JP2008028532A - Voice processor and voice processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To set the optimum amplification value of an analog voice signal and the optimum amplification value of a digital voice signal that make a voice signal to be inputted become desired sound volume in a voice processor which performs the amplification of the analog voice signals and that of the digital voice signals. <P>SOLUTION: When setting an analog amplification value and a digital amplification value respectively with respect to a target amplification value, the analog amplification value and digital amplification value are set so that an amplification value obtained by combining analog amplification with digital amplification is equal to the target amplification value and the analog amplification value becomes the maximum within a settable range. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a speech processing apparatus such as speech recognition that processes speech input by a user.

  Various audio processing devices perform processing for adjusting the volume of input audio. For example, some voice recorders used when taking notes during a conference or voice automatically adjust the volume to keep the recording level constant. Some voice recognition software that is commercially available has a function of adjusting the input volume for the purpose of improving the accuracy of voice detection and voice recognition.

In such volume adjustment, the volume of the input voice is adjusted so that the volume of the input voice becomes a predetermined volume. For example, in the speech recognition device disclosed in Patent Document 1, the gain of a microphone amplifier is changed according to the noise level. In the speech recognition apparatus disclosed in Patent Document 2, the amplification of the speech signal is adjusted according to the speech recognition rate. In the input level adjustment circuit and the voice communication terminal device disclosed in Patent Document 3, the amplification of the voice signal is adjusted according to the average input level of the voice section.
JP-A-6-67689 JP-A-6-337697 JP 2001-69200 A

  In the conventional technique, the input volume is adjusted by a single amplifier. However, when the audio processing apparatus includes an amplifier that amplifies the analog audio signal and an amplifier that amplifies the digital audio signal after A / D conversion, it discloses how to distribute the amplification value in each amplifier. There is nothing.

  When the digital audio signal is amplified, the quantization error at the time of A / D conversion is also amplified at the same time. Therefore, when both analog and digital amplification can be used, it is better to preferentially use the amplification by the analog audio signal.

  The present invention has been made in view of such circumstances. In an audio processing apparatus including an amplifier that amplifies an analog audio signal and an amplifier that amplifies a digital audio signal, quantization error during A / D conversion is reduced. It is an object to determine an amplification value of an analog audio signal and an amplification value of a digital audio signal so as to be reduced.

  In order to solve the above problems, an audio processing apparatus according to the present invention includes an analog amplifying unit that amplifies an analog audio signal, an A / D conversion unit that converts the analog audio signal into a digital audio signal, and amplifies the digital audio signal. A sound processing apparatus comprising: a digital amplifying unit that performs analog amplification with respect to a target amplification value; and an amplification value setting unit that sets the digital amplification value. The amplification value setting unit combines analog amplification and digital amplification. The analog amplification value and the digital amplification value are set so that the obtained amplification value is equal to the target amplification value and the analog amplification value is maximized within a settable range.

  In order to solve the above problems, an audio processing apparatus according to the present invention includes an analog amplifying unit that amplifies an analog audio signal, an A / D conversion unit that converts the analog audio signal into a digital audio signal, and a digital audio signal. An audio processing apparatus having a digital amplification means for amplification, an analog amplification value for a target amplification value, and an amplification value setting means for setting the digital amplification value, and is set to one or both of the analog amplification value and the digital amplification value When the value that can be obtained is a discrete value, the amplification value setting means is configured to select an analog amplification value from a combination of an analog amplification value and a digital amplification value in which a difference between the obtained amplification value and the target amplification value is within a predetermined range. Is set to the maximum combination.

  According to the present invention, in an audio processing apparatus including an amplifier that amplifies an analog audio signal and an amplifier that amplifies a digital audio signal, the amplification value of the analog audio signal is reduced so that quantization error during A / D conversion is reduced. The amplification value of the digital audio signal can be set.

  Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.

  FIG. 1 is a functional block diagram showing a schematic configuration of a speech processing apparatus according to Embodiment 1 of the present invention. In FIG. 1, 101 is an analog amplification unit, 102 is an A / D conversion unit, 103 is a digital amplification unit, 104 is an audio signal storage unit, 105 is an audio input control unit, 106 is an input volume adjustment unit, and 107 is an amplification value storage. Part.

