JP6238181B1 - Loudspeaker and control method thereof - Google Patents

Abstract

To enable a user to appropriately switch and output an original voice recorded from a voice spoken by a user and a synthesized voice of a source language generated from the original voice. SOLUTION: A microphone 3 that picks up speech uttered by a user, a recording unit 31 that records original sound picked up by the microphone, and synthesized speech of a source language corresponding to the source speech and synthesized speech of other languages are acquired. The synthesized speech acquisition unit 26, the speech output unit 27 that outputs the source speech, the synthesized speech of the source language and the synthesized speech of another language from the speaker, and the source language and source language as source language speech based on the user setting information An output control unit 36 that controls to output any one of the synthesized voices. [Selection] Figure 3

Description

  The present invention relates to a loudspeaker that outputs voice spoken by a user and a control method thereof.

  A loudspeaker that amplifies and outputs the voice spoken by the user is used to simultaneously notify a large number of people of messages related to guidance for evacuation, guidance and guidance for security, business instructions, etc. .

  On the other hand, many foreign tourists stay at airports, train stations, hotels, sightseeing spots, and many foreign workers work at factories, warehouses, construction sites, etc. In places, it is desirable to broadcast necessary messages in a foreign language that can be understood by foreigners.

  Conventionally, a plurality of message candidates are registered in advance, and the user selects an appropriate message from the plurality of message candidates, as a technique related to notification of such necessary messages in a foreign language voice. Thus, a technique is known in which a foreign language voice corresponding to a selected message is output (see Patent Document 1).

JP 2014-044349 A

  However, since the conventional technique selects and outputs an appropriate message from pre-registered messages, it is suitable for the situation of the site in the registered message in an unexpected situation. When there is nothing, there is a problem that it is not useful, and a configuration that can output an appropriate message in a foreign language according to the situation at the site is desired.

  Moreover, it is good to record the audio | voice which the user uttered and to output repeatedly the audio | voice with the foreign language audio | voice. At this time, it may be desirable to output the recorded original voice as it is depending on the situation at the site, or it may be desirable to output a synthesized voice generated from character information acquired by voice recognition of the original voice. For example, when it is necessary to give a sense of urgency in emergency evacuation guidance, it is desirable to output the original voice as it is, but in normal guidance, the voice in each language is unified with synthesized voice to eliminate the sense of incongruity. It is desirable to do.

  Therefore, according to the present invention, the user can appropriately switch and output the original voice recording the voice uttered by the user and the synthesized voice of the original language generated from the original voice depending on the situation at the site. The main object of the present invention is to provide a loudspeaker device and a control method thereof.

  The loudspeaker of the present invention is a loudspeaker that outputs a voice spoken by a user, a microphone that collects the voice spoken by the user, a recording unit that records the original voice collected by the microphone, and the original A synthesized speech acquisition unit that acquires a synthesized speech of a source language corresponding to speech and a synthesized speech of another language; and a speech output unit that outputs the synthesized speech of the source language, the synthesized speech of the source language, and the synthesized speech of the other language from a speaker; And an output control unit that controls to output either the original speech or the synthesized speech of the source language as the source language speech based on the user setting information.

  Further, the control method of the present invention is a control method of a loudspeaker that outputs the voice spoken by the user, picks up the voice spoken by the user with a microphone, records the original voice picked up by the microphone, When the original speech is selected as the original language speech, the original speech is output from a speaker, and then a synthesized speech of another language corresponding to the original speech is acquired, and a synthesized speech of another language is obtained. When the synthesized speech is selected as the source language speech, the source language synthesized speech corresponding to the source speech is acquired, and the source language synthesized speech is obtained from the speaker. Then, a synthesized speech in another language corresponding to the original speech is acquired, and a synthesized speech in another language is output from the speaker.

  According to the present invention, as the source language speech, either the source speech recorded by the user uttered speech or the source language synthesized speech generated from the source speech is selected according to the situation at the site. Can be appropriately switched and output.

