JPH05297897A - Voiced sound deciding method - Google Patents

Abstract

PURPOSE:To eliminate an offensive sound in MBE (multiband excited) encoding by generating a voiceless sound in the whole band in a block when the signal level of a block wherein a band decided as a voiced sound block is present is less than a 1st specific threshold value. CONSTITUTION:Data of one block on a time base are converted by an orthogonal conversion part 13 into data of one block on a frequency axis and divided by a band division part 14 into plural bands, and a V/UV decision part 15 decides whether the sound is voiced or not, band by band. When it is judged that at least one band is a voiced sound band, a level detection part 18 detects the signal level of one block including the voiced sound band and supplies it to a comparison part 19. This comparison part 19 compares it with a threshold value supplied to an input terminal 20 and when the level is less than the threshold value, a decision part 21 judges the signal of the one block judged by the decision part 15 as the voiced sound band is a granular noise level, i.e., a mean amplitude less than 1LSB (least order digit).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voiced sound for discriminating a voiced sound from noise or unvoiced sound from data on a frequency axis obtained by dividing an input voice signal into blocks and converting it into a frequency axis. Regarding the determination method.

[0002]

2. Description of the Related Art Speech is classified into voiced sound and unvoiced sound as a property of sound. Voiced sound is a voice accompanied by vocal cord vibration and is observed as periodic vibration. Unvoiced sound is observed as a non-periodic sound with no vocal cord vibration. Most of the normal voices are voiced sounds, and unvoiced sounds are only special consonants called unvoiced consonants. The period of voiced sound is determined by the period of vocal cord vibration, which is called the pitch period, and its reciprocal is called the pitch frequency. The pitch period and the pitch frequency (hereinafter, referred to as a pitch period when referred to as a pitch) are important factors that determine the pitch of the voice and intonation. Therefore, how accurately the pitch is captured affects the sound quality of the voice. However, when capturing the pitch, it is necessary to consider noise around the voice, so-called background noise, and quantization noise during quantization. Distinguishing these noises or unvoiced sounds from voiced sounds is important when coding speech signals.

Specific examples of the above-mentioned encoding of the voice signal include MBE (Multiband Excitation) encoding, SBE (Singleband Excitation) encoding, Harmonic encoding, and SBC ( Sub-band Coding), LP
There are C (Linear Predictive Coding), DCT (Discrete Cosine Transform), MDCT (Modified DCT), FFT (Fast Fourier Transform), and the like.

For example, in the above MBE encoding, when extracting a pitch from an input speech signal waveform, it is easy to capture the trajectory of the pitch even when a clear pitch does not appear. Then, on the decoding side (synthesis side), a cosine wave (cosin) wave synthesis is used to synthesize a voiced sound waveform on the time axis based on the above-mentioned pitch, and it is added and synthesized with a separately synthesized unvoiced sound waveform on the time axis and output. To do.

[0005]

By the way, the above MBE
In encoding, if hum noise, which is a low pitch frequency (50Hz) noise, is added to the background of the input speech signal, the synthesis side performs synthesis so that the peaks of the cosin wave overlap at that pitch. That is, each cosin wave is synthesized by addition of fixed phases as is done in the synthesis of voiced sounds. Then, by having the periodicity that the cosin waves are superposed with a fixed phase, the hum noise, which is a kind of background noise, becomes an impulse waveform with periodicity. In other words, since the intensity of the amplitude that should originally be scattered on the time axis is concentrated in one portion of the frame and has periodicity, it is reproduced as a very offensive noise. ..

The present invention has been made in view of the above circumstances, and is capable of distinguishing a voiced sound from noise or unvoiced sound with certainty, and capable of suppressing the generation of an abnormal sound on the synthesis side. The purpose is to provide a discrimination method.

[0007]

A voiced sound discrimination method according to the present invention comprises a step of dividing an input voice signal into blocks and converting them into a frequency axis to obtain frequency axis data.
Dividing the data on the frequency axis into a plurality of bands,
A step of determining whether or not there is voiced sound for each of the divided bands,
The step of detecting the level of the signal in block units in which the band determined to be the voiced sound exists by the above-mentioned discrimination, and the entire band is made unvoiced when the level of the detected signal becomes equal to or lower than a first predetermined threshold value. The above-mentioned problems are solved by the steps of:

A voiced sound discrimination method according to another invention is
There is a step of dividing the input audio signal into blocks and converting it into a frequency axis to obtain frequency axis data, a step of dividing the frequency axis data into a plurality of bands, and a step of dividing each of the divided bands. A step of determining whether or not it is a voice sound, a step of detecting the level of a signal of a block unit in which the band which is said to be a voiced sound by the above-mentioned determination exists, a step of obtaining a peak value of the signal in the block, A step of obtaining a normalized value of the peak value based on a signal level in a block, and a block unit in which the normalized value is equal to or less than a second predetermined threshold and a band in which the voiced sound is present exists To make the entire band unvoiced when the level of the signal of 1 becomes less than or equal to the first predetermined threshold value.

Here, the level of the signal may be detected from the data on the frequency axis of the block unit in which the band that is said to be voiced exists, and further detected from the data obtained by band-dividing the data on the frequency axis. You may.

Further, the above-mentioned discrimination of voiced sound is to discriminate between voiced sound, noise and unvoiced sound. It is possible to surely discriminate voiced sound and noise or unvoiced sound. That is, noise (including hum noise of a minute level) or unvoiced sound can be discriminated from the input voice signal. In such a case, for example, if the entire band of the input audio signal is forcibly made unvoiced, it is possible to suppress the generation of abnormal noise on the synthesis side.

[0011]

The input voice signal in block units is converted to the frequency axis, the data on the frequency axis is divided into a plurality of bands, and it is determined for each band whether or not it is a voiced sound. Then, when the level of the signal of the block in which the band discriminated as the voiced sound exists becomes equal to or lower than the first predetermined threshold value, the entire band in the block is made unvoiced. Therefore, generation of abnormal noise can be suppressed.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a voiced sound discrimination method according to the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic configuration of a voiced sound discrimination apparatus for explaining a voiced sound discrimination method according to a first embodiment of the present invention. In this embodiment, the input audio signal is divided into blocks, and the peak value of the signal in one block of the data on the frequency axis obtained by converting the data into the frequency axis is normalized based on the signal level in the block. And the one block of the data on the frequency axis is divided into a plurality of bands, voiced sound (V) / unvoiced sound (UV) is discriminated for each of the plurality of bands, and at least one band of the plurality of bands is determined. Is voiced, the entire band in the block is unvoiced according to the signal level of the block including the voiced band. Specifically, when the normalized value is less than or equal to a predetermined threshold value and the signal level is less than or equal to a predetermined threshold value, the entire band is unvoiced.

In FIG. 1, an input terminal 11 has at least a low frequency band for removing a so-called DC (direct current) offset component and band limitation (for example, 200-3400 Hz) by a filter such as an HPF (high-pass filter) not shown. An audio signal from which components (200 Hz or less) have been removed is supplied. This signal is sent to the windowing processing unit 12. In this windowing processing unit 12, for example, a Hamming window is applied to 1 block N samples (for example, N = 256), and this 1 block is processed for 1 frame L samples (for example, L = L).
= 160), it is moved sequentially in the time axis direction,
The overlap between blocks is NL samples (96 samples, for example). This windowing processing unit 12
The signal that is a block of N samples in
3 is supplied. The orthogonal transform unit 13 adds 0 data for 1792 samples to a sample sequence of 256 samples per block (so-called zero padding) 2
With 048 samples, the time-axis data sequence of 2048 samples is subjected to orthogonal transformation processing such as FFT (Fast Fourier Transform) and converted into a frequency-axis data sequence. The data on the frequency axis from the orthogonal transform unit 13 is the band division unit 1
4 and at the same time, it is also supplied to the signal level detection unit 18 described later. The band division unit 14
Divides the supplied data on the frequency axis into a plurality of bands according to the pitch extracted by a pitch extraction unit (not shown). The data on the frequency axis divided into each band is supplied to the V (voiced sound) / UV (unvoiced sound) determination unit 15. This V
/ UV discrimination unit 15 determines V / U for each of the divided bands.
V is being discriminated. This V / UV discrimination information is supplied to the correction unit 16 and the determination unit 21.

On the other hand, the signal level detecting section 18 detects the signal level of each block of the data on the time axis from the windowing processing section 12, and detects from the block containing one band of the voiced sound. The detected signal level is supplied to the comparison unit 19. The comparison unit 19 compares the first predetermined threshold value supplied from the input terminal 20 with the signal level, and supplies the comparison result to the determination unit 21.

Here, the signal level detection unit 18 may detect the level of the signal for each block from the data on the frequency axis from the orthogonal transformation unit 13, and the band from the band division unit 14 may be detected. It may be detected from the divided data on the frequency axis.

The data on the time axis from the windowing processing unit 12 is the peak detection unit 22 and the intra-block variance detection unit 2.
3 is also supplied. The peak detection unit 22 detects a peak value of data for each block on the time axis and supplies the peak value to the normalization unit 24. The intra-block variance detector 23 detects the variance of the signal level for each block on the time axis and supplies the variance to the normalizer 24. The normalization unit 24 calculates a normalized value from the peak value and the variance of the signal level. Then, it is supplied to the comparison unit 25. The comparison unit 25 compares the second predetermined threshold value supplied from the input terminal 26 with the normalized value, and supplies the comparison result to the determination unit 21.