  The analog amplifying unit 101 analog-amplifies the input analog audio signal and outputs the analog audio signal. The A / D converter 102 converts the analog audio signal output from the analog amplifier 101 into a digital audio signal and outputs the digital audio signal. The digital amplification unit 103 digitally amplifies the digital audio signal output from the A / D conversion unit 102 and outputs the digital audio signal. The audio signal storage unit 104 temporarily stores the digital audio signal output from the digital amplification unit 103. The audio input control unit 105 controls operations of the analog amplification unit 101, the A / D conversion unit 102, the digital amplification unit 103, and the audio signal storage unit 104. The audio input control unit 105 controls an amplification value (analog amplification value) when the analog amplification unit 101 amplifies the signal, and an amplification value (digital amplification value) when the digital amplification unit 103 amplifies the signal. . The audio input control unit 105 acquires a digital audio signal stored in the audio signal storage unit 104. The input volume adjustment unit 106 determines the analog amplification value and the digital amplification value so that the acquired audio signal (= digital audio signal stored in the audio signal storage unit 104) has a desired volume. A method for determining an analog amplification value and a digital amplification value (hereinafter, the two are collectively referred to as an amplification value) will be described in detail later. The amplification value storage unit 107 stores the amplification value determined by the input sound volume adjustment unit 106.

  FIG. 2 is a diagram illustrating a state in which an audio signal input in this apparatus is amplified through an analog amplification unit and a digital amplification unit. In FIG. 2, when an analog audio signal having a volume Go (201) is input to the audio processing apparatus, the analog amplification unit 101 amplifies the volume to Go + Aa (Aa (202) is an analog amplification value). The digital audio signal A / D converted by the A / D converter 102 is further amplified by the digital amplifier 103, and the finally obtained volume G (204) becomes G = Go + Aa + Ad (Ad (203) is Digital amplification value). The total amplification value A (205) obtained by combining analog amplification and digital amplification is A = Aa + Ad. Note that the unit of the volume and the amplified value is expressed in decibels (dB). Hereinafter, dB is used as a unit of volume and amplification value.

A procedure for adjusting the input sound volume in the sound processing apparatus having the above configuration will be described.
In this apparatus, sample sound is acquired, and an amplification value is set so that the volume of the sample sound approaches a desired volume. In this embodiment, the case where the maximum volume of the sample sound is brought close to the desired volume will be described. However, the procedure described below can be applied even when the volume calculation method such as the average volume or the minimum volume is changed. When the volume of the sample sound before amplification is Go and the desired volume is G ′, a target amplification value A ′ that is a target amplification value is obtained by the following equation.
A ′ = G′−Go (Formula 1)

Here, the volume Go before amplification of the sample sound cannot be directly obtained by this apparatus, but can be calculated by the following equation using the acquired volume G of the sample sound.
Go = GA (Formula 2)
In Equation 2, A is an amplification value used when obtaining sample audio.
Substituting Go of Equation 2 into Equation 1 yields the amplification value A ′ to be obtained by the following equation.
A ′ = G ′ − (GA)
= (G'-G) + A
= (G'-G) + (Aa + Ad) (Formula 3)
In Formula 3, Aa and Ad are an analog amplification value and a digital amplification value used when acquiring sample audio, respectively.

Next, the amplification value A ′ obtained by Expression 3 is distributed to the analog amplification value Aa ′ and the digital amplification value Ad ′.
A ′ = Aa ′ + Ad ′ (Formula 4)
In digital amplification, the quantization error during A / D conversion is also amplified, so the digital amplification value should be as small as possible. Also, in order to minimize the quantization error during A / D conversion, the analog amplification value should be as large as possible. Therefore, in this apparatus, in order to reduce the quantization error during A / D conversion, the amplification values are distributed so that the analog amplification value is as large as possible and the digital amplification value is as small as possible. When the maximum value of the digital amplification value is Ad_max, the minimum value is Ad_min, the maximum value of the analog amplification value is Aa_max, and the minimum value is Aa_min, the analog amplification value and the digital amplification value are obtained by the following equations.
(Case 1) When A ′ ≧ Aa_max + Ad_max Aa ′ = Aa_max (Formula 5)
Ad ′ = Ad_max (Formula 6)
(Case 2) When Aa_max + Ad_max> A ′ ≧ Aa_max + Ad_min Aa ′ = Aa_max (Formula 7)
Ad ′ = A′−Aa_max (Formula 8)
(Case 3) When Aa_max + Ad_min> A ′> Aa_min + Ad_min Aa ′ = A′−Ad_min (Formula 9)
Ad ′ = Ad_min (Formula 10)
(Case 4) When Aa_min + Ad_min ≧ A ′ Aa ′ = Aa_min (Formula 11)
Ad ′ = Ad_min (Formula 12)
By setting the amplification value in this way, the input sound volume can be adjusted to minimize the quantization error and obtain the desired sound volume G ′.

