JP3074046B2 - Voice / music sound identification circuit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice / musical tone discriminating circuit for discriminating a human voice signal from a musical tone signal of a musical instrument.

[0002]

2. Description of the Related Art Conventionally, techniques in such a field include:
For example, there is one described in the following literature. Literature; IEEJ Information and Systems Division National Convention, 1982, Proceedings 380 Hosoda et al., "A Study on Data Identification Method for DCME Overload Protection", p. Conventionally, as described in the above literature, in the field of international communication using satellites and the like, DSI (Digital Speach Intepolation) has been used.
) Technology and ADPCM (adaptive differential pulse-c)
2. Description of the Related Art A digital line multiplexer (hereinafter referred to as DCME) which combines ode modulation technology has attracted attention. In this DCME, when a transmission channel becomes overloaded, bits allocated to the transmission channel (usually 4 bits)
Is used to reduce the call by one bit, and the call lock-out is used. However, voiceband data (voicebanddata)
If the channel transmitting the modem signal consisting of 1 bit is reduced by 1 bit, the transmission characteristics deteriorate. Therefore, ADPC
It is necessary to identify whether the M input signal is voice band data. Therefore, in the technique of the above-mentioned document, a DCME overload protection data identification algorithm is used to identify whether an input signal is a voice signal or voice band data.

[0003]

However, if a voice / tone discriminating circuit for discriminating voice and tone is constructed by using a conventional method for discriminating between a voice signal and a modem signal (voicebanddata), the following will occur. There were challenges.

[0004] Here, a musical tone is a sound that can be separated from a human speaking voice signal. The audio signal is
It is a means of communicating intentions for realizing a conversation, and it is a sound that only needs to secure a certain level of quality. Speech is generated from the human vocal tract and is given spectral properties mainly by formants. A musical tone is one that gives a musically rich feeling, and has a characteristic that the vibration is periodic and the frequency of harmonics forms a positive ratio. As an example of these, FIG. 2 shows a time-series waveform diagram of an audio signal, and FIG.
2 shows time-series waveform charts of the tone signal. The tone signal of FIG. 3 is a waveform diagram of an orchestra. 2 and 3
The waveform diagram drawn below each waveform diagram is a partially enlarged view of the upper diagram.

When such an audio signal is distinguished from a tone signal, the characteristics of the two signals are too different, so that the audio signal is distinguished from the tone signal, and the quality of the band corresponding to the audio signal is adjusted. It is difficult to provide a high-quality and efficient voice / musical sound discriminating circuit by improving the intelligibility and intelligibility while handling the musical sound signal in a wide band corresponding to the musical sound signal.

SUMMARY OF THE INVENTION It is an object of the present invention to distinguish an audio signal from a tone signal, improve the clarity and intelligibility of the audio signal, and provide a high-quality and efficient tone signal. It is an object of the present invention to provide a voice / musical tone discriminating circuit which solves the difficulty of identifying a tone.

[0007]

To solve the previous SL problems SUMMARY OF THE INVENTION The first aspect of the present invention, speech and tone identify circuit identifies the audio signal and the musical tone signal from the digitized input signals of the present invention An autocorrelation coefficient calculating means for obtaining an autocorrelation coefficient of the input signal; a first threshold value judging and accumulating means for accumulating the number of times the autocorrelation coefficient has a value larger than a positive threshold value; a second threshold determination integrator for integrating the number of times the correlation coefficient is generated by the negative threshold value smaller than the integrated value of the first threshold determination integrating means and the second
The number-of-integration number determination means for determining whether or not the integrated value of the threshold value determination integration means exceeds a predetermined threshold value, and the determination result of the integration number determination means is the same determination result for a certain section.
Bei Eteiru counting identification means for identifying an audio signal or tone signal by detecting whether the. The second invention is the first invention
In the voice / music sound discriminating circuit of the invention,
The means comprises an integrated value of the first threshold value judgment integrating means and the second
The sum of the threshold value and the integrated value of the threshold value integrating means exceeds a predetermined threshold value.
It is configured to determine whether or not