The judgment unit 21 controls the correction unit 16. For example, when the signal level is equal to or lower than the first predetermined threshold value and the value obtained by normalizing the peak value is equal to or lower than the second predetermined threshold value, the correction unit 16 outputs one block of data on the frequency axis. The whole band is controlled to be unvoiced. Then, the unvoiced sound information is output from the output terminal 17.

The operation of this embodiment will be described below. The number of samples N in one block cut out by applying the Hamming window in the windowing processing unit 12 is 256 samples, and the input sample sequence is x (n). This one block (256 samples) of data on the time axis is converted into one block of data on the frequency axis by the orthogonal transform unit 13. The one-block data on the frequency axis is divided into a plurality of bands by the band dividing unit 14. The data on the frequency axis of one block divided into a plurality of bands is V /
The UV discrimination unit 15 discriminates V / UV for each band. Here, when there is no voiced band at all, that is, when the band is unvoiced, the correction section 16 does not make any particular correction.

On the other hand, when the V / UV discriminating section 15 determines that at least one band is voiced sound, the signal level of one block including the voiced sound band is detected by the signal level detecting section 18. To be detected. The detected signal level is supplied to the comparison unit 18 and compared with the threshold value from the input terminal 20. Then, the comparison result is supplied to the determination unit 21.

The case where at least one band is voiced by the V / UV discriminator 15 means that the level of the input voice signal is 1 LSB (least significant digit) of the quantization step. That is, the V / UV discrimination unit 15 discriminates noise of 1 LSB (granular noise), and when the signal level of one block of voice at this time is a minute level, the entire band is made unvoiced.

Specifically, the signal level detector 18 calculates the energy variance σ by

[0022]

[Equation 1]

It is calculated by the equation (1) shown by. Where x is

[0024]

[Equation 2]

It is the average value of x (n) represented by

The above σ is supplied to the comparator 19 when the V / UV discriminating unit 15 determines that at least one band is voiced sound. The comparison unit 19 compares the threshold value g _t (for example, g _t = 8) supplied to the input terminal 20 with the variance σ. Here, if σ is smaller than the threshold value g _t , the determination unit 21 causes the V / UV determination unit 15 to perform at least 1
It is determined that the signal of one block whose band is voiced has a granular noise level, that is, an average amplitude of 1 LSB or less.

Further, the peak detecting section 22 calculates the peak value P _{e of the} data on the frequency axis of one block by

[0028]

[Equation 3] It is detected by the equation (3). Here, x is the average value of x (n) as described above. Then, the normalization unit 24
Is obtained by dividing the peak value P _e by the variance σ shown in the equation (1) to obtain a normalized value P _n . That is, the normalized value P _n is represented by P _n = P _e / σ (4) Then, the comparison unit 25 compares the normalized value P _n with the threshold P _thn supplied from the input terminal 26, and supplies the comparison information to the determination unit 21. This determination unit 21
Also, the comparison information obtained by the comparison unit 19 comparing the variance σ and the threshold value g _t (for example, g _t = 8) supplied from the input terminal 20 is also supplied. Then, the judgment unit 21 is configured to compare the comparison unit 25 and the comparison unit 1 with each other.
A control signal is supplied to the correction unit 16 in accordance with the comparison information from 9. Specifically, the normalized value P _n is less than or _equal to a predetermined threshold P _t , and the signal level σ is also a predetermined threshold g _t.
In the following cases, the correction unit 16 is caused to function to make the entire band of the data on the frequency axis of one block unvoiced.

Therefore, in this embodiment, when the signal level of one block of the signals other than the unvoiced sound in the entire band, that is, the signal in which at least one band is determined to be the voiced sound is small and the peak value is small, In addition, by forcibly making all bands unvoiced, it is possible to prevent the generation of abnormal noise due to fixed phase addition that occurs on the combining side.

The voiced sound discrimination method according to the present invention uses M
It is applicable to vocoders such as BE. That is, when the data on the frequency axis of one block of the input audio signal is made noise or unvoiced sound, all the bands are forced to be unvoiced sound to prevent the generation of abnormal sound on the synthesis side of the vocoder such as MBE. You can

A specific example of an MBE (Multiband Excitation) vocoder, which is a kind of speech signal synthesis analysis coding apparatus (so-called vocoder) to which the voiced sound discrimination method according to the present invention is applicable, will be described below with reference to the drawings. It will be explained with reference to FIG. This MBE vocoder is DW Grif
fin and JS Lim, "Multiband Excitation Vocoder,"
IEEE Trans. Acoustics, Speech, and Signal Processing,
Vol.36, No.8, pp.1223-1235, Aug.1988, the conventional PARCOR (PARtial auto-C
In ORrelation: partial autocorrelation) vocoders, voiced sections and unvoiced sections were switched for each block or frame when modeling speech, whereas in MBE vocoder, the same time (in the same block or frame) Voiced section (Voiced) and unvoiced section (Unvoic)
ed) section is modeled on the assumption that and exist.

FIG. 2 is a block diagram showing the overall schematic configuration of the embodiment of the MBE vocoder. In FIG. 2, an audio signal is supplied to an input terminal 101, and this input audio signal is sent to a filter 102 such as an HPF (high-pass filter) to be a so-called DC signal.
Removal of (DC) offset and band limitation (for example, 200
At least low frequency component (2)
(Less than 00 Hz) is removed. The signal obtained through the filter 102 is sent to the pitch extraction unit 103 and the windowing processing unit 104, respectively. Pitch extraction unit 103
In, the input voice signal data is divided into blocks in units of a predetermined number N (for example, N = 256) (or cut out by a rectangular window), and pitch extraction is performed on the voice signals in this block. Such a cut block (256 samples) is moved in the time axis direction at a frame interval of L samples (for example, L = 160) as shown in A of FIG. 3, and the overlap between the blocks is N−. There are L samples (for example, 96 samples). In addition, in the windowing processing unit 104, 1 block N
A predetermined window function, for example, a Hamming window is applied to the sample, and this windowed block is sequentially moved in the time axis direction at intervals of 1 frame L sample.

When this windowing process is expressed by a mathematical expression, x _w (k, q) = x (q) w (kL-q) (5) In this equation (5), k is a block number,
q represents the time index (sample number) of the data, and the q-th data x (q) of the unprocessed input signal is windowed by the window function (w (kL-q)) of the kth block. It is shown that the data x _w (k, q) can be obtained by doing so. 3A in the pitch extraction unit 103
The window function w _r (r) in the case of the rectangular window is as follows: w _r (r) = 1 0 ≦ r <N (6) = 0 r <0, N ≦ r window function w _h (r) in the case of Hamming window as shown in B of FIG. 3 is a section _{104, w h (r) =} 0.54 - 0.46 cos (2πr / (N-1)) 0 ≦ r <N (7) = 0 r <0, N ≦ r. Such a window function w _r (r) or w (5) when using the _h (r) formula of the window function w (r) (= w (KL-
q)), the zero-zero interval is 0 ≦ kL−q <N, which is transformed into kL−N <q ≦ kL. Therefore, for example, in the case of the above rectangular window, the window function w _r (kL−q) =
As shown in FIG. 4, 1 becomes kL-N <q ≦ kL.
It will be when. Further, the above equations (5) to (7) are expressed by the length N
The window of (= 256) samples is shown to advance by L (= 160) samples. Below, above (6)
Each of the N points (0 ≦ r
The non-zero sample sequence of <N) is represented by x _wr (k, r) and x _wh
Let us denote it as (k, r).

In the windowing processing unit 104, as shown in FIG. 5, 179 is applied to the sample sequence x _wh (k, r) of one block of 256 samples to which the Hamming window of the equation (7) is applied.
Two samples of 0 data are added (so-called zero padding) to make 2048 samples, and the orthogonal transformation unit 105 performs orthogonal transformation such as FFT (Fast Fourier Transform) on the time-axis data sequence of 2048 samples. Processing is performed.

In the pitch extraction unit 103, the above x _wr (k, r)
Pitch extraction is performed based on the sample sequence (1 block N samples). The pitch extraction method is known to include periodicity of time waveform, periodic frequency structure of spectrum, and autocorrelation function.
The center correlation waveform autocorrelation method is used. Regarding the center clip level in the block at this time, one clip level may be set for one block, but the peak level of the signal of each part (each sub-block) obtained by subdividing the block is detected and When there is a large difference in the peak level of each sub-block, the clip level is changed stepwise or continuously within the block. The peak period is determined based on the peak position of the autocorrelation data of the center clip waveform. At this time, a plurality of peaks are obtained from the autocorrelation data belonging to the current frame (the autocorrelation is obtained for the data of N samples of one block), and the maximum peak among the plurality of peaks is equal to or larger than a predetermined threshold. In the case of, the maximum peak position is set as the pitch period, and in other cases, the pitch is within a pitch range that satisfies a predetermined relationship with the pitch other than the current frame, for example, the pitch of the previous frame and the pitch of the previous frame. As a result, a peak in the range of ± 20% is obtained, and the pitch of the current frame is determined based on this peak position. In this pitch extraction unit 103, a relatively rough pitch search is performed by an open loop, and the extracted pitch data has a high precision (fine) pitch search unit 10.
Then, the high precision pitch search (pitch fine search) is performed by the closed loop.