If Aa_max2 obtained by the following equation is used instead of Aa_max, the maximum analog amplification value that avoids clipping during A / D conversion can be set.
Aa_max2 = min (Aa_max, G_max−Go)
= Min (Aa_max, G_max− (GA))
(Formula 13)
In Expression 13, G_max is a maximum value of the volume that can be taken by the digital sound obtained after A / D conversion.

  The flow of the above input volume adjustment processing will be described with reference to the flowchart of FIG.

  When the input volume adjustment process is started, sample audio is first acquired. The input volume adjustment unit 106 sends an instruction to the audio input control unit 105 to acquire sample audio. Upon receiving the instruction, the audio input control unit 105 acquires the digital audio signal stored in the audio signal storage unit 104 via the analog control unit 101, the A / D conversion unit 102, and the digital amplification unit 103 (S101). Next, the volume of the digital audio signal acquired by the audio input control unit 105 is calculated by the input volume adjustment unit 106 (S102). Subsequently, the input volume adjustment unit 105 calculates an amplification value A ′ necessary for obtaining a desired volume (calculation of Formula 3, S103), and increases the analog amplification value according to the procedure described above. An analog amplification value and a digital amplification value are calculated (S104). The procedure for determining the analog amplification value and the digital amplification value is shown in the flowchart of FIG.

  In FIG. 4, A ′ is an amplification value calculated to obtain a desired sound volume calculated in S <b> 103.

  When the maximum amplification value in this apparatus, that is, the amplification value obtained by maximizing both analog amplification and digital amplification is less than the necessary amplification value A ′ or equal to A ′ (the above-mentioned case 1, S201 YES), digital Both the amplified value and the analog amplified value are set to the maximum values (Equation 5, Equation 6, S202).

  In the case of S201NO and when A ′ can be obtained by adjusting the digital amplification even when the analog amplification value is maximized (Case 2, S203 YES), the analog amplification value is set to the maximum value, and the necessary amplification value A ′ and analog The difference between the amplification values is adjusted with the digital amplification value (Equation 7, Equation 8, S204).

  In the case of S203NO and when A 'is larger than the minimum amplification value in the present apparatus (Case 3), the amplification value of digital amplification is minimized and the difference is set to the analog amplification value (Equation 9, Equation 10, S206).

  In the case of S206 NO, that is, when the necessary amplification value A ′ is equal to or smaller than the minimum amplification value of the present apparatus (case 4), both analog amplification and digital amplification are set to the minimum value (Formula 11, Formula 12, and S207).

  As described above, when clipping at the time of A / D conversion is avoided, the maximum value Aa_max2 of the analog amplification value in consideration of clipping obtained by Expression 13 is used instead of the maximum value Aa_max of the analog amplification value. Further, when the input volume adjustment is adjusted based on the average volume or the minimum volume, it is necessary to calculate the maximum volume at the same time when obtaining those volumes in S103.

  Returning to the flowchart of FIG. 3, the input volume adjustment unit 106 stores the determined amplification value in the amplification value storage unit 107 and ends (S105).

  As described above, by determining the analog amplification value and the digital amplification value, the analog amplification can be increased as much as possible to obtain a desired sound volume, thereby minimizing the quantization error during A / D conversion. it can.

  Some commonly used amplifier circuits can set only a few predetermined discrete amplification values. This is particularly the case with analog amplifier circuits. In this embodiment, as in the first embodiment, in an audio processing device that uses two amplification units of analog amplification and digital amplification, the amplification values in the analog amplification unit 101 and the digital amplification unit 103 can only be set discretely. explain.

  The processing from obtaining the sample sound, calculating the volume and calculating the necessary amplification value A ′ is the same as in the first embodiment (S101 to S103). Hereinafter, the process of S104 for determining the distribution of the analog amplification value and the digital amplification value will be described.

In the present embodiment, since the settable amplification values are discrete, it may be impossible to set the same amplification value as the necessary amplification value. Therefore, the analog amplification value Aa ′ and the digital amplification value Ad ′ with the least error with respect to the necessary amplification value are determined. That is, Aa ′ and Ad ′ that minimize the value of the following expression are determined in the ranges of Aa_max ≧ Aa ′ ≧ Aa_min and Ad_max ≧ Ad ′ ≧ Ad_min.
E = | A ′ − (Aa ′ + Ad ′) | (Formula 14)
When setting so as to avoid clipping at the time of A / D conversion, Aa_max3 ≧ Aa ≧ Aa_min using the maximum value Aa_max3 among settable analog amplification values not exceeding Aa_max2 calculated by Expression 1. Aa ′ is determined within the range of Aa_max3 is expressed by the following equation.
Aa_max3 = max {Aa∋S (Aa_max2)} (Formula 15)
In Equation 15, S (Aa_max2) is a set of amplification values that are equal to or smaller than Aa_max2 among the analog amplification values that can be set by this apparatus.

