JP2599974B2 - Voice detection method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a voice detection system.

[Prior Art] Conventionally, there are a number of methods for detecting the presence of speech in a noisy environment, such as detecting speech sections in communication as described in Japanese Patent Publication No. 57-12999, or using speech language contents. Although it is used for pre-processing of recognition, it is difficult to develop it for general use under high noise.For example, starting a response by voice of a hands-free phone while the ringing tone is ringing, etc. could not.

As a method of simply detecting the presence of voice in a noise environment, there has been a method of detecting the number of times an input signal crosses a reference axis within a predetermined time interval.

[Problems to be Solved by the Invention] However, in the above-mentioned method using the conventional voice detection method, it is generally assumed that the noise amplitude is smaller than the voice amplitude, and the noise amplitude is lower than the voice amplitude. If the amplitude is almost the same, the presence of voice cannot be detected.

SUMMARY OF THE INVENTION It is an object of the present invention to easily detect the presence of speech in a noise environment even when the amplitude of noise is large and the influence on speech detection is large.

[Means for Solving the Problems] According to the present invention as set forth in claims 1 and 2, the number of reference axis crossings of an input signal and a value relating to the amplitude distribution of a waveform are calculated as characteristic parameters, and this calculation result is used as a voiced sound. Is compared with the dictionary data on the specific noise to determine whether or not the input signal includes a voiced sound.

According to the first aspect of the present invention, for example, a peak value P represented by the following equation is used as a value relating to the amplitude distribution of the waveform.

P = 20 × 1 og ₁₀ (V _P / V _rms ) where V _P : the maximum value of the absolute value of the amplitude within a fixed time interval V _rms : the effective value of the amplitude within the same fixed time interval According to the present invention, as a value relating to the amplitude distribution of the waveform, for example, a peak value P represented by the following equation is used.
Is used.

_{P = 20 × 1og 10 (V} P / V a) However, V _P: maximum V _a of the absolute value of the amplitude within a certain time interval: average value of the absolute value of the amplitude in the same fixed time interval [Operation] claims In the present invention described in the items 1 and 2, a voice in a noise environment is detected as follows. In the present invention, voiced sounds (voices with vocal folds such as vowels, semi-vowels, and nasal sounds, and almost all voices uttered by humans include voiced sounds) are regarded as voices. .

(1) With respect to voiced sound and specific noise, the number of reference axis crossings (the number of times the signal crosses a predetermined reference level such as a zero level) and the value related to the amplitude distribution of the waveform within a certain time interval are used as characteristic parameters. Prepare dictionary data.

Examples of the dictionary data include the following (a), (b),
(C) is used.

(A) A set of feature parameters for voiced sounds obtained from multiple voices.

(B) A set of a large number of characteristic parameters obtained for a specific noise (for example, a ring tone of a specific telephone).

(C) A set of feature parameters obtained by adding a result of adding a voiced sound and a specific noise at a specific ratio for many voices.

The data (a), (b) and (c) are based on numerical data obtained by converting acoustic data into feature parameters, average values obtained by statistically processing numerical data, statistical data such as variance, or statistical data. It can be prepared in various modes such as discriminant data such as a boundary equation determined in advance.

(2) The input signal is sampled, and the number of reference axis crossings and the value related to the amplitude distribution of the waveform within a certain time interval of the input signal are calculated as characteristic parameters.

(3) The feature parameters calculated in (2) are compared with a standard pattern defined by the dictionary data defined in (1) in a parameter space, and pattern recognition is performed to determine whether or not the input signal includes a voiced sound. Determined by

The above-mentioned pattern is recognized using the dictionary data, for example, as follows.

The category "voiced sound" defined by the dictionary data (the voiced sound of (a) or the voiced sound category of the specific noise of (c) added at a specific ratio) and the category "others" are divided into two minutes. And determines which category the feature parameter of the input signal belongs to.

In addition, considering the possibility that the amplitude of the specific noise is large, which greatly affects the detection of voiced sound, in addition to the above, the boundary between the category "specific noise" and the category "voiced sound" is determined, and the input signal It is determined which category the feature parameter belongs to.

As a result of the above determination, it is determined that a voiced sound exists in the input signal on condition that the input signal belongs to the category “voiced sound” and does not belong to the category “specific noise”.

Thus, in the present invention described in claims 1 and 2,
Since two parameters, the number of reference axis crossings and the value related to the amplitude distribution of the waveform, are used as the feature parameters, the category “voiced sound” and the category “specific noise” can be clearly separated in the parameter space. Therefore, even when the amplitude of the specific noise is large and the influence on the voice detection is large, the presence of the voice under the noise environment can be easily detected at a high detection rate.