Side view of the loudspeaker 1 according to the present embodiment Block diagram showing a schematic configuration of the loudspeaker 1 Explanatory drawing which shows the outline of the process performed by the control part 6 Explanatory drawing which shows the fixed phrase display screen displayed on the display input panel 9 Explanatory drawing which shows the output audio | voice setting screen displayed on the display input panel 9 Explanatory drawing which shows the gap setting screen displayed on the display input panel 9 Explanatory diagram showing the output status of audio during playback Flow chart showing operation procedure of loudspeaker 1 Flow chart showing operation procedure in audio output (ST114) Flow chart showing operation procedure in audio output (ST114) Flow chart showing operation procedure in audio output (ST114)

  A first invention made to solve the above problem is a loudspeaker device that outputs a voice uttered by a user, a microphone that collects the voice uttered by the user, and an original voice collected by the microphone. A recording unit for recording, a synthesized speech acquisition unit for obtaining synthesized speech in a source language corresponding to the source speech and a synthesized speech in another language, and the source speech, the synthesized speech in the source language, and the synthesized speech in the other language A configuration comprising: an audio output unit that outputs from a speaker; and an output control unit that controls to output either the original speech or the synthesized speech of the original language as source language speech based on user setting information And

  According to this, as the source language speech, the user appropriately selects either the source speech recorded from the speech uttered by the user or the source language synthesized speech generated from the source speech according to the situation at the site. Can be output.

  Further, the second invention is configured such that the output control unit controls to output either a female synthetic voice or a male synthetic voice based on user setting information when outputting the synthetic voice. To do.

  According to this, since the user can appropriately switch the gender according to the situation at the site and output the synthesized speech, the convenience for the user can be improved.

  According to a third aspect of the present invention, the output control unit performs control so as to output a plurality of synthesized speech in other languages in the order specified by the user following the source language speech.

  According to this, since the user can designate the order in which the synthesized speech of another language is output according to the ratio of foreigners staying at the site, the convenience of the user can be improved.

  According to a fourth aspect of the present invention, the sound output unit inserts a silence period of a length specified by the user between the sound of each language when outputting the sound of the source language and the sound of the other language. The configuration is as follows.

  According to this, it becomes easy to hear the sound of each language.

  In addition, the fifth invention further includes a speech analysis unit that acquires feature information of the original speech, and a synthesized speech of the original speech and the other language when outputting the original speech and the synthesized speech of the other language. And a voice adjustment unit that performs processing for matching the voice characteristics.

  According to this, since the features of the voice (original voice and multilingual synthesized voice) continuously output from the loudspeaker are shared, it is possible to reduce the uncomfortable feeling given to the person who listens.

  According to a sixth aspect of the present invention, the voice adjustment unit is configured to match at least one of voice gender, tempo, volume, and height between the original voice and the synthesized voice of the other language.

  According to this, since the features of the voice (original voice and multilingual synthesized voice) continuously output from the loudspeaker are shared, it is possible to reduce the uncomfortable feeling given to the person who listens.

  The seventh invention is a method of controlling a loudspeaker that outputs a voice uttered by a user, the voice uttered by the user is picked up by a microphone, the original voice picked up by the microphone is recorded, and the original voice is recorded. When the original voice is selected as a language voice, the original voice is output from a speaker, and then a synthesized voice of another language corresponding to the original voice is acquired, and a synthesized voice of the other language is obtained. When the synthesized speech is selected as the source language speech output from the speaker, the source language synthesized speech corresponding to the source speech is acquired and the source language synthesized speech is output from the speaker. Subsequently, a synthesized speech in another language corresponding to the original speech is acquired, and a synthesized speech in another language is output from the speaker.

  According to this, similarly to the first invention, the user can appropriately switch and output the original voice recording the voice spoken by the user and the synthesized voice of the original language generated from the original voice.

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

  FIG. 1 is a side view of a loudspeaker 1 according to the present embodiment.

  In the loudspeaker 1, a microphone 3, a voice switching unit 5, a control unit 6, an amplifier 7, a speaker 8, and a display input panel 9 are accommodated in the housing 2. Further, a grip 10 held by the user is attached to the lower part of the housing 2. A conical horn portion 11 is provided on the front side of the speaker 8 in the housing 2. A battery (not shown) is also accommodated in the housing 2.

  The microphone 3 collects the voice uttered by the user.

  The control unit 6 records a standard sentence similar to the process of recording the original voice collected by the microphone 3, the process of converting the recorded original voice into character information (speech recognition), and the character information (original text) of the original voice. A search process (search), a process of generating synthesized speech from the text information of a fixed sentence (speech synthesis), a process of reproducing synthesized speech and original speech, and the like are performed.

  The voice switching unit 5 switches a voice path (voice path) in the megaphone mode (first operation mode) and a voice path in the translation mode (second operation mode). The amplifier 7 amplifies the sound output from the sound switching unit 5. The speaker 8 outputs the sound amplified by the amplifier 7.

  In the megaphone mode, the original sound picked up by the microphone 3 is directly amplified by the amplifier 7 and output from the speaker 8. In the translation mode, the original voice picked up by the microphone 3 is input to the control unit 6, and the synthesized voice or the original voice generated by the control unit 6 is amplified by the amplifier 7 and output by the speaker 8.