[0008]

According to the first and second aspects of the present invention, since the voice / musical tone discriminating circuit is configured as described above, when a digitized signal is input, the autocorrelation coefficient calculating means causes the autocorrelation coefficient calculating means to execute the self-phase correlation. The number is obtained and sent to the first and second threshold value judging and integrating means. The first threshold value judging and integrating means judges whether or not the autocorrelation coefficient is larger than a positive threshold value, and counts the number of occurrences of a value larger than the positive threshold value. Similarly, the second threshold value judging and integrating means judges whether or not the autocorrelation coefficient is smaller than a negative threshold value, and counts the number of occurrences of the small value.

The number-of-integration-number determining means determines whether each of the two integrated values of the first and second threshold value determining and integrating means exceeds a predetermined threshold value, or determines whether the sum of the two integrated values is a predetermined value. > it determines whether exceeds the threshold, sends the determination result to the counting identification means. The counting identification means is the counting number determination means .
A determination is made as to whether or not the determination result is the same for a certain section, and a voice signal and a musical sound signal are identified. Therefore, the above problem can be solved.

[0010]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a voice / audio system according to a first embodiment of the present invention.
FIG. 3 is a functional block diagram of a tone discrimination circuit. This voice / musical tone discriminating circuit calculates the power of a digitized input signal Xi to determine whether there is sound or no sound.
And an auto-correlation coefficient calculating means 2 for calculating an auto-correlation coefficient R (t) from the input signal Xi based on the soundness judgment of the power calculating means 1. On the output side of the autocorrelation coefficient calculation means 2, first and second threshold value judgment integration means 10, 2
0 is connected.

The first threshold value judging / integrating means 10 has a function of accumulating the number of times that the value of the autocorrelation coefficient R (t) is larger than the positive threshold value PK1, and calculates the autocorrelation coefficient R (t). A threshold judging means 11 for outputting when the autocorrelation coefficient R (t) is larger than a positive threshold PK1, and an integrated value M1 obtained by integrating the number of outputs of the threshold judging means 11 And the number-of-times-of-judgment accumulating means 12.
Similarly, the second threshold value judgment integration means 20 has a function of integrating the number of times that the value of the autocorrelation coefficient R (t) is smaller than the negative threshold value PK2, and the autocorrelation coefficient R (t) Is compared with a negative threshold value PK2 to output when the autocorrelation coefficient R (t) is small.
And a judgment number accumulating means 22 that accumulates the number of times of output of 1 and outputs an integrated value M2.

First and second threshold value judgment integrating means 10, 20
The output side is connected to the integrated number determination means 30, and the output side is connected to the count identification means 40. The number-of-integration number determination means 30 determines whether or not the integrated values M1 and M2 exceed the threshold values K1 and K2, and outputs a determination result of the number when the conditions of M1> K1 and M2> K2 hold. have.

The counting discriminating means 40 includes an integrated number judging means 3
Has the function of counting the continuity of the number of occurrences at 0,
The two count-up conditions (CNT1, CNT2) are connected to be input to the other resets R1, R2.
The counters 41 and 42, an inverter 43 for inverting the output of the integration number judging means 30, and the carrier C 1 of the counter 42 are reset by the carrier CR 1 of the counter 41.
A flip-flop (hereinafter, referred to as FF) 44 that is reset by R2 and outputs a tone signal Y1 or an audio signal Y2 from the normal phase output terminal and the negative phase output terminal.

Next, the operation of FIG. 1 will be described with reference to FIGS. FIG. 4 is an operation flowchart of FIG. 1, and S1 to S17 represent each processing step. FIG.
2 is a diagram showing the autocorrelation coefficient R (t) in FIG. 1, in which the output is normalized to 1 at the maximum.