High precision (fine) pitch search unit 106
Includes rough pitch data of integer (integer) values extracted by the pitch extraction unit 103 and the orthogonal transformation unit 10.
5, the data on the frequency axis subjected to FFT, for example, is supplied. In this high precision pitch search unit 106,
Centering on the above coarse pitch data value, it is ± 0.2 in increments of ±
Shake several samples at a time to drive to the optimum fine pitch data value with a decimal point (floating). As a fine search method at this time, a so-called analysis by synthesis method is used, and the pitch is selected so that the synthesized power spectrum is closest to the power spectrum of the original sound.

The fine search of the pitch will be described. First, in the MBE vocoder, the FF
S (j) as spectrum data on the frequency axis orthogonally transformed by T etc. is expressed as S (j) = H (j) | E (j) | 0 <j <J (8) It is assumed that the model. Here, J corresponds to πω _s = f _s / 2, and the sampling frequency f _s =
When 2πω _s is, for example, 8 kHz, it corresponds to 4 kHz.
In the above formula (8), when the spectrum data S (j) on the frequency axis has a waveform as shown in A of FIG. 6, H (j)
Is the original spectrum data S (j) as shown in B of FIG.
Shows the spectral envelope of E (j)
6 shows a spectrum of an excitation signal (excitation) which is periodic at an equal level as shown in C of FIG. That is, the FFT spectrum S (j) is the spectrum envelope H (j) and the power spectrum of the excitation signal | E (j) |
It is modeled as the product of and.

Power spectrum of the excitation signal | E (j)
Is a spectral waveform corresponding to the waveform of one band (band) for each band on the frequency axis in consideration of the periodicity (pitch structure) of the waveform on the frequency axis determined according to the pitch. It is formed by arranging it repeatedly. The waveform for one band is obtained by performing FFT by regarding a waveform obtained by adding (zero-filling) 0 data for 1792 samples to a Hamming window function of 256 samples as shown in FIG. 5 as a time axis signal. It can be formed by cutting out an impulse waveform having a certain bandwidth on the frequency axis according to the pitch.

Next, for each band divided according to the above pitch, a value (a kind of amplitude) | A that represents the above H (j) (minimizes the error for each band) | A _m
Ask for |. Here, for example, when the lower and upper points of the m-th band (band of the m-th harmonic) are a _m and b _m , respectively, the error ε _m of the m-th band is

[0040]

[Equation 4] Can be expressed as | A _m | that minimizes this error ε _m
Is

[0041]

[Equation 5] Therefore, when | A _m | in the equation (10), the error ε _m is minimized. Such an amplitude | A _m | is obtained for each band, and the obtained amplitude | A _m | is used to obtain an error ε _m for each band defined by the above equation (9). Next, the sum Σε of all the bands of such error ε _m for each band
_{Find m} . Further, such an error sum value Σε _m of all bands is obtained for some slightly different pitches, and a pitch that minimizes the error sum value Σε _m is obtained.

That is, with the rough pitch obtained by the pitch extraction unit 103 as the center, several types are prepared up and down in steps of, for example, 0.25. The error sum value Σε _m is obtained for each of these plural kinds of slightly different pitches. In this case, when the pitch is determined, the bandwidth is determined. From the above equation (10), the power spectrum | S (j) |
(j) | and the error ε _m in the above equation (9) can be obtained, and the sum total value Σε _m of all the bands can be obtained. This error sum value Σε _m is obtained for each pitch, and the pitch corresponding to the minimum error sum value is determined as the optimum pitch. As described above, the high-precision pitch search unit 106 obtains the optimum fine (eg, 0.25 step) pitch, and the amplitude | A _m | corresponding to this optimum pitch.
Is determined.

In the above description of the pitch fine search, in order to simplify the description, all bands are voiced (Vo
Assuming the case of iced), as described above, MBE
In the vocoder, unvoiced sound (Un
Since a model in which a voiced) area exists is used, it is necessary to distinguish voiced sound / unvoiced sound for each band.

The optimum pitch and amplitude | A _m | data from the high precision pitch search unit 106 is sent to the voiced sound / unvoiced sound determination unit 107, and the voiced sound / unvoiced sound is discriminated for each band. NSR (noise to signal ratio) is used for this determination. That is, the NSR of the m-th band is

[0045]

[Equation 6] When this NSR value is larger than a predetermined threshold value (for example, 0.3) (error is large), | S (j) | is approximated by | A _m || E (j) | in that band. Is not good (the above excitation signal | E (j) | is unsuitable as a basis)
Therefore, the band is determined to be UV (Unvoiced, unvoiced sound). In other cases, it can be determined that the approximation is performed to some extent, and the band is determined to be V (Voiced, voiced sound).

Next, the amplitude re-evaluation unit 108 has the amplitude on the frequency axis data from the orthogonal transformation unit 105 and the amplitude | A _m evaluated as the fine pitch from the high precision pitch search unit 106.
| And each voiced sound / unvoiced sound discrimination unit 107
V / UV (voiced sound / unvoiced sound) discrimination data from The amplitude re-evaluation unit 108 re-calculates the amplitude of the band determined to be unvoiced sound (UV) by the voiced sound / unvoiced sound determination unit 107. The amplitude | A _m | _UV for this UV band is

[0047]

[Equation 7] Required at.

The data from the amplitude re-evaluation unit 108 is
Data number conversion (a kind of sampling rate conversion) unit 10
Sent to 9. This data number conversion unit 109 is for making the number constant, considering that the number of divided bands on the frequency axis differs according to the pitch and the number of data (especially the number of amplitude data) differs. is there. That is, for example, if the effective band is up to 3400 Hz, the effective band is divided into 8 bands to 63 bands according to the pitch, and the amplitude | A obtained for each of these bands | A
_{The number of m m} (including UV band amplitude | A _m | _UV ) data m _MX +1 also changes from 8 to 63. Therefore, the data number conversion unit 109 converts the variable number m _MX +1 of amplitude data into a fixed number N _C (for example, 44) of data.

In this embodiment, dummy data for interpolating the values from the last data in the block to the first data in the block with respect to the amplitude data of one block of the effective band on the frequency axis. Is added to expand the number of data to N _F , and then the bandwidth-limited K _OS times (for example, 8
Obtain an amplitude data of K _OS times the number by performing oversampling multiplied), the K _OS times the number ((m _MX +
1) x K _OS pieces of amplitude data are linearly interpolated and expanded to a larger number of N _M pieces (for example, 2048 pieces), and the N _M pieces of data are thinned out to obtain the above-mentioned fixed number N _C (for example, 44 pieces). Convert to data.

The data from the data number conversion unit 109 (a fixed number N _C of the amplitude data) is sent to the vector quantization unit 110, and a predetermined number of data is collected into a vector, and vector quantization is performed. Is given. The quantized output data from the vector quantizer 110 is output to the output terminal 1
It is taken out via 11. The high-precision (fine) pitch data from the high-precision pitch search unit 106 is coded by the pitch coding unit 115, and the output terminal 11
It is taken out via 2. Further, the voiced sound / unvoiced sound (V / UV) discrimination data from the voiced sound / unvoiced sound discrimination unit 107 is taken out through the output terminal 113. The data from these output terminals 111 to 113 are transmitted as signals in a predetermined transmission format.

Each of these data is obtained by processing the data in the block of N samples (for example, 256 samples), but the block is a frame of L samples on the time axis. , The data to be transmitted is obtained in the frame unit. That is, the pitch data, the V / UV discrimination data, and the amplitude data are updated at the above frame cycle.

Next, a synthesizing side (decoding side) for synthesizing an audio signal on the basis of the above-mentioned respective data transmitted and obtained.
The general configuration of will be described with reference to FIG. In FIG. 7, the vector-quantized amplitude data is input to the input terminal 121, the encoded pitch data is input to the input terminal 122, and the V-value is input to the input terminal 123.
/ UV discrimination data is supplied respectively. Input terminal 12
The quantized amplitude data from 1 is the inverse vector quantization unit 12
4 and is inversely quantized, is then sent to the data number inverse transform unit 125 and is inversely transformed, and the obtained amplitude data is sent to the voiced sound synthesis unit 126 and the unvoiced sound synthesis unit 127. The encoded pitch data from the input terminal 122 is the pitch decoding unit 1
It is decoded at 28 and sent to the data number inverse conversion unit 125, the voiced sound synthesis unit 126, and the unvoiced sound synthesis unit 127. The V / UV discrimination data from the input terminal 123 is sent to the voiced sound synthesis unit 126 and the unvoiced sound synthesis unit 127.

In the voiced sound synthesis unit 126, for example, cosine (cosin
e) A voiced sound waveform on the time axis is synthesized by wave synthesis, and in the unvoiced sound synthesis unit 127, for example, white noise is filtered by a bandpass filter to synthesize the unvoiced sound waveform on the time axis, and these voiced sound synthesized waveforms are combined. The unvoiced sound synthesized waveform is added and synthesized by the addition unit 129 and is taken out from the output terminal 130. In this case, the amplitude data, the pitch data, and the V / UV discrimination data are updated and given for each frame (L sample, for example, 160 samples) at the time of the analysis, but the continuity between the frames is improved (smoothed). Therefore, each value of the amplitude data and the pitch data is set as each data value at, for example, the center position in one frame, and each data value up to the center position of the next frame (one frame at the time of composition) is interpolated. Ask by. That is,
In one frame (for example, from the center of the above analysis frame to the center of the next analysis frame) at the time of synthesis, each data value at the tip sample point and the end (the tip of the next synthesis frame)
Each data value at the sample point is given, and each data value between these sample points is obtained by interpolation.