  A mode change switch 12 is provided on the side surface of the housing 2. The mode switch 12 switches between the megaphone mode and the translation mode, and the voice switching unit 5 switches between the voice path in the megaphone mode and the voice path in the translation mode according to the operation of the mode switch 12. It is done.

  The display input panel 9 (display unit, input unit) is configured by a so-called touch panel display in which a touch panel and a liquid crystal display are combined, and the projection 18 formed on the upper portion of the housing 2 with the screen facing upward. Contained.

  A power switch 13 is provided on the side surface of the convex portion 18.

  A recording switch 14 and a volume adjustment switch 15 are provided on the front side of the grip 10. On the opposite side of the grip 10, an output switch 16 and a lock switch 17 are provided. The user can operate the recording switch with the thumb of the hand while holding the grip 10 with one hand, and can operate the output switch with the index finger.

  The recording switch 14 instructs to record the sound collected by the microphone 3. The volume adjustment switch 15 is for adjusting the volume of the sound output from the speaker 8.

  The output switch 16 performs an operation for operating the amplifier 7 in the megaphone mode. When the output switch 16 is pressed, the original sound collected by the microphone 3 is amplified by the amplifier 7 and output. On the other hand, in the translation mode, the output switch 16 operates the amplifier 7 and instructs the control unit 6 to reproduce the sound. When the output switch 16 is pressed, the control unit 6 reproduces the sound, The sound is amplified by the amplifier 7 and output. At this time, the sound is repeatedly reproduced by continuously pressing the output switch 16.

  The lock switch 17 holds the output switch 16 in a pressed state. Thereby, even if the user does not keep pressing the output switch 16, the sound output can be continued.

  The mode switch 12 is a rocker switch, the recording switch 14 and the output switch 16 are pushbutton switches, and the volume control switch 15 is a rotary switch.

  Next, a schematic configuration of the loudspeaker 1 will be described. FIG. 2 is a block diagram showing a schematic configuration of the loudspeaker 1. FIG. 3 is an explanatory diagram showing an outline of the processing performed by the control unit 6.

  A signal from the mode switch 12 is input to the voice switching unit 5. The voice switching unit 5 switches the voice path in the megaphone mode and the voice path in the translation mode in accordance with the operation of the mode switch 12. The voice switching unit 5 includes an input switching unit 21 and an output switching unit 22. I have. The input switching unit 21 outputs the sound output from the microphone 3 to either the output switching unit 22 side or the control unit 6 side. The output switching unit 22 outputs the sound input from either the input switching unit 21 or the control unit 6 to the amplifier 7 side.

  A level adjustment unit 23 is provided between the voice switching unit 5 and the control unit 6. In the level adjusting unit 23, the level of the sound output from the input switching unit 21 of the sound switching unit 5 is adjusted.

  A signal from the volume adjustment switch 15 is input to the volume adjustment unit 24. The volume adjustment unit 24 is provided between the audio switching unit 5 and the amplifier 7. In the volume adjusting unit 24, the volume of the sound output from the output switching unit 22 of the sound switching unit 5 is adjusted according to the operation of the volume adjustment switch 15.

  A signal from the output switch 16 is input to the amplifier 7. The output switch 16 functions as a switch for intermittently supplying power to the amplifier 7, and when the output switch 16 is pressed in the megaphone mode, the amplifier 7 is energized to enter a sound output state, and the original sound input from the microphone 3 is received. Amplified by the amplifier 7 and output. On the other hand, when the output switch 16 is pressed in the translation mode, the amplifier 7 is energized to enter the audio output state, and the reproduction unit 35 reproduces the audio, and the audio output from the reproduction unit 35 is amplified by the amplifier 7. Is output.

  The storage unit 25 temporarily stores, in the control unit 6, the original voice that has been recorded by the user, the original sentence acquired by voice recognition of the original voice, and the synthesized voice converted from the character information of the fixed sentence. The storage unit 25 stores a fixed phrase database. Many fixed sentences are registered in this fixed sentence database.

  In addition, the storage unit 25 stores information input by the user on the setting screen displayed on the display input panel 9 as user setting information. In the present embodiment, as user setting information, information on whether or not to output the original voice as the voice of the original language, information on the gender (female or male) of the synthesized voice, a plurality of other languages (English, Chinese, etc.) ) Is stored.

  The control unit 6 includes a recording unit 31, a voice recognition unit 32, a search unit 33, a voice synthesis unit 34, a reproduction unit 35, an output control unit 36, a voice analysis unit 37, a voice adjustment unit 38, It has. The control unit 6 includes a processor, and each unit of the control unit 6 is realized by executing a program stored in the storage unit 25.