  In case 1 described in the first embodiment, that is, when the necessary amplification value is equal to or larger than the maximum value of the settable amplification value, the analog amplification value and the digital amplification value are set to the maximum as in the first embodiment. In the case 4, that is, when the necessary amplification value is less than or equal to the minimum value of the settable amplification value, the analog amplification value and the digital amplification value are set to the minimum as in the first embodiment.

  In case 2 and case 3, a combination that minimizes the error in the equation 14 is selected from the combination of the analog amplification value and the digital amplification value that satisfy the equation 14. When there are a plurality of combinations that minimize the error, a combination with a large analog amplification value is selected. A specific example will be described in which a combination with a small digital amplification value is selected when the analog amplification value is the same.

  As an example in which only discrete amplification values can be set, a case where analog amplification can be set every 10 dB in the range of −40 dB to 40 dB and digital amplification can be set every 2 dB in the range of −20 dB to 20 dB will be described.

  When the necessary amplification value A ′ is 46.3 dB, the analog amplification value is set to 40 dB and the digital amplification value is set to 6 dB. At this time, the error E that can be expressed by Equation 14 is 0.3 dB, and no other combination has the error E smaller than this, so this combination is selected.

  When the necessary amplification value is 45.0 dB, there are the following four combinations that minimize the error.

  These combinations of candidates are narrowed down to two combinations (1) and (2) with large analog amplification values. (1) and (2) are compared, and (2) having a small digital amplification value is determined as a final combination of amplification values. By doing so, it is possible to determine an amplification value that minimizes the quantization error during A / D conversion and reduces the amplification of the quantization error during digital amplification.

  The procedure for determining the analog amplification value and the digital amplification value in S104 of the present embodiment described above is shown in the flowchart of FIG. In the flowchart of FIG. 5, S201, S202, S205, and S207 are the same as those in the first embodiment.

  In FIG. 5, A ′ is an amplification value necessary for obtaining the desired volume calculated in S <b> 103. When the maximum amplification value in this apparatus, that is, the amplification value obtained by maximizing both analog amplification and digital amplification is less than the necessary amplification value A ′ or equal to A ′ (the above-mentioned case 1, S201 YES), digital Both the amplified value and the analog amplified value are set to the maximum values (Equation 5, Equation 6, S202).

  In the case of S201NO and S205NO, that is, when the necessary amplification value A 'is less than or equal to the minimum amplification value of the present apparatus (case 4), both analog amplification and digital amplification are set to the minimum values (Formula 11, Formula 12, and S207).

  On the other hand, in the case of S201NO and when A ′ is larger than the minimum amplification value in the present apparatus (Case 2 and Case 3), the combination of the analog amplification value and the digital amplification value that minimizes the error E expressed by Equation 14 is obtained. Calculate (S208).

  If there is only one combination that results in the minimum error, an amplification value is determined based on the calculated combination (NO in S209). On the other hand, when there are a plurality of combination candidates having the smallest error (YES in S209), the combination having the largest analog amplification value is selected from these candidates (S210).

  When the candidate selected in S210 is one set, the amplification value is determined by the selected combination (NO in S210). If there are a plurality of candidates selected in S210 (YES in S211), the combination having the smallest digital amplification value is selected from these candidates and the amplification value is determined (S212).

  In addition, when avoiding clipping at the time of A / D conversion as in the first embodiment, Aa_max3 obtained by Expression 15 is used instead of the maximum analog amplification value Aa_max.

  Returning to the flowchart of FIG. 3, the input volume adjustment unit 106 stores the determined amplification value in the amplification value storage unit 107 and ends (S105).

  As described above, by determining the analog amplification value and the digital amplification value, the quantization error can be reduced when the numerical values that can be set for the analog amplification value and the digital amplification value are discrete.

  In Example 1 and Example 2, sample sound was acquired and the amplification value was determined from the volume of the acquired sample sound. However, in this apparatus, there is no means for confirming whether or not the acquired sample sound includes sound. For this reason, there is a problem that the amplified value cannot be calculated correctly if an unexpected input is made, such as when the user does not utter or when sudden internal noise is mixed at the time of sample sound acquisition. In this embodiment, in order to cope with this problem, an example will be described in which a means for confirming whether or not an appropriate volume can be acquired with a set amplification value and a means for setting a predetermined standard amplification value are provided. To do.

  In the present embodiment, the operation of the voice recognition device incorporating the input sound volume device described in the first and second embodiments will be described as an example. In the present embodiment, for the sake of explanation, a simple voice recognition apparatus that executes commands by voice input will be described as an example. However, the present invention is not limited to this, and the present invention is applicable to all apparatuses and systems that involve voice input. Is possible.

