of:
according to formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k) Where k is the candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, of said candidate pitch period kn(k) β and γ are weighting factors, β + γ is 1.0; and determining the pitch estimation set according to the pitch estimation value under each candidate pitch period.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and determining normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
In one embodiment, the computer program when executed by the processor further performs the steps of:
according to the formula
Calculating normalized amplitude difference energy D between the audio signal to be detected and each offset audio signaln(k) Wherein k is the candidate pitch period, s (N) is the audio signal to be tested, s (N + k) is the offset audio signal, and N is the window function length.
In one embodiment, the computer program when executed by the processor further performs the steps of: determining a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period; determining a second autocorrelation function sum between the audio signal to be detected and each second offset audio signal according to the audio signal to be detected and each second offset audio signal; the second offset audio signal is a signal obtained after the audio signal to be detected is offset according to the frequency multiplication of the candidate fundamental tone period; and determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
In one embodiment, the computer program when executed by the processor further performs the steps of:
according to the formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k) Wherein k is the candidate pitch period, s (n) is the audio signal to be detected, t is the frequency multiplication factor of the candidate pitch period, when t is 1, s (n + t x k) is the first offset audio signal, t is the first offset audio signal, and>at time 1, s (n + T × k) is the second shifted audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the embodiments described above may be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the embodiments described above are not described, but should be considered as being within the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (11)
1. A method for pitch period determination, the method comprising:
acquiring normalized amplitude difference energy and an autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
performing extremum conversion operation on the autocorrelation function under each candidate pitch period to obtain autocorrelation error under each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
determining a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error of each candidate pitch period; the pitch estimation set comprises a pitch estimation value under each candidate pitch period;
and determining a candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as a target pitch period.
2. The method according to claim 1, wherein said performing an extremum transformation operation on the sum of the autocorrelation functions for each of the candidate pitch periods to obtain an autocorrelation error for each of the candidate pitch periods comprises:
and acquiring the autocorrelation error under each candidate pitch period according to the audio signal to be detected and each autocorrelation function sum.
3. The method according to claim 2, wherein said obtaining an autocorrelation error for each of said candidate pitch periods based on said audio signal under test, each of said autocorrelation function sums comprises:
according to the formula
Calculating an autocorrelation error E for each of said candidate pitch periodsn(k)，
Wherein k is the candidate pitch period, s (n) is the audio signal to be tested, Rn(k) Sum of autocorrelation functions for said candidate pitch period k, w (n)Is the signal weighting window, N is the window function length, and α is the noise correction factor.
4. The method according to claim 1, wherein said determining a set of pitch estimates from normalized amplitude difference energy and autocorrelation error at each of said candidate pitch periods comprises:
according to formula Pn(k)＝β*Dn(k)+γ*En(k) Calculating the pitch estimation value P of each candidate pitch periodn(k)，
Where k is the candidate pitch period, Dn(k) Is the normalized amplitude difference energy, E, of said candidate pitch period kn(k) β and γ are weighting factors, β + γ is 1.0;
and determining the pitch estimation set according to the pitch estimation value under each candidate pitch period.
5. The method according to any one of claims 1 to 3, wherein the obtaining the normalized amplitude difference energy between the audio signal under test and each offset audio signal at each preset candidate pitch period comprises:
and determining normalized amplitude difference energy between the audio signal to be detected and each offset audio signal according to the audio signal to be detected and each offset audio signal.
6. The method of claim 5, wherein determining the normalized amplitude difference energy between the audio signal under test and each of the offset audio signals according to the audio signal under test and each of the offset audio signals comprises:
according to the formula
Calculating normalized amplitude difference energy D between the audio signal to be detected and each offset audio signaln(k)，
Wherein k is the candidate pitch period, s (N) is the audio signal to be detected, s (N + k) is the offset audio signal, and N is the window function length.
7. The method according to any one of claims 1 to 3, wherein obtaining a sum of autocorrelation functions between the audio signal under test and each of the shifted audio signals at each of the preset candidate pitch periods comprises:
determining a first autocorrelation function sum between the audio signal to be detected and each first offset audio signal according to the audio signal to be detected and each first offset audio signal; the first offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
determining a second autocorrelation function sum between the audio signal to be detected and each second offset audio signal according to the audio signal to be detected and each second offset audio signal; the second offset audio signal is a signal obtained after the audio signal to be detected is offset according to the frequency multiplication of the candidate fundamental tone period;
and determining the autocorrelation function sum corresponding to the audio signal to be tested in each candidate pitch period according to each first autocorrelation function sum and each second autocorrelation function sum.