  The voice recognition unit 32, the search unit 33, and the voice synthesis unit 34 constitute a synthesized voice acquisition unit 26. In addition, the audio output unit 27 is configured by the reproduction unit 35, the amplifier 7, and the speaker 8 (see FIG. 3).

  The control unit 6 receives a signal from the mode change switch 12, a signal from the recording switch 14, and a signal from the output switch 16.

  The recording unit 31 records the original sound output from the level adjustment unit 23. In this recording process, an audio signal (analog signal) is A / D converted into audio data and stored in the storage unit 25. The recording unit 31 starts the recording process when the recording switch 14 is pressed, and ends the recording process when the recording switch 14 is released.

  The voice recognition unit 32 performs voice recognition for converting the original voice recorded by the recording unit 31 into character information, and acquires the original sentence (character information of the original voice) as the voice recognition result. This original text is temporarily stored in the storage unit 25.

  The search unit 33 searches for a fixed phrase having the highest similarity to the original sentence from fixed phrases in the original language (for example, Japanese) registered in the fixed phrase database (see FIG. 3).

  The speech synthesizer 34 acquires a source language fixed sentence from a fixed sentence database, and generates a synthesized speech of the source language by speech synthesis from the character information of the fixed sentence of the source language (for example, Japanese). A fixed sentence in another language is acquired from a fixed sentence database, and synthesized speech in another language is generated by speech synthesis from character information of the fixed sentence in another language (for example, English, Chinese, etc.). The synthesized speech generated by the speech synthesizer 34 is temporarily stored in the storage unit 25.

  The output control unit 36 controls to start outputting sound when the output switch 16 is pressed and to stop outputting sound when the output switch 16 is released. At this time, the output control unit 36 instructs the speech synthesizer 34 to synthesize speech of the source language fixed phrases acquired by the search unit 33 and the fixed phrases of other languages corresponding thereto, and the speech synthesizer 34 further generates them. The reproduction unit 35 is instructed to reproduce the synthesized speech (see FIG. 3). In addition, the output control unit 36 outputs the synthesized speech of a plurality of other languages in the order designated by the user, following the source language speech, based on the user setting information. To instruct.

  Further, the output control unit 36 controls to output either the original speech or the synthesized speech of the original language (for example, Japanese) as the source language speech based on the user setting information. Here, when outputting the original sound, the original sound is acquired from the storage unit 25 and the reproduction unit 35 is instructed to reproduce the original sound (see FIG. 3). On the other hand, when outputting synthesized speech in the source language, the standard sentence in the source language is acquired from the storage unit 25, the speech synthesis unit 34 is instructed to synthesize speech in the source language fixed sentence, and the speech synthesis unit The reproduction unit 35 is instructed to reproduce the synthesized speech of the source language generated in 34 (see FIG. 3).

  When outputting the original voice, the voice spoken by the user collected by the microphone 3 (for example, “Distribute materials here”) itself is output, but the synthesized speech in the original language is output. In the case of doing so, the standard sentence having the highest similarity to the speech uttered by the user is acquired from the standard sentence stored in the storage unit 25 (for example, “We distribute materials here.”). However, since the speech synthesis of this fixed sentence is output, it may be slightly different from the content of the speech that the user actually uttered.

  Further, when outputting the synthesized speech, the output control unit 36 performs control so as to output either the female synthesized speech or the male synthesized speech based on the user setting information. In the case of outputting female synthesized speech, the voice synthesizer 34 is instructed to generate female speech synthesis. In the case of outputting male synthesized speech, the voice synthesizer 34 is instructed to generate male synthesized speech. To do.

  The playback unit 35 plays back the original voice recorded by the recording unit 31 and the synthesized voice generated by the voice synthesis unit 34. In this reproduction process, a process of D / A converting original voice and synthesized voice data into a voice signal (analog signal) is performed. Note that the original speech and synthesized speech are temporarily stored in the storage unit 25, and when the output switch 16 is pressed, the source language speech (original speech or synthesized speech) and the synthesized speech of other languages are in a predetermined order. Is played back repeatedly continuously.

  Further, when outputting the speech in the original language (original speech or synthesized speech) and the speech in another language, the playback unit 35 creates a gap (silence period) of a length specified by the user between the speech in each language. insert. This gap (silence period) is also preferably stored as user setting information based on the user's specification, and the output control unit 36 may notify the playback unit 35 of the gap.