  FIG. 6 is a block diagram showing a schematic configuration of the speech recognition apparatus according to the present embodiment. Reference numerals 101 to 107 are the same as those of the sound processing apparatus described in the first and second embodiments, and thus description thereof is omitted.

  In FIG. 6, reference numeral 108 denotes a voice recognition control unit which controls operations of a voice detection unit 109 and a voice recognition unit 110 described later. In addition, the audio signal stored in the audio signal storage unit 104 is acquired via the audio input control unit 105, and the audio signal is passed to the audio detection unit 109. The voice detection unit 109 detects a voice section uttered by the user from a voice signal obtained via the voice recognition control unit 109. The voice recognition 110 performs an acoustic analysis of the voice signal in the section detected by the voice detection unit 109, performs a decoding process using necessary data such as an acoustic model and a language model (not shown), and obtains a voice recognition result. The result obtained by the voice recognition unit 110 is referred to via the voice recognition control unit 108. The voice output unit 111 controls guidance output to the user by voice including talkback of the recognition result. The voice guidance to the user may be realized by using voice synthesis (not shown). The display control unit 112 controls display such as GUI used in the apparatus and display of recognition results. The operation control unit 113 controls operation means provided to the user, such as a mouse, a kibo board, and a button. The control unit 114 controls the operation of the entire apparatus.

  The operation of the speech recognition apparatus having the above configuration will be described with reference to the flowchart of FIG. When this apparatus is activated, the amplification value is set when the activation process is performed by the control unit 114. The control unit 114 reads the analog amplification value and the digital amplification value stored in the amplification value storage unit 107, and sets the amplification value via the voice input control unit 105 (S301). In this apparatus, the amplification value once set is stored in the amplification value storage unit 107, and the same amplification value is applied at the next startup even when the apparatus is terminated. The amplification value stored in the amplification value 107 is held in the amplification value storage unit 107 unless the amplification value is updated by input volume adjustment described later.

  When the activation process ends, the display control unit 112 displays as shown in FIG. 8 and waits until the user performs an operation (NO in S302). FIG. 8 shows an example of a display in a standby state, in which a guide 302, an utterance button 303, an input volume adjustment button 304, and an end button 305 are displayed in a window 301. The guidance 302 displays contents to be notified to the user such as an operation guide. Here, an operation guide for command input is displayed.

  When the user operates the displayed button, the operation control unit 113 detects a user operation (YES in S302), and the control unit 114 controls processing according to the operation content (S303 to S310). In FIG. 8, when the user presses the utterance button 303 (S303 NO, S305 YES), voice recognition is executed (S309), and processing corresponding to the obtained recognition result is performed (S309). Similarly, when the user presses the input volume adjustment button 304, the input volume adjustment is executed (S307). Details of the input volume adjustment in S307 will be described later.

  When the end button 305 is pressed (S302 YES, S303 NO, S305 NO, S306 YES), the execution of the speech recognition apparatus is ended. When providing an operation other than the displayed button (for example, an operation using a keyboard), the control unit 114 performs a corresponding process when the operation is detected (S302 YES, S303 NO, S305 NO, S306 NO) (S310). .

  Next, the detailed operation of the input volume adjustment in S307 in this apparatus will be described with reference to FIGS. FIG. 9 is a flowchart showing the flow of input volume adjustment processing in this apparatus. 10 to 13 are examples of displays controlled by the display control unit 112 during the input volume adjustment process.

  When the input volume adjustment is started (S307 in FIG. 7), the display control unit 112 switches the display to an input volume adjustment UI (user interface) and waits until the user operates (NO in S401).

  FIG. 10 shows an example of the input volume adjustment UI. A start button 306, a standard value setting button 307, and a cancel button 308 are displayed in the window 301. The guidance 302 displays an input volume adjustment operation method.

  Returning to FIG. 9, when the operation control unit 113 detects a user's operation (YES in S401), the control unit 114 executes processing according to the user's operation (S402 to S419).

  10, when the user presses the start button 306 (YES in S401, S402 YES), the input volume adjustment using the sample voice is started in S403 and thereafter. On the other hand, when the standard value setting button 307 is pressed (S401 YES, S402 NO, S417 YES), the input volume adjustment unit 106 reads the standard value of the amplification value stored in the amplification value storage unit 107 and passes through the voice input control unit 105. The read amplification value is set (S418). Subsequently, the set amplification value is stored in the amplification value storage unit (S419), and the input volume adjustment is terminated.