The synthesis processing in the voiced sound synthesis unit 126 will be described in detail below. The one combined frame (L sample, for example, 160 samples) on the time axis in the m-th band (band of the m-th harmonic) determined to be V (voiced sound)
When the voiced sound for a minute is V _m (n), V _m (n) = A _m (n) cos (θ _m (n)) 0 using the time index (sample number) n in this composite frame. ≦ n <L can be expressed as (13). The final voiced sound V (n) is synthesized by adding (ΣV _m (n)) the voiced sounds of all the bands which are determined to be V (voiced sound) of all the bands.

A _m (n) in the equation (13) is the amplitude of the m-th harmonic wave which is interpolated from the beginning to the end of the composite frame. The simplest way is to linearly interpolate the value of the m-th harmonic of the amplitude data updated in frame units.
That is, the m-th frame at the tip (n = 0) of the composite frame
The amplitude value of the harmonic is A _0m , the end of the composite frame (n =
L: the amplitude value of the m-th harmonic at the next synthetic frame) is A _Lm , then A _m (n) = (Ln) A _0m / L + nA _Lm / L ... (14) It suffices to calculate A _m (n).

Next, the phase θ _m (n) in the above equation (13) is calculated by θ _m (0) = mω _O1 n + n ² m (ω _L1 −ω ₀₁ ) / 2L + φ _{0 m} + Δω n (15) You can ask. In this equation (15), φ _0m represents the phase (frame initial phase) of the m-th harmonic at the top (n = 0) of the composite frame, and ω ₀₁ represents the top of the composite frame (n = 0). The fundamental angular frequency, ω _L1, represents the fundamental angular frequency at the end of the combined frame (n = L: the leading end of the next combined frame). Δω in the equation (15) is
Set a minimum Δω such that the phase φ _{Lm at} n = L is equal to θ _m (L).

Hereinafter, how to obtain the amplitude A _m (n) and the phase θ _m (n) according to the V / UV discrimination result when n = 0 and n = L in an arbitrary m-th band will be described. To do. When the m-th band is V (voiced sound) for both n = 0 and n = L, the amplitude A _m (n) is the transmitted amplitude value A _0m by the above-mentioned equation (10), Linear interpolation of A _Lm and amplitude A
_It suffices to calculate _m (n). The phase θ _m (n) is θ when n = 0
Δω is set so that θ _m (L) becomes φ _Lm when _m (0) = φ _{0 m} and n = L.

Next, when n = 0, V (voiced sound) and n =
Amplitude A _m (n) when UV (unvoiced sound) when L
Is linearly interpolated so that 0 A _m (L) from the transmission amplitude value A _{0 m} of A _m (0). The transmission amplitude value A _{Lm when} n = L is the amplitude value of unvoiced sound and is used in unvoiced sound synthesis described later. The phase θ _m (n) is θ _m (0) = φ _{0 m} , and Δω =
Set to 0.

Further, when n = 0, UV (unvoiced sound)
When V = voiced sound when n = L, amplitude A _m
(n) is linearly interpolated so that the amplitude A _m (0) at n = 0 is 0 and the transmitted amplitude value A _Lm is n = L. Phase θ
_{For m} (n), using the phase value φ _Lm at the end of the frame as the phase θ _m (0) at n = 0, θ _m (0) = φ _Lm −m (ω _O1 + ω _L1 ) L / 2 ... (16) and Δω = 0.

When both n = 0 and n = L are V (voiced sound), Δω is set so that θ _m (L) becomes φ _Lm.
A method of setting will be described. In the above formula (15), n
= L, then θ _m (L) = mω _O1 L + L ² m (ω _L1 − ω ₀₁ ) / 2L + φ _0m + ΔωL = m (ω _O1 + ω _L1 ) L / 2 + φ _0m + ΔωL = φ _Lm . Then, Δω becomes Δω = (mod2π ((φ _Lm −φ _0m ) −mL (ω _O1 + ω _L1 ) / 2) / L (17). In this equation (17), mod 2π (x) Is the principal value of x
It is a function that returns a value between π and + π. For example, x = 1.3
mod2 π (x) = -0.7π when π, mod2 when x = 2.3π
When π (x) = 0.3π and x = -1.3π, mod2π (x) = 0.7
π, and so on.

Here, A of FIG. 8 shows an example of the spectrum of the voice signal, and the bands with band numbers (harmonics number) m of 8, 9, and 10 are UV (unvoiced sound), and other bands. Is V (voiced sound). This V
The time axis signal of the (voiced sound) band is the voiced sound synthesis unit 1 described above.
26, and the time axis signal of the UV (unvoiced sound) band is synthesized by the unvoiced sound synthesis unit 127.

The unvoiced sound synthesizing process in the unvoiced sound synthesizing section 127 will be described below. The white noise signal waveform on the time axis from the white noise generation unit 131 is windowed by a suitable window function (for example, Hamming window) with a predetermined length (for example, 256 samples), and the STFT processing unit 132 performs STFT (short circuit). By performing the (Term Fourier Transform) process, a power spectrum of white noise on the frequency axis as shown in B of FIG. 8 is obtained. The power spectrum from the STFT processing unit 132 is converted to the band amplitude processing unit 1
33, and as shown in FIG. 8C, the above-mentioned amplitude | A _m | _UV is multiplied with respect to the UV (unvoiced) band (for example, m = 8, 9, 10), and another V (voiced sound) is generated. ) Is set to 0. This band amplitude processing unit 1
The above amplitude data, pitch data, and V / UV discrimination data are supplied to 33. The output from the band amplitude processing unit 133 is sent to the ISTFT processing unit 134, and the phase is converted into a signal on the time axis by performing inverse STFT processing using the phase of the original white noise. The output from the ISTFT processing unit 134 is sent to the overlap addition unit 135, and the overlap and addition are repeated while performing appropriate weighting (so that the original continuous noise waveform can be restored) on the time axis, and continuous. Time-domain waveforms are synthesized.
The output signal from the overlap adder 135 is sent to the adder 129.

As described above, the signals of the voiced sound portion and the unvoiced sound portion which are synthesized by the synthesis units 126 and 127 and are returned to the time axis are added by the addition unit 129 at an appropriate fixed mixing ratio, The reproduced audio signal is taken out from the output terminal 130.

Regarding the configuration on the speech analysis side (encoding side) in FIG. 2 and the configuration on the speech synthesis side (decoding side) in FIG. 8, although each unit is described as hardware, a so-called DSP (digital It is also possible to realize it by a software program using a signal processor or the like.

The voiced sound discrimination method according to the present invention is also used as a means for detecting background noise when, for example, it is desired to reduce environmental noise (background noise etc.) on the transmission side of a car telephone. That is, it is also applied to noise detection in so-called speech enhancement in which low-quality speech disturbed by noise is processed, the influence of noise is removed, and a sound that is easy to hear is obtained.

[0066]

The voiced sound discrimination method according to the present invention converts an input voice signal in block units into a frequency axis, divides the frequency axis into a plurality of bands, and determines whether or not each band is a voiced sound. To determine. Then, when the signal level of the block in which the band discriminated as the voiced sound is equal to or lower than the first predetermined threshold value, the entire band in the block is made unvoiced sound, for example, in the MBE encoding, the input speech It is possible to prevent very annoying sound such as abnormal noise due to addition of fixed phase on the synthesis side when hum noise that is low pitch frequency (50Hz) is added to the background of the signal.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a schematic configuration of a voiced sound discrimination apparatus for explaining an embodiment of a voiced sound discrimination method according to the present invention.

FIG. 2 is a functional block diagram showing a schematic configuration of an analysis side (encoding side) of a voice signal synthesis analysis coding method as a specific example of an apparatus to which a voiced sound discrimination method according to the present invention is applicable.

FIG. 3 is a diagram for explaining a windowing process.

FIG. 4 is a diagram for explaining a relationship between windowing processing and a window function.

FIG. 5 is a diagram showing time axis data as an object of orthogonal transform (FFT) processing.

FIG. 6 is a diagram showing spectrum data on a frequency axis, a spectrum envelope (envelope), and a power spectrum of an excitation signal.

FIG. 7 is a functional block diagram showing a schematic configuration on a synthesis side (decoding side) of a speech signal synthesis analysis coding apparatus as a specific example of an apparatus to which a voiced sound discrimination method according to the present invention is applicable.

FIG. 8 is a diagram for explaining unvoiced sound synthesis when synthesizing a voice signal.

[Explanation of symbols]

 12 ... Window processing unit 13 ... Orthogonal transformation unit 14 ... Band division unit 15 ... V / UV discrimination unit 16 ... Correction unit 18 ...・ ・ ・ Signal level detection unit 19 ・ ・ ・ ・ ・ Comparison unit (signal level) 21 ・ ・ ・ Determination unit 22 ・ ・ ・ Peak detection unit 23 ・ ・ ・ In-block variance detection unit 24 ・ ・・ ・ ・ Normalization part 25 ・ ・ ・ Comparison part (normalized peak)

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] May 12, 1993

[Procedure Amendment 1]

[Document name to be amended] Statement

[Name of item to be corrected] Full text

[Correction method] Change

[Correction content]

[Document name] Statement

Title of invention Voiced sound discrimination method

[Claims]

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voiced sound for discriminating a voiced sound from noise or unvoiced sound from data on a frequency axis obtained by dividing an input voice signal into blocks and converting it into a frequency axis. Regarding the determination method.