According to the first aspect of the present invention, since the peak value as described above is used as the value related to the amplitude distribution of the waveform, the parameter value that faithfully reflects the sharp waveform which is a characteristic of voiced sound is used. As a result, there is a merit that noise discrimination is improved.

According to the second aspect of the present invention, since the peak value as described above is used as the value relating to the amplitude distribution of the waveform, the amount of calculation can be reduced as compared with the second aspect of the present invention. In addition, since a parameter value that reflects a sharp waveform that is a characteristic of voiced sound relatively faithfully is used, there is a merit that noise discrimination is improved. Note that a small amount of calculation means that the response speed is fast.

[Embodiment] FIG. 1 is a block diagram showing an example of a speech detection device used in the embodiment of the present invention, and FIG. 2 is a schematic diagram showing a parameter space formed by characteristic parameters of the present invention.

In FIG. 1, 11 is a microphone, 12 is an amplifier, 13 is a low-pass filter, 14 is an A / D converter, 15 is a parameter calculation unit, 16 is a dictionary data storage unit, 17 is a judgment unit, and 18 is a result output unit. . In this embodiment, speech in a noisy environment is detected as follows.

(1) For voiced sound and specific noise, 20mS of those signals
A reference axis intersecting the number X ₁ between, the value X ₂ relating to the amplitude distribution of the waveform to prepare the dictionary data, wherein parameters allowed to store it in the dictionary data storage unit 16.

Here, the value X ₂ relating to the amplitude distribution of the waveform can be used below, one of the.

The peak value P represented by the following equation.

P = 20 × 1og ₁₀ (V _P / V _rms ) where V _P : Maximum value of the absolute value of the amplitude within a fixed time interval V _rms : RMS value of the amplitude within the same fixed time interval P.

_{P = 20 × 1og 10 (V} P / V a) However, V _P: maximum V _a of the absolute value of the amplitude within a certain time interval: average above the peak value of the absolute value of the amplitude in the same fixed time interval Is used, a parameter value that reflects a sharp waveform that is a feature of voiced sound relatively faithfully is used, and there is an advantage that noise discrimination is improved.

When the above peak value is used, the amount of calculation can be reduced as compared to the above peak value, and a parameter value that faithfully reflects a sharp waveform that is a characteristic of voiced sound is used. There is a merit that is improved.

As the dictionary data, for example, the following (a):
(B) and (c) are created.

(A) A set of feature parameters for voiced sound [A] obtained from many voices.

(B) A set of a large number of characteristic parameters obtained for a specific noise (ringing sound of a specific telephone).

(C) The voiced sound [a] and the specific noise are defined as 20 × 1 og ₁₀ (S _rms / N _rms ) [dB], and the voiced sound-to-specific noise ratio is defined as 3,0, −3, −6, −10. [D
A set of feature parameters obtained by adding the results obtained in [B] for many voices. Note that S _rms represents the effective value of the amplitude of the voiced sound “A”, and N _rms represents the effective value of the amplitude of the specific noise.

(2) The input signal is sampled by the microphone 11, the input signal is amplified by the amplifier 12, and the component above 4.2 KHz is cut by passing through the low-pass filter 13, and the A / D converter 1
The signal is converted into a digital signal having a sampling frequency of 10 KHz and a conversion bit number of 16 bits by 4 and is sent to the parameter calculation unit 15.
Parameter calculating unit 15 calculates a reference axis intersecting the number X ₁ between 20mS of the input signal, and a value X ₂ relating to the amplitude distribution of the waveform as the feature parameter.

(3) The feature parameter calculated in (2) is compared with the standard pattern defined by the dictionary data defined in (1) in the determination unit 17 to determine whether the input signal includes a voiced sound. The result is output from the result output unit 18.

Here, the pattern recognition using the aforementioned dictionary data is as follows.
For example, the following is performed on the parameter space of FIG.

Incidentally, FIG. 2 are those taken in X ₁ 2 single feature parameters respectively axial and X ₂ axis between the peak value zero crossing number (obtained by setting the reference axis level to zero level). In FIG.
μ ₁ , σ ₁₁ , and σ ₁₂ are voiced sound dictionaries (the voiced sound [a] of (a) or the voiced sound obtained by adding the specific noise of (c) at a specific voiced sound to specific noise ratio). the average value of the parameter, the standard deviation of the X ₁ axis component represents the standard deviation of the X ₂ axis _{component, μ 2, σ 21, σ} 22 represents the same values for the dictionary parameter of each specific noise.