  The voice analysis unit 37 acquires the original voice from the storage unit 25 and acquires the feature information of the original voice. In the present embodiment, information on gender (male voice or female voice), tempo (speed), volume, and height (tone) is acquired as characteristic information of the original voice.

  When outputting the original voice and the synthesized voice of another language, the voice adjustment unit 38 performs a process of matching the voice characteristics between the original voice and the synthesized voice. In the present embodiment, adjustments relating to at least one of gender (male voice or female voice), tempo (speed), volume, and height (tone) are performed as voice characteristics.

  Further, in the present embodiment, there are a mode in which the synthesized speech is matched with the original speech and a mode in which the synthesized speech is matched with the synthesized speech, and the user can select either mode. In the mode in which the synthesized speech is matched with the original speech, the speech synthesizer 34 generates the synthesized speech based on the feature information of the original speech so that the gender, tempo, volume and height of the synthesized speech match the original speech. In the mode in which the original voice is matched with the synthesized voice, the original voice is converted so that the original voice matches the standard tempo, volume, and height adopted in the initial setting of the synthesized voice.

  Thus, if the features of the speech are made to match between the original speech and the synthesized speech, the features of the speech (original speech and synthesized speech) that are continuously output from the loudspeaker 1 are shared, so The unpleasant feeling given can be reduced.

  Next, the fixed phrase display screen displayed on the display input panel 9 will be described. FIG. 4 is an explanatory diagram showing a fixed phrase display screen.

  A fixed sentence display unit 41 is provided on the fixed sentence display screen. In the present embodiment, the search unit 33 searches for a fixed phrase having the highest similarity to the original sentence generated by speech recognition of the original speech uttered by the user, and the fixed sentence of the original language (Japanese) found here is searched. Is displayed on the fixed phrase display unit 41.

  In addition, a reproduction order display unit 42 is provided on the fixed sentence display screen. The playback order display unit 42 displays the playback order of the source language (Japanese) and other languages (English, Chinese, Korean, etc.) based on the user setting information. The reproduction order display unit 42 is provided with a national flag icon 43. By operating the national flag icon 43, character information of other languages is displayed on the standard sentence display unit 41. Note that the user sets the playback order on a setting screen (not shown).

  When the output switch 16 is pressed while the standard sentence display screen is displayed, the original language voice and the other language voice related to the standard sentence displayed on the standard sentence display screen are displayed on the reproduction order display unit 42. Are output in order.

  Next, the output audio setting screen displayed on the display input panel 9 will be described. FIG. 5 is an explanatory diagram showing an output audio setting screen.

  An output sound selection unit 51 is provided on the output sound setting screen. The output voice selection unit 51 is provided with two radio buttons 52. By operating the radio button 52, a user can select either a mode for outputting a female synthesized voice or a mode for outputting a male synthesized voice. Can be selected. In addition, the output sound selection unit 51 is provided with a check box 53, and by the operation of the check box 53, the user determines whether or not to output the original sound in the original language (Japanese) sound. You can choose.

  In addition, a sound adjustment selection unit 54 is provided on the output sound setting screen. The sound adjustment selection unit 54 is provided with a check box 55, and the user can select whether or not to perform sound adjustment by operating the check box 55. Further, the radio adjustment selection unit 54 is provided with two radio buttons 56, and by operating these radio buttons 56, it is determined whether the synthesized voice is adjusted to the original voice or the original voice is adjusted to the synthesized voice when the voice is adjusted. The user can select.

  The output sound setting screen is provided with a cancel button 57 and an OK button 58. When the cancel button 57 is operated, the contents selected by the user in the output sound selection unit 51 and the sound adjustment selection unit 54 are discarded, and the screen returns to the setting menu screen (not shown). When the OK button 58 is operated, the user setting information in the storage unit 25 is updated with the contents selected by the user in the output audio selection unit 51 and the audio adjustment selection unit 54, and the screen returns to the setting menu screen (not shown).

  As described above, in the present embodiment, as the source language speech, the user selects either the original speech recorded from the speech uttered by the user or the synthesized speech synthesized from the text information of the fixed sentence corresponding to the original speech. You can select and output. For example, when it is necessary to give a sense of urgency in emergency evacuation guidance, it may be set to output the original voice. Further, in order to eliminate a sense of incongruity in normal guidance, it is preferable to select synthesized speech as source language speech and unify the speech of each language with synthesized speech.

  In the present embodiment, the user can select the sex (male voice or female voice) of the synthesized voice. For this reason, it is possible to output a synthesized voice having sex suitable for the use situation. For example, a male voice may be selected when a sense of urgency is required for emergency evacuation guidance, and a female voice may be selected for normal guidance.