[0002]

2. Description of the Related Art Speech is classified into voiced sound and unvoiced sound as a property of sound. Voiced sound is a voice accompanied by vocal cord vibration and is observed as periodic vibration. Unvoiced sound is observed as a non-periodic sound with no vocal cord vibration. Most of the normal voices are voiced sounds, and unvoiced sounds are only special consonants called unvoiced consonants. The period of voiced sound is determined by the period of vocal cord vibration, which is called the pitch period, and its reciprocal is called the pitch frequency. The pitch period and the pitch frequency (hereinafter, referred to as a pitch period when referred to as a pitch) are important factors that determine the pitch of the voice and intonation. Therefore, how accurately the pitch is captured affects the sound quality of the voice. However, when capturing the pitch, it is necessary to consider noise around the voice, so-called background noise, and quantization noise during quantization. Distinguishing these noises or unvoiced sounds from voiced sounds is important when coding speech signals.

Specific examples of the above-mentioned encoding of the voice signal include MBE (Multiband Excitation) encoding, SBE (Singleband Excitation) encoding, Harmonic encoding, and SBC ( Sub-band Coding), LP
There are C (Linear Predictive Coding), DCT (Discrete Cosine Transform), MDCT (Modified DCT), FFT (Fast Fourier Transform), and the like.

For example, in the above MBE encoding, when extracting a pitch from an input speech signal waveform, it is easy to capture the trajectory of the pitch even when a clear pitch does not appear. Then, on the decoding side (synthesis side), a cosine wave (cosin) wave synthesis is used to synthesize a voiced sound waveform on the time axis based on the above-mentioned pitch, and it is added and synthesized with a separately synthesized unvoiced sound waveform on the time axis and output. To do.

[0005]

By the way, the above MBE
In encoding, if hum noise, which is a low pitch frequency (50Hz) noise, is added to the background of the input speech signal, the synthesis side performs synthesis so that the peaks of the cosin wave overlap at that pitch. That is, each cosin wave is synthesized by addition of fixed phases as is done in the synthesis of voiced sounds. Then, by having the periodicity that the cosin waves are superposed with a fixed phase, the hum noise, which is a kind of background noise, becomes an impulse waveform with periodicity. In other words, since the intensity of the amplitude that should originally be scattered on the time axis is concentrated in one portion of the frame and has periodicity, it is reproduced as a very offensive noise. ..

The present invention has been made in view of the above circumstances, and is capable of distinguishing a voiced sound from noise or unvoiced sound with certainty, and capable of suppressing the generation of an abnormal sound on the synthesis side. The purpose is to provide a discrimination method.

[0007]

A voiced sound discrimination method according to the present invention comprises a step of dividing an input voice signal into blocks and converting them into a frequency axis to obtain frequency axis data.
Dividing the data on the frequency axis into a plurality of bands,
A step of determining whether or not there is voiced sound for each of the divided bands,
The step of detecting the level of the signal in block units in which the band determined to be the voiced sound exists by the above-mentioned discrimination, and the entire band is made unvoiced when the level of the detected signal becomes equal to or lower than a first predetermined threshold value. The above-mentioned problems are solved by the steps of:

A voiced sound discrimination method according to another invention is
There is a step of dividing the input audio signal into blocks and converting it into a frequency axis to obtain frequency axis data, a step of dividing the frequency axis data into a plurality of bands, and a step of dividing each of the divided bands. A step of determining whether or not it is a voice sound, a step of detecting the level of a signal of a block unit in which the band which is said to be a voiced sound by the above-mentioned determination exists, a step of obtaining a peak value of the signal in the block, A step of obtaining a normalized value of the peak value based on a signal level in a block, and a block unit in which the normalized value is equal to or less than a second predetermined threshold and a band in which the voiced sound is present exists To make the entire band unvoiced when the level of the signal of 1 becomes less than or equal to the first predetermined threshold value.

Here, the level of the signal may be detected from the data on the frequency axis of the block unit in which the band that is said to be voiced exists, and further detected from the data obtained by band-dividing the data on the frequency axis. You may.

Further, the above-mentioned discrimination of voiced sound is to discriminate between voiced sound, noise and unvoiced sound. It is possible to surely discriminate voiced sound and noise or unvoiced sound. That is, noise (including hum noise of a minute level) or unvoiced sound can be discriminated from the input voice signal. In such a case, for example, if the entire band of the input audio signal is forcibly made unvoiced, it is possible to suppress the generation of abnormal noise on the synthesis side.

[0011]

The input voice signal in block units is converted to the frequency axis, the data on the frequency axis is divided into a plurality of bands, and it is determined for each band whether or not it is a voiced sound. Then, when the level of the signal of the block in which the band discriminated as the voiced sound exists becomes equal to or lower than the first predetermined threshold value, the entire band in the block is made unvoiced. Therefore, generation of abnormal noise can be suppressed.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a voiced sound discrimination method according to the present invention will be described below with reference to the drawings. FIG. 1 shows a schematic configuration of a voiced sound discrimination apparatus for explaining a voiced sound discrimination method according to a first embodiment of the present invention. In this embodiment, the input audio signal is divided into blocks, and the peak value of the signal in one block of the data on the frequency axis obtained by converting the data into the frequency axis is normalized based on the signal level in the block. And the one block of the data on the frequency axis is divided into a plurality of bands, voiced sound (V) / unvoiced sound (UV) is discriminated for each of the plurality of bands, and at least one band of the plurality of bands is determined. Is voiced, the entire band in the block is unvoiced according to the signal level of the block including the voiced band. Specifically, when the normalized value is less than or equal to a predetermined threshold value and the signal level is less than or equal to a predetermined threshold value, the entire band is unvoiced.

In FIG. 1, an input terminal 11 has at least a low frequency band for removing a so-called DC (direct current) offset component and band limitation (for example, 200-3400 Hz) by a filter such as an HPF (high-pass filter) not shown. An audio signal from which components (200 Hz or less) have been removed is supplied. This signal is sent to the windowing processing unit 12. In this windowing processing unit 12, for example, a Hamming window is applied to 1 block N samples (for example, N = 256), and this 1 block is processed for 1 frame L samples (for example, L = L).
= 160), it is moved sequentially in the time axis direction,
The overlap between blocks is NL samples (96 samples, for example). This windowing processing unit 12
The signal that is a block of N samples in
3 is supplied. The orthogonal transform unit 13 adds 0 data for 1792 samples to a sample sequence of 256 samples per block (so-called zero padding) 2
With 048 samples, the time-axis data sequence of 2048 samples is subjected to orthogonal transformation processing such as FFT (Fast Fourier Transform) and converted into a frequency-axis data sequence. The data on the frequency axis from the orthogonal transform unit 13 is the band division unit 1
4 and at the same time, it is also supplied to the signal level detection unit 18 described later. The band division unit 14
Divides the supplied data on the frequency axis into a plurality of bands according to the pitch extracted by a pitch extraction unit (not shown). The data on the frequency axis divided into each band is supplied to the V (voiced sound) / UV (unvoiced sound) determination unit 15. This V
/ UV discrimination unit 15 determines V / U for each of the divided bands.
V is being discriminated. This V / UV discrimination information is supplied to the correction unit 16 and the determination unit 21.

On the other hand, the signal level detecting section 18 detects the signal level of each block of the data on the time axis from the windowing processing section 12, and detects from the block containing one band of the voiced sound. The detected signal level is supplied to the comparison unit 19. The comparison unit 19 compares the first predetermined threshold value supplied from the input terminal 20 with the signal level, and supplies the comparison result to the determination unit 21.

Here, the signal level detection unit 18 may detect the level of the signal for each block from the data on the frequency axis from the orthogonal transformation unit 13, and the band from the band division unit 14 may be detected. It may be detected from the divided data on the frequency axis.

The data on the time axis from the windowing processing unit 12 is the peak detection unit 22 and the intra-block variance detection unit 2.
3 is also supplied. The peak detection unit 22 detects a peak value of data for each block on the time axis and supplies the peak value to the normalization unit 24. The intra-block variance detector 23 detects the variance of the signal level for each block on the time axis and supplies the variance to the normalizer 24. The normalization unit 24 calculates a normalized value from the peak value and the variance of the signal level. Then, it is supplied to the comparison unit 25. The comparison unit 25 compares the second predetermined threshold value supplied from the input terminal 26 with the normalized value, and supplies the comparison result to the determination unit 21.

The judgment unit 21 controls the correction unit 16. For example, when the signal level is equal to or lower than the first predetermined threshold value and the value obtained by normalizing the peak value is equal to or lower than the second predetermined threshold value, the correction unit 16 outputs one block of data on the frequency axis. The whole band is controlled to be unvoiced. Then, the unvoiced sound information is output from the output terminal 17.