  In FIG. 10, when the user presses the cancel button 308 (S401 YES, S402 NO, 417 NO), the input volume adjustment is immediately finished.

  Next, the processing after S403 will be described. First, in step S <b> 403, the input volume adjustment unit 107 sets an amplification value for obtaining sample audio via the audio input control unit 105. When acquiring sample audio, a small amplification value is used so as not to clip during A / D conversion. Amplification value for sample audio acquisition is not clipped even when the user's voice is loud, and an appropriate value can be obtained experimentally in advance so that an audio signal can be acquired even when the user's voice is low, and the amplification value is stored. Stored in the unit 107.

  Next, sample audio used for input volume adjustment is acquired. The input volume adjustment unit 106 sends an instruction to the audio input control unit 105 to acquire sample audio. Upon receiving the instruction, the audio input control unit 105 acquires the digital audio signal stored in the audio signal storage unit 104 via the analog control unit 101, the A / D conversion unit 102, and the digital amplification unit 103 (S404).

  Next, the input volume adjusting unit 106 calculates the volume of the digital audio signal acquired by the audio input control unit 105 (S405).

  Subsequently, the input volume adjusting unit 105 calculates an amplification value A ′ necessary to obtain a desired volume (S406), and calculates an analog amplification value and a digital amplification value (S407).

  The procedure described in the first or second embodiment is used for the calculation of the amplification value A ′ in S406 and the digital amplification value and the analog calculation in S407.

  In step S <b> 408, the input volume adjustment unit 105 sets the analog amplification value and the digital amplification value calculated in step S <b> 407 via the voice input control unit 105.

  With the above processing, the digital amplification value and the analog amplification value are changed by any one of the methods of the first and second embodiments.

  Hereinafter, processing (S409 to S416) for confirming whether or not an appropriate sound volume can be acquired with the set amplification value will be described.

  When the setting of the amplification value is completed (S408), the display control unit 112 switches to a display for prompting the user to speak in order to obtain confirmation sound, and waits until the user operates (NO in S409).

  FIG. 11 shows an example of a display for prompting a utterance for confirmation. A start button 309 and a cancel button 310 are displayed in the window 301. Further, the guidance 302 displays contents for prompting confirmation.

  Returning to FIG. 9, when the operation control unit 113 detects a user operation (YES in S409), the control unit 114 executes a process according to the user operation (S410 to S416).

  In FIG. 11, when the user presses the start button 309 (S409 YES, S410 YES), an instruction is sent from the input volume adjustment unit 106 to the voice input control unit 105 and the confirmation voice is acquired in the same manner as the sample voice acquisition in S404. (S411).

  Subsequently, the volume G ″ of the confirmation sound acquired in S411 is calculated, and it is determined whether or not the volume G ″ calculated in S413 is appropriate.

  Here, the minimum value Mmin and the maximum value Mmax of the allowable range with respect to the difference from the desired sound volume G ′ are determined in advance, and it is determined that Mmin <G ″ −G ′ <Mmax is appropriate. The determination criterion is not limited to this, and a simple method such as a difference between G ″ and G ′ within 5 dB may be used. In addition to these determinations, it may be determined that the confirmation sound is not appropriate when clipping.

  If the volume of the confirmation sound is appropriate (YES in S413), the set amplification value is stored in the amplification value storage unit 107, and the input volume adjustment is terminated (S414).

  As described above, the amplification value calculated in S407 is stored in the amplification value storage unit 107 after confirming that the sound volume obtained by setting the amplification value is appropriate. The original amplification value before starting the input volume adjustment is stored in the amplification value storage unit 107. Accordingly, when the user presses the cancel button 310 in FIG. 11 and cancels the input volume adjustment halfway (NO in S410), or when the volume of the confirmation sound described later is not appropriate, the stored amplification value is read out again. By setting, it is possible to return to the amplified value before the input volume adjustment.

  If the volume of the confirmation voice is not appropriate in S413 (NO), the user is notified of the failure to adjust the input volume and the expected cause (S415). FIG. 12 is an example of a display for notifying the input volume adjustment failure and its cause by the display control unit 112.

  Following the cause notification in S415, the amplification value set for performing the input volume adjustment is returned to the original value. The input volume adjustment unit 106 reads the amplification value stored in the amplification value storage unit 107, sets the read amplification value via the voice input control unit (S416), and ends the input volume adjustment processing.

  As described above, according to the speech recognition apparatus of the third embodiment, when acquiring sample speech and setting an amplification value, a confirmation speech is acquired using the calculated amplification value, and the set amplification value is Judging whether it is appropriate. For this reason, it is possible to reduce the possibility that an inappropriate amplification value is set due to an unexpected input at the time of sample sound acquisition.