  Further, in the present embodiment, the user can select whether or not to perform sound adjustment, and further, the user can select whether the synthesized speech is matched with the original speech or the original speech is matched with the synthesized speech. . For this reason, the sound suitable for the use situation can be output. For example, if the original voice is fast, the original voice is synthesized so that the tempo of the original voice is slow so that it is easy to hear. The tempo of the synthesized voice can be increased according to the voice.

  Next, the gap setting screen displayed on the display input panel 9 will be described. FIG. 6 is an explanatory diagram showing a gap setting screen.

  In this gap setting screen, a plurality of radio buttons 61 are provided, and the length of a gap (silence period) inserted between voices of each language when voices are output by operating the radio buttons 61. The (gap time) can be selected by the user. In the example shown in FIG. 6, four radio buttons 61 are provided, and one of 0.5 seconds, 1.0 seconds, 2.0 seconds, and 3.0 seconds can be selected.

  In addition, a cancel button 62 and an OK button 63 are provided on the gap setting screen. When the cancel button 62 is operated, the content selected by the user is discarded and the screen returns to the setting menu screen (not shown). When the OK button 63 is operated, the user setting information in the storage unit 25 is updated with the contents selected by the user, and the screen returns to the setting menu screen (not shown).

  Next, the audio output status during reproduction will be described. FIG. 7 is an explanatory diagram showing the output status of audio during playback.

  In the present embodiment, the user can select whether to output the original voice with priority and whether to output the female or male synthetic voice, and this information is stored in the storage unit 25 as user setting information. When the speech is stored and output, necessary speech synthesis processing is performed based on the user setting information, and the speech in each language is output in order. FIG. 7 shows an example in which Japanese is selected as the source language and English and Chinese are selected as the other languages.

  Here, when the original voice is given priority and the female synthetic voice is set to be output, as shown in FIG. 7A, the original voice (Japanese), the English female synthetic voice, Chinese female synthesized speech is output in order. Further, when the original voice is given priority and the male synthetic voice is set to be output, as shown in FIG. 7B, the original voice (Japanese), the English male synthetic voice, the Chinese The male synthesized speech of words is output in order.

  Further, when the original voice is not given priority and the female synthetic voice is set to be output, as shown in FIG. 7C, the Japanese female synthetic voice, the English female synthetic voice, Chinese female synthesized speech is output in order. Further, when the original voice is not given priority and the male synthetic voice is set to be output, as shown in FIG. 7D, a Japanese male synthetic voice, an English male synthetic voice, Chinese male synthesized speech is output in order.

  Further, when outputting the sound of each language, a gap (silence period) having a length designated by the user is inserted between the sounds of each language. For this reason, it becomes easy to hear the sound of each language.

  Next, the operation of the loudspeaker 1 will be described. FIG. 8 is a flowchart showing the operation procedure of the loudspeaker 1.

  In the loudspeaker 1, first, if the mode switch 12 is not in the translation mode, that is, in the megaphone mode (No in ST101), the voice switching unit 5 outputs the original voice collected by the microphone 3 as it is. When the output switch 16 is pressed (Yes in ST102), the amplifier 7 enters a sound output state and starts outputting the original sound (ST103). At this time, the original voice spoken by the user is directly amplified by the amplifier 7 and output from the speaker 8. When the output switch 16 is returned (Yes in ST104), the output of the original voice is stopped (ST105). When the output switch 16 is not pressed (No in ST102), no special operation is performed.

  On the other hand, when the mode switch 12 is in the translation mode (Yes in ST101), the control unit 6 next determines whether or not the recording switch 14 is pressed (ST106). If the recording switch 14 is pressed (Yes in ST106), the recording unit 31 starts a process of recording the original sound collected by the microphone 3 (ST107). At this time, the user may be notified of the start of recording by vibration or notification sound. When the recording switch 14 is returned (Yes in ST108), recording is stopped (ST109).

  Next, the voice recognition unit 32 performs voice recognition for converting the recorded original voice into character information (ST110). Next, the search unit 33 performs a search for searching for a fixed phrase most similar to the original sentence (character information of the original voice) (ST111). Then, a fixed sentence display screen (see FIG. 4) for displaying the fixed sentence found by the search unit 33 is displayed on the display input panel 9 (ST112).

  Next, it is determined whether or not the output switch 16 is pressed (ST113). Here, when the output switch 16 is pressed (Yes in ST113), the speech synthesizer 34 generates synthesized speech from the standard sentence, and the playback unit 35 starts processing to reproduce the synthesized speech. The synthesized voice is output from the speaker 8 (ST114). At this time, the sound of each language is repeatedly reproduced in order by the reproduction unit 35. When the output switch 16 is returned (Yes in ST115), the audio output is stopped (ST116).