The operation of this embodiment will be described below. The number of samples N in one block cut out by applying the Hamming window in the windowing processing unit 12 is 256 samples, and the input sample sequence is x (n). This one block (256 samples) of data on the time axis is converted into one block of data on the frequency axis by the orthogonal transform unit 13. The one-block data on the frequency axis is divided into a plurality of bands by the band dividing unit 14. The data on the frequency axis of one block divided into a plurality of bands is V /
The UV discrimination unit 15 discriminates V / UV for each band. Here, when there is no voiced band at all, that is, when the band is unvoiced, the correction section 16 does not make any particular correction.

On the other hand, when the V / UV discriminating section 15 determines that at least one band is voiced sound, the signal level of one block including the voiced sound band is detected by the signal level detecting section 18. To be detected. The detected signal level is supplied to the comparison unit 19 and compared with the threshold value from the input terminal 20. Then, the comparison result is supplied to the determination unit 21.

[0020] When at least one-zone above V / UV discrimination unit 15 is a voiced sound, the level of an audio signal input is a 1LSB (least significant digit) fraction quantization step
Whether or not the signal level detection unit 18 determines
In the following cases, the entire band is unvoiced.

Specifically, the signal level detector 18 calculates the standard deviation σ of energy as

[0022]

[Equation 1]

It is calculated by the equation (1) shown by. Where x is

[0024]

[Equation 2]

It is the average value of x (n) represented by

The above σ is supplied to the comparison unit 19 when the V / UV discrimination unit 15 determines that at least one band is voiced sound. The comparison unit 19 compares the threshold value g _t (for example, g _t = 8) supplied to the input terminal 20 with the above σ . Here, if σ is smaller than the threshold value g _t , the determination unit 21 determines that the signal of one block in which at least one band is voiced by the V / UV determination unit 15 is a granular noise level, that is, 1 LSB. The following average amplitudes are determined.

Further, the peak detecting section 22 calculates the peak value P _e of the data of one block as

[0028]

[Equation 3]

It is detected by the equation (3) shown by. Here, x is the average value of x (n) as described above. And
The normalization unit 24 divides the peak value P _e by the standard deviation σ shown in the equation (1) to obtain a normalized value P _n . That is, the normalized value P _n is represented by P _n = P _e / σ (4) Then, the comparison unit 25 compares the normalized value P _n with the threshold P _thn supplied from the input terminal 26, and supplies the comparison information to the determination unit 21. This determination unit 21
Is compared with the standard deviation σ and the input terminal 20.
The comparison information obtained by comparing with the threshold value g _t (for example, g _t = 8) supplied from the above is also supplied. Then, the judgment unit 21 supplies a control signal to the correction unit 16 according to the comparison information from the comparison unit 25 and the comparison unit 19. Specifically, the normalized value P _n is less than or _equal to a predetermined threshold P _t , and the signal level σ is also a predetermined threshold g.
_{When t is} less than or equal to _{t, the} correction unit 16 is caused to function to make the entire band of the data on the frequency axis of one block unvoiced.

Therefore, in this embodiment, when the signal level of one block of the signals other than the unvoiced sound in the entire band, that is, the signal in which at least one band is determined to be the voiced sound is small and the peak value is small, In addition, by forcibly making all bands unvoiced, it is possible to prevent the generation of abnormal noise due to fixed phase addition that occurs on the combining side.

The voiced sound discrimination method according to the present invention uses M
It is applicable to vocoders such as BE. That is, when the data on the frequency axis of one block of the input audio signal is made noise or unvoiced sound, all the bands are forced to be unvoiced sound to prevent the generation of abnormal sound on the synthesis side of the vocoder such as MBE. You can

A specific example of an MBE (Multiband Excitation) vocoder, which is a kind of speech signal synthesis analysis coding apparatus (so-called vocoder) to which the voiced sound discrimination method according to the present invention is applicable, will be described below with reference to the drawings. It will be explained with reference to FIG. This MBE vocoder is DW Grif
fin and JS Lim, "Multiband Excitation Vocoder,"
IEEE Trans. Acoustics, Speech, and Signal Processing,
Vol.36, No.8, pp.1223-1235, Aug.1988, the conventional PARCOR (PARtial auto-C
In ORrelation: partial autocorrelation) vocoders, voiced sections and unvoiced sections were switched for each block or frame when modeling speech, whereas in MBE vocoder, the same time (in the same block or frame) Voiced section (Voiced) and unvoiced section (Unvoic)
ed) section is modeled on the assumption that and exist.

FIG. 2 is a block diagram showing the overall schematic configuration of the embodiment of the MBE vocoder. In FIG. 2, an audio signal is supplied to an input terminal 101, and this input audio signal is sent to a filter 102 such as an HPF (high-pass filter) to be a so-called DC signal.
Removal of (DC) offset and band limitation (for example, 200
At least low frequency component (2)
(Less than 00 Hz) is removed. The signal obtained through the filter 102 is sent to the pitch extraction unit 103 and the windowing processing unit 104, respectively. Pitch extraction unit 103
In, the input voice signal data is divided into blocks in units of a predetermined number N (for example, N = 256) (or cut out by a rectangular window), and pitch extraction is performed on the voice signals in this block. Such a cut block (256 samples) is moved in the time axis direction at a frame interval of L samples (for example, L = 160) as shown in A of FIG. 3, and the overlap between the blocks is N−. There are L samples (for example, 96 samples). In addition, in the windowing processing unit 104, 1 block N
A predetermined window function, for example, a Hamming window is applied to the sample, and this windowed block is sequentially moved in the time axis direction at intervals of 1 frame L sample.

When this windowing process is expressed by a mathematical expression, x _w (k, q) = x (q) w (kL-q) (5) In this equation (5), k is a block number,
q represents the time index (sample number) of the data, and the q-th data x (q) of the unprocessed input signal is windowed by the window function (w (kL-q)) of the kth block. It is shown that the data x _w (k, q) can be obtained by doing so. 3A in the pitch extraction unit 103
The window function w _r (r) in the case of the rectangular window is as follows: w _r (r) = 1 0 ≦ r <N (6) = 0 r <0, N ≦ r window function w _h (r) in the case of Hamming window as shown in B of FIG. 3 is a section _{104, w h (r) =} 0.54 - 0.46 cos (2πr / (N-1)) 0 ≦ r <N (7) = 0 r <0, N ≦ r. Such a window function w _r (r) or w (5) when using the _h (r) formula of the window function w (r) (= w (KL-
q)), the zero-zero interval is 0 ≦ kL−q <N, which is transformed into kL−N <q ≦ kL. Therefore, for example, in the case of the above rectangular window, the window function w _r (kL−q) =
As shown in FIG. 4, 1 becomes kL-N <q ≦ kL.
It will be when. Further, the above equations (5) to (7) are expressed by the length N
The window of (= 256) samples is shown to advance by L (= 160) samples. Below, above (6)
Each of the N points (0 ≦ r
The non-zero sample sequence of <N) is represented by x _wr (k, r) and x _wh
Let us denote it as (k, r).

In the windowing processing unit 104, as shown in FIG. 5, 179 is applied to the sample sequence x _wh (k, r) of one block of 256 samples to which the Hamming window of the equation (7) is applied.
Two samples of 0 data are added (so-called zero padding) to make 2048 samples, and the orthogonal transformation unit 105 performs orthogonal transformation such as FFT (Fast Fourier Transform) on the time-axis data sequence of 2048 samples. Processing is performed.

In the pitch extraction unit 103, the above x _wr (k, r)
Pitch extraction is performed based on the sample sequence (1 block N samples). The pitch extraction method is known to include periodicity of time waveform, periodic frequency structure of spectrum, and autocorrelation function.
The center correlation waveform autocorrelation method is used. Regarding the center clip level in the block at this time, one clip level may be set for one block, but the peak level of the signal of each part (each sub-block) obtained by subdividing the block is detected and When there is a large difference in the peak level of each sub-block, the clip level is changed stepwise or continuously within the block. The pitch period is determined based on the peak position of the autocorrelation data of this center clip waveform. At this time, a plurality of peaks are obtained from the autocorrelation data belonging to the current frame (the autocorrelation is obtained for the data of N samples of one block), and the maximum peak among the plurality of peaks is equal to or larger than a predetermined threshold. In the case of, the maximum peak position is set as the pitch period, and in other cases, the pitch is within a pitch range that satisfies a predetermined relationship with the pitch other than the current frame, for example, the pitch of the previous frame and the pitch of the previous frame. As a result, a peak in the range of ± 20% is obtained, and the pitch of the current frame is determined based on this peak position. In this pitch extraction unit 103, a relatively rough pitch search is performed by an open loop, and the extracted pitch data has a high precision (fine) pitch search unit 10.
Then, the high precision pitch search (pitch fine search) is performed by the closed loop.

High precision (fine) pitch search section 106
Includes rough pitch data of integer (integer) values extracted by the pitch extraction unit 103 and the orthogonal transformation unit 10.
5, the data on the frequency axis subjected to FFT, for example, is supplied. In this high precision pitch search unit 106,
Centering on the above coarse pitch data value, it is ± 0.2 in increments of ±
Shake several samples at a time to drive to the optimum fine pitch data value with a decimal point (floating). As a fine search method at this time, a so-called analysis by synthesis method is used, and the pitch is selected so that the synthesized power spectrum is closest to the power spectrum of the original sound.