A category “voiced sound” defined by the dictionary data (a voiced sound [a] of the above (a) or a voiced sound category obtained by adding the specific noise of the above (c) at a specific ratio);
A boundary 1 that divides the category “other” into two is determined. In the boundary 1, the side including the average value mu ₁ of the dictionary data voiced is category "voiced". This boundary 1 is a concentrated ellipse indicating how much voiced dictionary data is concentrated around the average value. By changing the length of the axis, the ratio of the voiced dictionary data entering the ellipse is changed. be able to. In the case of this embodiment, the length of the axis is determined so that 90% of the voiced dictionary data falls within the ellipse. The broken line represents the concept of the category “voiced sound” defined by μ and σ. That is, in this process, it is determined which side of the boundary 1 the characteristic parameter of the input signal belongs to.

Next, in consideration of the possibility that the amplitude of the specific noise is large, which greatly affects the detection of voiced sound, in addition to the above, the category "specific noise" and the category "voiced sound"
The boundary 2 of is determined. In the boundary 2, is the side that contains the average value mu ₂ specific noise becomes category "specific noise". The boundary 2 is a group of points having the same likelihood for the category “voiced sound” and the category “specific noise”. In the case of this embodiment, the standard deviation of the specific noise is the ring tone of the artificially created telephone, and the dictionary data of the voiced sound and the specific noise added at the specific voiced sound to specific noise ratio. Is generally smaller than the standard deviation of, the category "specific noise" is a closed space. The broken line represents the concept of the category “specific noise” defined by μ and σ. That is, in this process, it is determined which side of the boundary 2 the characteristic parameter of the input signal belongs to.

Above, the results of the determination of the input signal, in feature parameter space, belonging to the mu ₁ side of the boundary 1 in, when not belonging to mu ₂ side of the boundary 2 in and belongs to the category "voiced" input signal Is determined. That is, it is determined that a voiced sound exists in the input signal.

In the above embodiment, since two parameters, ie, the number of reference axis crossings and the value related to the amplitude distribution of the waveform, are used as the characteristic parameters, the category “voiced sound” and the category “specific noise” are clearly defined in the parameter space. Can be separated. Therefore, even when the amplitude of the specific noise is large and the influence on the voice detection is large, the presence of the voice under the noise environment can be easily detected at a high detection rate. Especially,
In the above example, it was confirmed that the voiced sound-to-specific noise ratio exhibited a high voiced sound detection rate even at −6 dB, and a detection rate close to 100% at −3 dB.

In the above embodiment, a concentrated ellipse and a set of points at which the likelihoods of the two categories are equal are used as the boundary defining the standard pattern in the feature parameter space. General pattern recognition technique can be used. For example,
Instead of a collection of points where the likelihood for the category "voiced sound" and the category "specific noise" are equal, Maharanobi
A group of points where the s distance and the Euclid distance are equal can be used.

[Effects of the Invention] As described above, according to the present invention, even when the amplitude of noise is large and the influence on voice detection is large, the presence of voice in a noise environment can be easily detected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example of a voice detection device used for carrying out the present invention, and FIG. 2 is a schematic diagram showing a parameter space formed by characteristic parameters of the present invention. 11: microphone, 15: parameter calculation unit, 16: dictionary data storage unit, 17: determination unit, 18: result output unit

──────────────────────────────────────────────────の Continuation of the front page (51) Int.Cl. ⁶ Identification number Office reference number FI Technical display location G10L 9/18 301 G10L 9/18 301A (56) References JP-A-58-7196 (JP, A) JP-A-60-200300 (JP, A) JP-A-61-292199 (JP, A) JP-B-38-11568 (JP, B1) JP-T-63-500339 (JP, A)

Claims (2)

(57) [Claims] 1. The method according to claim 1, wherein the number of reference axis crossings of the input signal and a value relating to the amplitude distribution of the waveform are calculated as characteristic parameters, and the calculation result is compared with dictionary data for voiced sound and specific noise. And a value relating to the amplitude distribution of the waveform, expressed as a ratio of the effective value of the amplitude within the fixed time interval to the maximum value of the absolute value of the amplitude within the fixed time interval. A voice detection method using peak values. 2. The method according to claim 1, wherein the number of reference axis crossings of the input signal and a value relating to the amplitude distribution of the waveform are calculated as characteristic parameters, and the calculation result is compared with dictionary data of voiced sound and specific noise. A ratio of the average value of the absolute value of the amplitude within a certain time interval to the maximum value of the absolute value of the amplitude within a certain time interval as a value related to the amplitude distribution of the waveform. A voice detection method using the peak value represented by.