(Other examples)
In the above description, the content of the guidance 302 displayed in the window 301 described with reference to FIGS. 8, 10, 11, and 12 can be guided by the synthesized output using speech synthesis (not shown) by the speech output control unit 111. Good.

  The object of the present invention can also be achieved as follows. That is, a storage medium in which a program code of software that realizes the functions of the above-described embodiments is recorded is supplied to the system or apparatus. Then, the computer (or CPU or MPU) of the system or apparatus reads and executes the program code stored in the storage medium. It goes without saying that the purpose is achieved even in this way.

  In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the program code constitutes the present invention.

  As a storage medium for supplying the program code, for example, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

  Further, the embodiments according to the present invention are not limited to the case where the functions of the above-described embodiments are realized by executing the program code read by the computer. For example, an OS (operating system) running on a computer performs part or all of actual processing based on an instruction of the program code, and the functions of the above-described embodiments may be realized by the processing. Needless to say, it is included.

  Furthermore, the functions of the embodiment according to the present invention are also realized as follows. That is, the program code read from the storage medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer. Then, based on the instruction of the program code, the CPU provided in the function expansion board or function expansion unit performs part or all of the actual processing. It goes without saying that the functions of the above-described embodiments are realized by this processing.

It is a figure which shows schematic structure of the audio | voice processing apparatus which concerns on Example 1 and Example 2 of this invention. It is a figure explaining a mode that the audio | voice signal input with the speech processing apparatus which concerns on Example 1 and Example 2 of this invention, and the speech recognition apparatus which concerns on Example 3 is amplified through an analog amplification part and a digital amplification part. . It is a flowchart showing the operation | movement procedure of the audio | voice processing apparatus which concerns on Example 1 and Example 2 of this invention. 4 is a flowchart illustrating a procedure for calculating an analog amplification value and a digital amplification value in the sound processing apparatus according to Embodiment 1 of the present invention. It is a flowchart which shows the procedure which calculates an analog amplification value and a digital amplification value in the audio processing apparatus which concerns on Example 2 of this invention. It is a figure which shows schematic structure of the speech recognition apparatus which concerns on Example 2 of this invention. It is a flowchart which shows an example of the operation | movement procedure of the speech recognition apparatus which concerns on Example 3 of this invention. It is a figure explaining an example of operation | movement of the speech recognition apparatus which concerns on Example 3 of this invention. It is a flowchart which shows the operation | movement procedure of the speech recognition apparatus which concerns on Example 3 of this invention. It is a figure explaining an example of operation | movement at the time of performing input volume adjustment in the speech recognition apparatus which concerns on Example 3 of this invention. It is a figure explaining an example of operation | movement at the time of performing input volume adjustment in the speech recognition apparatus which concerns on Example 3 of this invention. It is a figure explaining an example of operation | movement at the time of performing input volume adjustment in the speech recognition apparatus which concerns on Example 3 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 Analog amplification part 102 A / D conversion part 103 Digital amplification part 104 Audio | voice signal storage part 105 Audio | voice input control part 106 Input volume adjustment part 107 Amplification value storage part

Claims (19)