In FIG. 4, when the operation of the voice / tone discriminating circuit of FIG. 1 starts, the digitized input signal Xi
Are stored in a memory (not shown) in block units (frame units) in step S1. The input signal Xi for each block stored in the memory is input to the power calculation means 1 and the autocorrelation coefficient calculation means 2. The power calculation means 1 calculates the power of the input signal Xi in step S2, and in step S3, sets the signal power to the threshold value Pt.
h, the signal power is determined to be greater than the threshold Pth
If it is larger, it is regarded as a sound, and a calculation instruction is issued to the autocorrelation coefficient calculating means 2. Other than that, it is regarded as silence, and the operation ends without discriminating the audio signal and the tone signal in this block section.

In the autocorrelation coefficient calculating means 2, step S
4, the autocorrelation coefficient R (t) is calculated by the following equation.
Is calculated.

[0017]

(Equation 1)

The calculated autocorrelation coefficient R (t) is shown in FIG.
Are sent to the first and second threshold value judging and integrating means 10 and 20.

In step S5 of FIG. 4, the integrated values M1 and M2 of the judgment number integrating means 12 and 22 are initialized to zero. In step S7 through step S6, the threshold value judging means 11 sets the autocorrelation coefficient R (t) to a positive threshold value PK1.
And the autocorrelation coefficient R is larger than the positive threshold PK1.
When (t) is large, it is output to the judgment number accumulating means 12. The judgment number accumulating means 12 counts up in step S8. On the other hand, in step S9, the threshold value judging means 21 compares the autocorrelation coefficient R (t) with the negative threshold value P
K2, and the autocorrelation coefficient R (t) is set to a negative threshold value P
When it is smaller than K2, it is output to the judgment number accumulating means 22. The judgment number accumulating means 22 counts up in step S10. The integration of these determination times is performed in step S
6 is repeated by the number N of signals in the block.

The integrated value M of the judgment number integrating means 12 and 22
1 and M2 are sent to the integration number determination means 30. In the integration number determination means 30, when one block is completed, step S1 is executed.
At 1, it is determined whether or not the conditions of M1> K1 and M2> K2 are satisfied. The determination result is provided to the counter 41 in the counting and identifying means 40, and is inverted by the inverter 43 and provided to the counter 42. If the conditions of M1> K1 and M2> K2 are satisfied in step S11, step S12
The counter 41 counts up by the reset R1
Resets the counter 42 to zero. Conversely, if the conditions of M1> K1 and M2> K2 are not satisfied in step S11, the counter 42 counts up in step S13, and the counter R4 is reset by the reset R2.
1 is reset to 0.

When each of the counters 41 and 42 counts a predetermined number of times, it outputs the carriers CR1 and CR2 as an overflow. These counters 41 and 42
Then, the continuity of the number of times generated by the number-of-times-to-be-integrated number determination means 30 is usually counted. That is, the counter 41
Counts up when the determination by the integration number determination means 30 is true, outputs the carrier CR1 when the count overflows, sets the FF 44, and determines the tone signal Y1 from the in-phase output of the FF 44 in step S14. Similarly, the counter 42 counts up when the determination by the integration number determination means 30 is false, resets the FF 44 when the count overflows, and determines in step S16 and S17 that the audio signal Y2 is output from the inverted phase output of the FF 44. .

When the identification processing in one block is completed in this way, the identification processing of the next block is performed. As described above, in the present embodiment, the number of times that the value of the autocorrelation coefficient R (t) is larger than the positive threshold value PK1 is counted, and the number of times that the value of the autocorrelation coefficient R (t) is smaller than the negative threshold value PK2 is counted. Since the integration number determination means 30 determines whether the positive and negative count values exceed the threshold value, the tone signal Y1 and the audio signal Y2 can be clearly detected. Moreover, the tone signal Y1 and the audio signal Y2 can be identified in a short time, for example, 3 to 4 block times. Further, since the amount of calculation may be small, the circuit scale can be reduced.

FIG. 6 is a functional block diagram of a voice / musical tone discriminating circuit according to a second embodiment of the present invention, in which components common to those in FIG. 1 are denoted by common reference numerals.