  Next, an operation procedure in audio output (ST114) will be described. 9, FIG. 10 and FIG. 11 are flowcharts showing an operation procedure in audio output (ST114). Here, an example is shown in which the source language is Japanese and the other languages are English and Chinese.

  In the audio output (ST114), first, as shown in FIG. 9, the control unit 6 determines whether or not it is a setting for performing audio adjustment based on the user setting information (ST201).

  If the setting is not for performing audio adjustment (No in ST201), it is next determined whether or not the setting is for preferential output of the original audio (ST202). If the setting is to output the original voice preferentially (Yes in ST202), it is next determined whether or not the setting is to output the female synthesized voice (ST203).

  If the setting is to output female synthesized speech (Yes in ST203), first, the original speech is acquired from the storage unit 25, and the speech output unit 27 outputs the original speech (ST204). Subsequently, the English text (English character information) of the fixed sentence specified by the user as the output target is acquired from the fixed sentence database, and the female synthesized speech is generated from the English text in the speech synthesizer 34. A female synthesized voice is output (ST205). Next, Chinese text (Chinese character information) is acquired from the fixed phrase database, and a female synthesized speech is generated from the Chinese text and output (ST206).

  On the other hand, if it is not set to output female synthesized speech, that is, is set to output synthesized male speech (No in ST203), first, the original speech is acquired from the storage unit 25 and the original speech is output. (ST207). Next, an English text is acquired from the fixed phrase database, and male synthesized speech is generated from the English text and output (ST208). Next, Chinese text is acquired from the fixed phrase database, and male synthesized speech is generated from the Chinese text and output (ST209).

  If it is not set to output the original voice preferentially (No in ST202), it is next determined whether or not it is set to output the female synthesized voice as shown in FIG. 10 (ST210).

  If the setting is to output female synthesized speech (Yes in ST210), first, the Japanese text of the standard text specified as the output target by the user (Japanese character information) is acquired from the standard text database. Then, female synthesized speech is generated from the Japanese text and output (ST211). Next, an English text is acquired from the fixed phrase database, and a female synthesized speech is generated and output from the English text (ST212). Next, Chinese text is acquired from the fixed phrase database, and female synthesized speech is generated from the Chinese text and output (ST213).

  On the other hand, if it is not set to output female synthesized speech, that is, it is set to output male synthesized speech (No in ST210), first, the Japanese text of the standard text specified by the user as the output target is the standard text database. , And generates and outputs male synthesized speech from the Japanese text (ST214). Next, an English text is acquired from the fixed phrase database, and male synthesized speech is generated from the English text and output (ST215). Next, Chinese text is acquired from the fixed phrase database, and male synthesized speech is generated from the Chinese text and output (ST216).

  Also, as shown in FIG. 9, if the setting is to perform voice adjustment (Yes in ST201), then as shown in FIG. 11, the voice analysis unit 37 performs the characteristics of the original voice (gender, tempo). , Volume and height) are detected (ST217).

  Next, the voice adjustment unit 38 determines whether or not the synthesized voice is set to match the original voice based on the user setting information (ST218).

  Here, when the synthetic voice is set to match the original voice (Yes in ST218), the original voice is first acquired from the storage unit 25, and the voice output unit 27 outputs the original voice (ST219). Next, the English text is acquired from the fixed phrase database, and the speech synthesizer 34 generates synthesized speech from the English text so as to match the gender, tempo, volume and height of the original speech, and the speech output unit 27 Synthetic speech is output (ST220). Next, Chinese text is acquired from the fixed phrase database, synthesized speech is generated from the Chinese text to match the gender, tempo, volume, and height of the original speech, and the synthesized speech is output (ST221). .

  On the other hand, if the synthesized voice is not set to match the original voice, that is, it is set to match the original voice to the synthesized voice (No in ST218), the original voice information is acquired from the storage unit 25, and the initial setting for the synthesized voice is obtained. The original voice is converted so as to have the standard tempo, volume and height adopted in the above, and the original voice is output (ST222). Next, an English text is acquired from the fixed phrase database, a synthesized speech is generated from the English text with the gender of the original speech, and the synthesized speech is output (ST223). Next, Chinese text is acquired from the fixed phrase database, synthesized speech is generated from the Chinese text with the gender of the original speech, and the synthesized speech is output (ST224).