Analog amplification means for amplifying an analog audio signal;
A / D conversion means for converting an analog audio signal into a digital audio signal;
An audio processing apparatus comprising digital amplification means for amplifying a digital audio signal, and an amplification value setting means for setting an analog amplification value and a digital amplification value for a target amplification value,
The amplification value setting means sets the analog amplification value and the digital amplification value so that the amplification value obtained by combining analog amplification and digital amplification is equal to the target amplification value and becomes the maximum within the settable range of the analog amplification value. An audio processing apparatus characterized by setting. Analog amplification means for amplifying an analog audio signal;
A / D conversion means for converting an analog audio signal into a digital audio signal;
An audio processing apparatus comprising digital amplification means for amplifying a digital audio signal, and an amplification value setting means for setting an analog amplification value and a digital amplification value for a target amplification value,
If the value that can be set to either one or both of the analog amplification value and the digital amplification value is a discrete value,
The amplification value setting means sets a combination having the maximum analog amplification value among a combination of an analog amplification value and a digital amplification value in which a difference between an obtained amplification value and a target amplification value falls within a predetermined range. A voice processing device.   3. The audio processing apparatus according to claim 2, wherein the amplification value setting means sets a combination that minimizes the amplification value of the digital audio signal when there are a plurality of combinations having the maximum amplification value of the analog signal. .   4. The audio processing apparatus according to claim 1, wherein the amplification value setting unit sets an amplification value of an analog signal so as not to be clipped during A / D conversion. An audio acquisition means for acquiring an audio signal;
Volume calculation means for calculating the volume of the acquired audio signal;
A target amplification value calculating means for calculating a target amplification value for setting the volume of the acquired audio signal to a predetermined volume;
The voice acquisition means acquires a first voice signal;
The volume calculation means calculates the volume of the acquired first audio signal,
The target amplification value calculation means calculates a target amplification value using a predetermined volume, an analog amplification value when the first audio signal is acquired, and a digital amplification value when the first audio signal is acquired,
5. The audio processing apparatus according to claim 1, wherein the amplification value setting unit sets an analog amplification value and a digital amplification value for the target amplification value calculated by the target amplification value calculation unit. .   When the first audio signal is acquired by the audio acquisition unit, audio is acquired using a predetermined analog amplification value and digital amplification value that are predetermined so that the acquired audio signal is not clipped. The speech processing apparatus according to claim 5. Acquiring a second audio signal by the audio acquisition means;
The volume of the second audio signal is calculated by the volume calculation means,
A check means for determining whether or not the calculated volume of the second audio signal is appropriate;
The sound processing apparatus according to claim 5, wherein the confirmation unit determines that the sound volume value of the second sound signal is appropriate when the sound value is within a predetermined range.   8. The user according to claim 7, wherein when the volume of the second audio signal is not determined to be appropriate by the confirmation means, the user is notified of the reason that the volume of the second audio signal is not appropriate. Voice processing device.   9. The speech processing apparatus according to claim 1, further comprising standard value setting means for setting the analog amplification value and the digital amplification value to a predetermined standard value. Analog amplification means for amplifying an analog audio signal;
A / D conversion means for converting an analog audio signal into a digital audio signal;
An audio processing method in an audio processing apparatus comprising digital amplification means for amplifying a digital audio signal,
An amplification value setting step for setting an analog amplification value and a digital amplification value for a target amplification value is provided.
In the amplification value setting step, the analog amplification value and the digital amplification value are set so that the amplification value obtained by combining analog amplification and digital amplification is equal to the target amplification value and the analog amplification value is maximized within a settable range. An audio processing method comprising: setting. Analog amplification means for amplifying an analog audio signal;
A / D conversion means for converting an analog audio signal into a digital audio signal;
An audio processing method in an audio processing apparatus comprising digital amplification means for amplifying a digital audio signal,
An amplification value setting step for setting an analog amplification value and a digital amplification value for a target amplification value
If the value that can be set to either one or both of the analog amplification value and the digital amplification value is a discrete value,
In the amplification value setting step, the combination of the analog amplification value and the digital amplification value that sets the difference between the obtained amplification value and the target amplification value within a predetermined range is set to the maximum analog amplification value. Voice processing method.   12. The audio processing method according to claim 11, wherein the amplification value setting step sets a combination that minimizes the amplification value of the digital audio signal when there are a plurality of combinations having the maximum amplification value of the analog signal. .   13. The audio processing method according to claim 10, wherein the amplification value setting step sets an amplification value of an analog signal so as not to be clipped at the time of A / D conversion. A voice acquisition step of acquiring a voice signal;
A volume calculation step for calculating the volume of the acquired audio signal;
A target amplification value calculating step of calculating a target amplification value for setting the volume of the acquired audio signal to a predetermined volume;
The voice acquisition step acquires a first voice signal,
The volume calculation step calculates the volume of the acquired first audio signal,
The target amplification value calculation step calculates a target amplification value using a predetermined volume, an analog amplification value when the first audio signal is acquired, and a digital amplification value when the first audio signal is acquired,
The audio processing method according to claim 10, wherein the amplification value setting step sets an analog amplification value and a digital amplification value for the target amplification value calculated by the target amplification value calculation step. .   When acquiring the first audio signal in the audio acquisition step, the audio is acquired using a predetermined analog amplification value and digital amplification value that are predetermined so that the acquired audio signal is not clipped. The voice processing method according to claim 14. A second audio signal is acquired by the audio acquisition step,
The volume of the second audio signal is calculated by the volume calculation step,
And further comprising a confirmation step of determining whether or not the calculated volume of the second audio signal is appropriate,
16. The sound processing method according to claim 14, wherein the confirmation step determines that the sound volume is appropriate when the volume value of the second sound signal is within a predetermined range.   17. The user is notified of the reason why the volume of the second audio signal is not appropriate when the volume of the second audio signal is not determined to be appropriate in the confirmation step. Voice processing method.   18. The voice processing method according to claim 10, further comprising a standard value setting step of setting the analog amplification value and the digital amplification value to a predetermined standard value. A program for causing a computer to implement the voice processing method according to any one of claims 10 to 18.

Images