In this voice / musical tone discriminating circuit, integrated frequency determining means 30A having a different configuration is provided in place of integrated frequency determining means 30 of FIG. The integrated number determining means 30A determines whether the sum of the integrated values M1 and M2 of the two positive and negative determined number integrating means 12 and 22 exceeds a threshold value K, and the output is true when M1 + M2> K. Thus, the same operation and effect as those of the first embodiment can be obtained.

The present invention is not limited to the above embodiment,
Various modifications are possible. For example, there are the following modifications. (A) The counting and identifying means 40 in FIGS. 1 and 6 may be configured using other circuit configurations. (B) The circuits in FIGS. 1 and 6 may be configured as individual circuits using an integrated circuit or the like, or may be digital signal
It may be executed by program control using a processor (DSP) or the like. When the program is executed under program control, the circuit scale can be reduced as compared with an individual circuit.

[0026]

As described in detail above, the first and second embodiments are described .
According to the second aspect, the autocorrelation coefficient of the input signal is obtained, the number of times the autocorrelation coefficient has a value larger than the positive threshold is integrated, and the number of times the value of the autocorrelation coefficient is smaller than the negative threshold is calculated. And the two integrated values each reach a predetermined threshold.
Whether it exceeds, or the sum of the two integrated values
It is determined whether or not a predetermined threshold is exceeded, and if the determination result is
A voice or a musical sound is identified by detecting whether or not the same determination result is obtained in the fixed section . Therefore, the audio signal and the tone signal are clearly distinguished, and the audio signal is improved in intelligibility and intelligibility by securing the quality of the band corresponding to the audio signal, and further, the tone signal is adapted to the tone signal. By handling in a wide band, high-quality and efficient identification processing becomes possible.

Further, since the identification processing can be performed in a short time and the amount of calculation is small, the circuit size can be reduced. Therefore, the present invention can be applied to various signal processing devices and the like that process voice signals and musical tone signals using digital signals.

[Brief description of the drawings]

FIG. 1 is a functional block diagram of a voice / music sound discriminating circuit showing a first embodiment of the present invention.

FIG. 2 is a time-series waveform diagram of an audio signal.

FIG. 3 is a time-series waveform diagram of a tone signal.

FIG. 4 is an operation flowchart of FIG.

FIG. 5 is a waveform diagram showing an autocorrelation coefficient of FIG.

FIG. 6 is a functional block diagram of a voice / music sound discriminating circuit according to a second embodiment of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 power calculating means 2 autocorrelation coefficient calculating means 10, 20 first and second threshold value judging means 11, 21 threshold value judging means 12, 22 judgment number accumulating means 30, 30 A accumulation number judging means 40 count identifying means

Continuation of the front page (58) Field surveyed (Int. Cl. ⁷ , DB name) G10L 11/00-21/06 JICST file (JOIS)

Claims (2)

(57) [Claims] 1. An autocorrelation coefficient calculating means for obtaining an autocorrelation coefficient of a digitized input signal, and a first threshold value judgment for integrating the number of times the autocorrelation coefficient has a value larger than a positive threshold value Integrating means; second threshold determination integrating means for integrating the number of times the autocorrelation coefficient has a value smaller than a negative threshold ; integrated value of the first threshold determining integrating means and the second threshold determination An integration count determining means for determining whether or not the integrated values of the integrating means each exceed a predetermined threshold; and a determination result of the integration number determining means being the same for a certain section.
Speech and tone discrimination circuit characterized by comprising a counting identification means for identifying an audio signal or tone signal by detecting whether the result. 2. A voice / musical tone discriminating circuit according to claim 1.
And wherein the number-of-times-of-integration determining means is the first threshold-value determining / integrating means.
Of the integrated value of the second threshold value and the integrated value of the second threshold determination integrating means
Is configured to determine whether or not exceeds a predetermined threshold
A voice / musical tone discriminating circuit characterized by the following.