The fine search of the pitch will be described. First, in the MBE vocoder, the FF
S (j) as spectrum data on the frequency axis orthogonally transformed by T etc. is expressed as S (j) = H (j) | E (j) | 0 <j <J (8) It is assumed that the model. Where J is
corresponding to ω _s / 4π = f _s / 2, and the sampling frequency f
_{When s} = ω _s / 2π is 8 kHz, for example, it corresponds to 4 kHz. In the equation (8), when the spectrum data S (j) on the frequency axis has a waveform as shown in A of FIG.
(j) is the original spectrum data S as shown in B of FIG.
The spectral envelope of (j) is shown as E
(j) shows the spectrum of the excitation signal (excitation) which is cyclic at the same level as shown in C of FIG. That is, the FFT spectrum S (j) is the spectrum envelope H (j) and the power spectrum | E of the excitation signal.
(j) | is modeled as the product.

Power spectrum of the excitation signal | E (j)
Is a spectral waveform corresponding to the waveform of one band (band) for each band on the frequency axis in consideration of the periodicity (pitch structure) of the waveform on the frequency axis determined according to the pitch. It is formed by arranging it repeatedly. The waveform for one band is obtained by performing FFT by regarding a waveform obtained by adding (zero-filling) 0 data for 1792 samples to a Hamming window function of 256 samples as shown in FIG. 5 as a time axis signal. It can be formed by cutting out an impulse waveform having a certain bandwidth on the frequency axis according to the pitch.

Next, for each band divided according to the above pitch, a value (a kind of amplitude) | A that represents the above H (j) (minimizes the error for each band) | A _m
Ask for |. Here, for example, when the lower and upper points of the m-th band (band of the m-th harmonic) are a _m and b _m , respectively, the error ε _m of the m-th band is

[0041]

[Equation 4] Can be expressed as | A _m | that minimizes this error ε _m
Is

[0042]

[Equation 5]

Therefore, when | A _m | in this equation (10),
Minimize the error ε _m . Such an amplitude | A _m | is obtained for each band, and the obtained amplitude | A _m | is used to obtain the error ε m for each band defined by the above equation (9). Next, the sum total value Σε _m of all the bands of such error ε _m for each band is obtained. Further, such an error sum value Σε _m of all bands is obtained for some slightly different pitches, and a pitch that minimizes the error sum value Σε _m is obtained.

That is, several kinds of vertical pitches are prepared with the rough pitch obtained by the pitch extraction unit 103 as the center, for example, in steps of 0.25. The error sum value Σε _m is obtained for each of these plural kinds of slightly different pitches. In this case, when the pitch is determined, the bandwidth is determined. From the above equation (10), the power spectrum | S (j) |
(j) | and the error ε _m in the above equation (9) can be obtained, and the sum total value Σε _m of all the bands can be obtained. This error sum value Σε _m is obtained for each pitch, and the pitch corresponding to the minimum error sum value is determined as the optimum pitch. As described above, the high-precision pitch search unit 106 obtains the optimum fine (eg, 0.25 step) pitch, and the amplitude | A _m | corresponding to this optimum pitch.
Is determined.

In the above description of the pitch fine search, in order to simplify the explanation, all bands are voiced (Vo
Assuming the case of iced), as described above, MBE
In the vocoder, unvoiced sound (Un
Since a model in which a voiced) area exists is used, it is necessary to distinguish voiced sound / unvoiced sound for each band.

The optimum pitch and amplitude | A _m | data from the high precision pitch search unit 106 is sent to the voiced sound / unvoiced sound determination unit 107, and the voiced sound / unvoiced sound is discriminated for each band. NSR (noise to signal ratio) is used for this determination. That is, the NSR of the m-th band is

[0047]

[Equation 6]

If this NSR value is larger than a predetermined threshold value (eg, 0.3) (error is large), | A _m || E (j) | due to | S (j) in that band It can be judged that the approximation of | is not good (the above excitation signal | E (j) | is unsuitable as a basis), and the band is UV (Unvoice).
d, unvoiced sound). In other cases, it can be judged that the approximation has been performed to some extent, and the band is set to V
(Voiced, voiced sound).

Next, in the amplitude re-evaluation section 108, the amplitude on the frequency axis data from the orthogonal transformation section 105 and the amplitude | A _m evaluated as the fine pitch from the high precision pitch search section 106.
| And each voiced sound / unvoiced sound discrimination unit 107
V / UV (voiced sound / unvoiced sound) discrimination data from The amplitude re-evaluation unit 108 re-calculates the amplitude of the band determined to be unvoiced sound (UV) by the voiced sound / unvoiced sound determination unit 107. The amplitude | A _m | _UV for this UV band is

[0050]

[Equation 7] Required at.

The data from the amplitude re-evaluation unit 108 is
Data number conversion (a kind of sampling rate conversion) unit 10
Sent to 9. This data number conversion unit 109 is for making the number constant, considering that the number of divided bands on the frequency axis differs according to the pitch and the number of data (especially the number of amplitude data) differs. is there. That is, for example, if the effective band is up to 3400 Hz, the effective band is divided into 8 bands to 63 bands according to the pitch, and the amplitude | A obtained for each of these bands | A
_{The number of m m} (including UV band amplitude | A _m | _UV ) data m _MX +1 also changes from 8 to 63. Therefore, the data number conversion unit 109 converts the variable number m _MX +1 of amplitude data into a fixed number N _C (for example, 44) of data.

Here, in this embodiment, dummy data for interpolating values from the last data in the block to the first data in the block for the amplitude data of one block of the effective band on the frequency axis. Is added to expand the number of data to N _F , and then the bandwidth-limited K _OS times (for example, 8
Obtain an amplitude data of K _OS times the number by performing oversampling multiplied), the K _OS times the number ((m _MX +
1) amplitude data of × K _OS pieces) extended to linear interpolation to more NM number (e.g. 2048), the data of the N thinned out _M data the predetermined number N _C (e.g. 44) Convert to.

The data from the data number conversion unit 109 (the above-mentioned fixed number N _C of amplitude data) is sent to the vector quantization unit 110, and a predetermined number of data is collected into a vector, and vector quantization is performed. Is given. The quantized output data from the vector quantizer 110 is output to the output terminal 1
It is taken out via 11. The high-precision (fine) pitch data from the high-precision pitch search unit 106 is coded by the pitch coding unit 115, and the output terminal 11
It is taken out via 2. Further, the voiced sound / unvoiced sound (V / UV) discrimination data from the voiced sound / unvoiced sound discrimination unit 107 is taken out through the output terminal 113. The data from these output terminals 111 to 113 are transmitted as signals in a predetermined transmission format.

Note that each of these data is obtained by processing the data in the block of N samples (for example, 256 samples), but the block is a frame of L samples on the time axis. , The data to be transmitted is obtained in the frame unit. That is, the pitch data, the V / UV discrimination data, and the amplitude data are updated at the above frame cycle.

Next, a synthesizing side (decoding side) for synthesizing a voice signal on the basis of the above-mentioned respective data transmitted and obtained.
The general configuration of will be described with reference to FIG. In FIG. 7, the vector-quantized amplitude data is input to the input terminal 121, the encoded pitch data is input to the input terminal 122, and the V-value is input to the input terminal 123.
/ UV discrimination data is supplied respectively. Input terminal 12
The quantized amplitude data from 1 is the inverse vector quantization unit 12
4 and is inversely quantized, is then sent to the data number inverse transform unit 125 and is inversely transformed, and the obtained amplitude data is sent to the voiced sound synthesis unit 126 and the unvoiced sound synthesis unit 127. The encoded pitch data from the input terminal 122 is the pitch decoding unit 1
It is decoded at 28 and sent to the data number inverse conversion unit 125, the voiced sound synthesis unit 126, and the unvoiced sound synthesis unit 127. The V / UV discrimination data from the input terminal 123 is sent to the voiced sound synthesis unit 126 and the unvoiced sound synthesis unit 127.

In the voiced sound synthesis unit 126, for example, cosine (cosin)
e) A voiced sound waveform on the time axis is synthesized by wave synthesis, and in the unvoiced sound synthesis unit 127, for example, white noise is filtered by a bandpass filter to synthesize the unvoiced sound waveform on the time axis, and these voiced sound synthesized waveforms are combined. The unvoiced sound synthesized waveform is added and synthesized by the addition unit 129 and is taken out from the output terminal 130. In this case, the amplitude data, the pitch data, and the V / UV discrimination data are updated and given for each frame (L sample, for example, 160 samples) at the time of the analysis, but the continuity between the frames is improved (smoothed). Therefore, each value of the amplitude data and the pitch data is set as each data value at, for example, the center position in one frame, and each data value up to the center position of the next frame (one frame at the time of composition) is interpolated. Ask by. That is,
In one frame (for example, from the center of the above analysis frame to the center of the next analysis frame) at the time of synthesis, each data value at the tip sample point and the end (the tip of the next synthesis frame)
Each data value at the sample point is given, and each data value between these sample points is obtained by interpolation.

The synthesis processing in the voiced sound synthesis unit 126 will be described in detail below. The one combined frame (L sample, for example, 160 samples) on the time axis in the m-th band (band of the m-th harmonic) determined to be V (voiced sound)
When the voiced sound for a minute is V _m (n), V _m (n) = A _m (n) cos (θ _m (n)) 0 using the time index (sample number) n in this composite frame. ≦ n <L can be expressed as (13). The final voiced sound V (n) is synthesized by adding (ΣV _m (n)) the voiced sounds of all the bands which are determined to be V (voiced sound) of all the bands.