8. The method according to claim 7, wherein said determining a corresponding sum of autocorrelation functions of said audio signal under test at each of said candidate pitch periods based on each of said first and second sums of autocorrelation functions comprises:
according to the formula
Calculating the autocorrelation function and R corresponding to the audio signal to be tested in each candidate pitch periodn(k)，
Wherein k is the candidate pitch period, s (n) is the audio signal to be measured, T is the frequency multiplication factor of the candidate pitch period, when T is 1, s (n + T × k) is the first offset audio signal, when T >1, s (n + T × k) is the second offset audio signal, w (n) and w (n + T × k) are signal weighting windows, T is the largest candidate pitch period, and M is the weighting window length.
9. An apparatus for pitch period determination, comprising:
the acquiring module is used for acquiring normalized amplitude difference energy and autocorrelation function sum between the audio signal to be detected and each offset audio signal in each preset candidate pitch period; the offset audio signal is a signal obtained after the audio signal to be detected is offset according to the candidate pitch period;
a conversion module, configured to perform extremum conversion on the autocorrelation function and the extremum conversion in each candidate pitch period to obtain an autocorrelation error in each candidate pitch period; the extremum transforming operation is to transform a maximum of the autocorrelation function sum to a minimum;
a determining module, configured to determine a pitch estimation set according to the normalized amplitude difference energy and the autocorrelation error in each of the candidate pitch periods; the pitch estimation set comprises a pitch estimation value under each candidate pitch period; and determining a candidate pitch period corresponding to the minimum pitch estimation value in the pitch estimation set as a target pitch period.
10. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 9 when executing the computer program.
11. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 9.
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 CN108831504A (en) | 2018-11-16 |
| CN108831504B true 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) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| 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 |
| 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 |
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 |
|---|
| iPEEH: Improving pitch estimation by enhancing harmonics;Kebin Wu 等;《Expert Systems with Applications》;20160803;第317-319页 * |
| Noise-Robust Pitch Detection using Auto-correlation Function with Enhancements;Saudi Arabia;《Journal of King Saud University - Computer and Information Sciences》;20090506;第13-28页 * |
| 加权短时自相关函数的基音周期估计算法;沈瑜 等;《计算机工程与应用》;20121011;第1-7页 * |
| 改进的基音周期检测算法研究;潘峥嵘 等;《计算机工程与应用》;20131210;第223-226页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108831504A (en) | 2018-11-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112489677B (en) | Voice endpoint detection method, device, equipment and medium based on neural network | |
| CN113610232B (en) | Network model quantization method and device, computer equipment and storage medium | |
| JP6272433B2 (en) | Method and apparatus for detecting pitch cycle accuracy | |
| CN108831504B (en) | Method and device for determining pitch period, computer equipment and storage medium | |
| Low et al. | Long-term extreme response analysis of offshore structures by combining importance sampling with subset simulation | |
| KR101944429B1 (en) | Method for frequency analysis and apparatus supporting the same | |
| CN110188862A (en) | Searching method, the device, system of model hyper parameter for data processing | |
| US11042880B1 (en) | Authenticating users in the presence of small transaction volumes | |
| CN107393549A (en) | Delay time estimation method and device | |
| JP5994639B2 (en) | Sound section detection device, sound section detection method, and sound section detection program | |
| CN111161880A (en) | Medical information classification method and device based on classification model and computer equipment | |
| CN112634870B (en) | Keyword detection method, device, equipment and storage medium | |
| CN112800813B (en) | Target identification method and device | |
| CN116959495A (en) | Speech signal-to-noise ratio estimation method and system | |
| US20210264342A1 (en) | Method, Computer Apparatus, And Storage Medium For Evaluating Product Reliability | |
| CN115705486A (en) | Method and device for training quantitative model, electronic equipment and readable storage medium | |
| WO2021138898A1 (en) | Speech recognition result detection method and apparatus, and storage medium | |
| CN108831509B (en) | Method and device for determining pitch period, computer equipment and storage medium | |
| CN111368044A (en) | Intelligent question answering method and device, computer equipment and storage medium | |
| CN115862686A (en) | Voice endpoint detection model construction method and device and computer equipment | |
| CN118193298B (en) | Test method and device of memory calculation integrated chip, electronic equipment and storage medium | |
| CN117910456B (en) | Method and system for evaluating software cost | |
| CN113160795B (en) | Language feature extraction model training method, device, equipment and storage medium | |
| RU2745717C2 (en) | Equipment and method for determining fundamental tone information | |
| TW200428355A (en) | Method for calculation a pitch period estimation of speech signals with variable step size |
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 |