  In the present embodiment, when the original sound is set to be combined with the synthesized voice, characteristics (tempo, volume, and height) other than gender are the standard voice generation conditions employed in the initial setting for the synthesized voice. The original voice is adjusted so that it fits, and with regard to the gender, the synthesized voice is generated with the gender of the original voice, but the original gender (for example, female), You may make it convert into the audio | voice of the sex designated by the user. In this case, synthesized speech in other languages (English, Chinese) is also generated with the initial gender or the gender specified by the user.

  In addition, in order to match the characteristics of the voice between the original voice and the synthesized voice, the synthesized voice is matched with the original voice or the original voice is matched with the synthesized voice. You may make it match | combine with the audio | voice of a predetermined characteristic.

  As described above, the embodiments have been described as examples of the technology disclosed in the present application. However, the technology in the present disclosure is not limited to this, and can be applied to embodiments in which changes, replacements, additions, omissions, and the like have been performed. Moreover, it is also possible to combine each component demonstrated by said embodiment into a new embodiment.

  For example, in the above-described embodiment, a fixed sentence in the source language (Japanese) having a high similarity to the original sentence acquired by speech recognition of speech uttered by the user is searched, and another language corresponding to the fixed sentence in the original language is searched. Was obtained, and synthesized speech of other languages was generated from the fixed phrases of other languages.However, by translating the original sentence using a translation engine, the sentences of other languages were obtained, You may make it produce | generate the synthetic speech of another language from a text.

  In the above-described embodiment, each process such as synthetic voice acquisition (speech recognition, search, voice synthesis), voice analysis, and voice adjustment is performed by the loudspeaker 1. For example, the original voice) may be transmitted to the server device, and the server device may perform all or part of the above-described processes. Moreover, you may make it make a server apparatus perform character translation using the said translation engine.

  According to the loudspeaker and the control method thereof according to the present invention, the user appropriately selects the original voice that is recorded from the voice spoken by the user and the synthesized voice of the original language that is generated from the original voice according to the situation at the site. It has the effect of being able to be switched and output, and is useful as a loudspeaker that outputs the voice spoken by the user and its control method.

DESCRIPTION OF SYMBOLS 1 Loudspeaker 3 Microphone 6 Control part 7 Amplifier 8 Speaker 14 Recording switch 16 Output switch 25 Storage part 26 Synthetic voice acquisition part 27 Voice output part 31 Recording part 32 Voice recognition part 33 Search part 34 Speech synthesis part 35 Playback part 36 Output control Part 37 voice analysis part 38 voice adjustment part

Claims (7)

A loudspeaker that outputs a voice spoken by a user,
A microphone that picks up the voice spoken by the user;
A recording unit for recording the original sound picked up by the microphone;
A synthesized speech acquisition unit that acquires a synthesized speech of a source language corresponding to the original speech and a synthesized speech of another language;
A voice output unit for outputting the original voice, the synthesized voice of the source language and the synthesized voice of the other language from a speaker;
An output control unit that controls to output either the original speech or the synthesized speech of the source language as the source language speech based on the user setting information;
A loudspeaker comprising:   2. The output control unit according to claim 1, wherein when outputting the synthesized speech, the output control unit controls to output either a female synthesized speech or a male synthesized speech based on user setting information. Loudspeaker.   3. The control according to claim 1, wherein the output control unit performs control so as to output a plurality of synthesized speech of the other languages in an order designated by a user following the speech of the source language. Loudspeaker.   The voice output unit inserts a silence period of a length specified by a user between voices of each language when outputting the voice of the source language and the voice of the other language. The loudspeaker according to any one of claims 1 to 3. Furthermore, a voice analysis unit that acquires feature information of the original voice;
And a voice adjustment unit configured to perform a process of matching voice characteristics between the original voice and the synthesized voice of the other language when outputting the original voice and the synthesized voice of the other language. The loudspeaker according to any one of claims 1 to 4.   6. The loudspeaker according to claim 5, wherein the voice adjustment unit matches at least one of voice gender, tempo, volume, and height between the original voice and the synthesized voice of the other language. A method for controlling a loudspeaker that outputs voice spoken by a user,
The voice uttered by the user is picked up by the microphone,
Record the original voice picked up by the microphone,
When the original speech is selected as the original language speech, the original speech is output from a speaker, and then a synthesized speech of another language corresponding to the original speech is acquired, and a synthesized speech of another language is obtained. Is output from the speaker,
When synthesized speech is selected as the source language speech, the source language synthesized speech corresponding to the source speech is acquired, and the source language synthesized speech is output from the speaker. A control method comprising: obtaining synthesized speech of another language corresponding to the original speech, and outputting synthesized speech of another language from the speaker.

Images