A _m (n) in the equation (13) is the amplitude of the m-th harmonic wave interpolated from the beginning to the end of the composite frame. The simplest way is to linearly interpolate the value of the m-th harmonic of the amplitude data updated in frame units.
That is, the m-th frame at the tip (n = 0) of the composite frame
The amplitude value of the harmonic is A _0m , the end of the composite frame (n =
L: the amplitude value of the m-th harmonic at the next synthetic frame) is A _Lm , then A _m (n) = (Ln) A _0m / L + nA _Lm / L ... (14) It suffices to calculate A _m (n).

Next, the phase θ _m (n) in the above equation (13) is calculated by θ _m (n) = mω _O1 n + n ² m (ω _L1 −ω ₀₁ ) / 2L + φ _{0 m} + Δω n (15) You can ask. In this equation (15), φ _0m represents the phase (frame initial phase) of the m-th harmonic at the top (n = 0) of the composite frame, and ω ₀₁ represents the top of the composite frame (n = 0). The fundamental angular frequency, ω _L1, represents the fundamental angular frequency at the end of the combined frame (n = L: the leading end of the next combined frame). Δω in the equation (15) is
Set a minimum Δω such that the phase φ _{Lm at} n = L is equal to θ _m (L).

Hereinafter, how to obtain the amplitude A _m (n) and the phase θ _m (n) according to the V / UV discrimination result when n = 0 and n = L in an arbitrary m-th band will be described. To do. When the m-th band is V (voiced sound) for both n = 0 and n = L, the amplitude A _m (n) is the transmitted amplitude value A _0m by the above-mentioned equation (10), Linear interpolation of A _Lm and amplitude A
_It suffices to calculate _m (n). The phase θ _m (n) is θ when n = 0
Δω is set so that θ _m (L) becomes φ _Lm when _m (0) = φ _{0 m} and n = L.

Next, when n = 0, V (voiced sound) and n =
Amplitude A _m (n) when UV (unvoiced sound) when L
Is linearly interpolated so that 0 A _m (L) from the transmission amplitude value A _{0 m} of A _m (0). The transmission amplitude value A _{Lm when} n = L is the amplitude value of unvoiced sound and is used in unvoiced sound synthesis described later. The phase θ _m (n) is θ _m (0) = φ _{0 m} , and Δω =
Set to 0.

Further, when n = 0, UV (unvoiced sound)
When V = voiced sound when n = L, amplitude A _m
(n) is linearly interpolated so that the amplitude A _m (0) at n = 0 is 0 and the transmitted amplitude value A _Lm is n = L. Phase θ
_{For m} (n), using the phase value φ _Lm at the end of the frame as the phase θ _m (0) at n = 0, θ _m (0) = φ _Lm −m (ω _O1 + ω _L1 ) L / 2 ... (16) and Δω = 0.

When both n = 0 and n = L are V (voiced sound), Δω is set so that θ _m (L) becomes φ _Lm.
A method of setting will be described. In the above formula (15), n
= L, then θ _m (L) = mω _O1 L + L ² m (ω _L1 −ω ₀₁ ) / 2L + φ _0m + ΔωL = m (ω _O1 + ω _L1 ) L / 2 + φ _0m + ΔωL = φLm. , Δω is Δω = (mod 2π ((φ _Lm −φ _0m ) −mL (ω _O1 + ω _L1 ) / 2) / L (17). In this equation (17), mod 2π (x) is , The main value of x is −
It is a function that returns a value between π and + π. For example, x = 1.3
mod2 π (x) = -0.7π when π, mod2 when x = 2.3π
When π (x) = 0.3π and x = -1.3π, mod2π (x) = 0.7
π, and so on.

Here, A of FIG. 8 shows an example of a spectrum of a voice signal. Each band of band numbers (harmonics number) m of 8, 9 and 10 is UV (unvoiced sound), and other bands. Is V (voiced sound). This V
The time axis signal of the (voiced sound) band is the voiced sound synthesis unit 1 described above.
26, and the time axis signal of the UV (unvoiced sound) band is synthesized by the unvoiced sound synthesis unit 127.

The unvoiced sound synthesizing process in the unvoiced sound synthesizing section 127 will be described below. The white noise signal waveform on the time axis from the white noise generation unit 131 is windowed by a suitable window function (for example, Hamming window) with a predetermined length (for example, 256 samples), and the STFT processing unit 132 performs STFT (short circuit). By performing the (Term Fourier Transform) process, a power spectrum of white noise on the frequency axis as shown in B of FIG. 8 is obtained. The power spectrum from the STFT processing unit 132 is converted to the band amplitude processing unit 1
33, and as shown in FIG. 8C, the above-mentioned amplitude | A _m | _UV is multiplied with respect to the UV (unvoiced) band (for example, m = 8, 9, 10), and another V (voiced sound) is generated. ) Is set to 0. This band amplitude processing unit 1
The above amplitude data, pitch data, and V / UV discrimination data are supplied to 33. The output from the band amplitude processing unit 133 is sent to the ISTFT processing unit 134, and the phase is converted into a signal on the time axis by performing inverse STFT processing using the phase of the original white noise. The output from the ISTFT processing unit 134 is sent to the overlap addition unit 135, and the overlap and addition are repeated while performing appropriate weighting (so that the original continuous noise waveform can be restored) on the time axis, and continuous. Time-domain waveforms are synthesized.
The output signal from the overlap adder 135 is sent to the adder 129.

In this way, the signals of the voiced sound portion and the unvoiced sound portion which are synthesized in the respective synthesis units 126 and 127 and are returned to the time axis are added by the addition unit 129 at an appropriate fixed mixing ratio, The reproduced audio signal is taken out from the output terminal 130.

Regarding the configuration on the voice analysis side (encoding side) in FIG. 2 and the configuration on the voice synthesis side (decoding side) in FIG. 8, each part is described as hardware, but a so-called DSP (digital It is also possible to realize it by a software program using a signal processor or the like.

The voiced sound discrimination method according to the present invention is also used as means for detecting background noise when it is desired to reduce environmental noise (background noise or the like) on the transmitting side of a car telephone. That is, it is also applied to noise detection in so-called speech enhancement in which low-quality speech disturbed by noise is processed, the influence of noise is removed, and a sound that is easy to hear is obtained.

[0069]

The voiced sound discrimination method according to the present invention converts an input voice signal in block units into a frequency axis, divides the frequency axis into a plurality of bands, and determines whether or not each band is a voiced sound. To determine. Then, when the signal level of the block in which the band discriminated as the voiced sound is equal to or lower than the first predetermined threshold value, the entire band in the block is made unvoiced sound, for example, in the MBE encoding, the input speech It is possible to prevent very annoying sound such as abnormal noise due to addition of fixed phase on the synthesis side when hum noise that is low pitch frequency (50Hz) is added to the background of the signal.

[Brief description of drawings]

FIG. 1 is a functional block diagram showing a schematic configuration of a voiced sound discrimination apparatus for explaining an embodiment of a voiced sound discrimination method according to the present invention.

FIG. 2 is a functional block diagram showing a schematic configuration of an analysis side (encoding side) of a voice signal synthesis analysis coding method as a specific example of an apparatus to which a voiced sound discrimination method according to the present invention is applicable.

FIG. 3 is a diagram for explaining a windowing process.

FIG. 4 is a diagram for explaining a relationship between windowing processing and a window function.

FIG. 5 is a diagram showing time axis data as an object of orthogonal transform (FFT) processing.

FIG. 6 is a diagram showing spectrum data on a frequency axis, a spectrum envelope (envelope), and a power spectrum of an excitation signal.

FIG. 7 is a functional block diagram showing a schematic configuration on a synthesis side (decoding side) of a speech signal synthesis analysis coding apparatus as a specific example of an apparatus to which a voiced sound discrimination method according to the present invention is applicable.

FIG. 8 is a diagram for explaining unvoiced sound synthesis when synthesizing a voice signal.

[Explanation of Codes] 12 ... Windowing processing unit 13 ... Orthogonal transformation unit 14 ... Band division unit 15 ... V / UV discrimination unit 16 ... Modifying unit 18-Signal level detecting unit 19-Comparison unit (signal level) 21-Judging unit 22-Peak detecting unit 23-In-block variance Detection unit 24 ... Normalization unit 25 ... Comparison unit (normalized peak)

Claims (2)

[Claims] 1. A step of dividing an input audio signal into blocks and converting it into a frequency axis to obtain frequency axis data; a step of dividing the frequency axis data into a plurality of bands; A step of determining whether each band is a voiced sound, a step of detecting a signal level of a block unit in which a band determined to be a voiced sound by the above determination exists, and a level of the detected signal is A voiced sound determination method, wherein the whole band is made unvoiced when it becomes equal to or less than a predetermined threshold value of 1. 2. A step of dividing an input audio signal in block units and converting it into a frequency axis to obtain frequency axis data, and a step of dividing the frequency axis data into a plurality of bands. A step of determining whether or not each band is voiced sound, a step of detecting a signal level of a block unit in which the band determined to be voiced by the above determination exists, and a peak value of the signal in the block And a step in which the peak value is normalized based on the signal level in the block, and a band in which the normalized value is equal to or less than a second predetermined threshold and is the voiced sound. A voiced sound discrimination method, wherein the whole band is made unvoiced when the level of the block-based signal in which is present is equal to or lower than a first predetermined threshold value.