JP6522679B2 - Speech control apparatus, method, speech system, and program - Google Patents

Description

  The present invention relates to a speech control apparatus, method, speech system, and program for determining the contents of speech to be made to speech in a speech apparatus.

  In recent years, methods have been known in which information desired by the user is provided by the device emitting speech. As speech to be used in this case, synthetic speech in which speech stored in advance in a device is synthesized by signal processing is known.

  For example, Patent Document 1 includes a voice synthesis database in which voices of a plurality of speakers are separately recorded, and a playback device that switches to a voice synthesis database desired by the user according to an instruction from the user, and plays the voice. Is disclosed.

JP 2005-321706 A (released on November 17, 2005)

  Generally, in human-to-human conversations, different phrases are used with the same meaning depending on the mood and character of the person who emits. For example, it may be said that "Good morning" and then return "Good morning", and may return phrases other than "Good morning", or change the "Good morning" intonation and feel the mood at that time. There is also a case to express.

  On the other hand, in the technology described in Patent Document 1, although the voice color of the human voice to be reproduced can be changed to the voice color desired by the user, the user can not change the emitted phrase or intonation. It was difficult to achieve smooth communication between

  The present invention has been made in view of the above problems, and an object thereof is to provide a speech control apparatus capable of achieving smooth communication with a user as compared with the prior art.

  In order to solve the above-mentioned subject, a speech control device concerning one mode of the present invention is a speech control device which determines the contents of speech which makes a speech device utter, and voice information which acquires input speech information from the above-mentioned speech device According to an acquisition unit, a voice information recognition unit for recognizing the input voice information, a mode set in the utterance control device, and recognition information recognized by the voice information recognition unit An utterance content determination unit that determines the utterance content by referring to a speech output unit that outputs the utterance content determined by the utterance content determination unit to the utterance device.

  In order to solve the above problems, a method according to an aspect of the present invention is a method of determining the content of speech to be uttered by a speech device, and a voice information acquisition step of acquiring input speech information from the speech device; The utterance content is determined by referring to one or more databases according to the voice information recognition process for recognizing the input voice information, the set mode, and the recognition information recognized in the voice information recognition process. And an audio output step of outputting the utterance content determined in the utterance content determination step to the utterance device.

  In order to solve the above-mentioned subject, a speech system concerning one mode of the present invention is a speech system provided with a speech device and a speech control device, and the above-mentioned speech control device is an input voice from the above-mentioned speech device According to the voice information acquisition unit for obtaining information, the voice information recognition unit for recognizing the input voice information, the mode set in the speech control device, and the recognition information recognized by the voice information recognition unit The utterance device includes an utterance content determination unit that determines utterance content by referring to one or more databases, and an audio output unit that outputs the utterance content determined by the utterance content determination unit to the utterance device. And an utterance content acquisition unit that acquires the utterance content output from the utterance control device, and an utterance unit that utters the acquired utterance content.

  According to one aspect of the present invention, the speech control device can communicate more smoothly with the user than in the related art.

It is a block diagram which shows the structure of the speech system which concerns on Embodiment 1 of this invention. In the speech system concerning Embodiment 1 of the present invention, it is an example of a database which a speech contents deciding part refers to in order to decide speech contents. It is an example of the recognition phrase and reply phrase in the speech system concerning Embodiment 1 of the present invention. In the speech system concerning Embodiment 5 of the present invention, it is an example of the ambiguous database which a speech contents deciding part refers to in order to decide speech contents. It is an example of the ambiguous phrase in the speech system concerning Embodiment 5 of the present invention. In the speech system concerning Embodiment 6 of the present invention, it is an example of a database which a speech contents deciding part refers to in order to decide speech contents. It is a flowchart which shows the flow of the process which changes the mode set to the server in the speech system which concerns on Embodiment 7 of this invention.

Embodiment 1
Hereinafter, Embodiment 1 of the present invention will be described in detail.

(Configuration of Utterance System 1)
FIG. 1 is a block diagram showing the configuration of the speech system 1 according to the first embodiment of the present invention. The speech system 1 includes a speech device 10 and a server (speech control device) 20, as shown in FIG.

  In the speech system 1, the server 20 has a plurality of modes. The server 20 determines the content of the utterance according to the mode selected and set among the plurality of modes (hereinafter, the set mode is also referred to as “the utterance mode”), and the utterance content is uttered from the uttering device 10 Let

  Here, the utterance content is information including a phrase to be uttered by the utterance device 10 and intonation. The utterance content also includes the voice color, the volume, the speech speed, and the pitch of the speech in which the speech apparatus 10 utters the phrase in the intonation.

  Further, the phrase may be composed of one or more words, and is not limited to the number of words or the composition of the sentence, and it may be a single word or a sentence composed of a plurality of words. , May be multiple sentences.

  In addition, intonation is not limited to sentence breaks (the length of a silence period between each pronunciation) and the pitch of sentences, and includes word accents. For example, speaking "Good morning" flatly without accent as in the case of Tokyo dialect and speaking with accenting "y" like the Kansai dialect are different in intonation. In addition, although the phrase always includes an intonation, in the present embodiment and the other embodiments, the reference to the intonation is omitted in the case where it may be any intonation.

  A mode is associated with a phrase or a phrase and intonation, more specifically, for example, mode 1 in which the phrase is "Good morning" and the intonation is flat, the phrase is "Good morning", and the intonation is in Mode 2 is accented with "Yo", mode 3 in which the phrase is "Good morning" and the intonation is flat. Therefore, if the selected mode is different, at least one of the phrase and the intonation of the corresponding utterance content will be different.

  Also, the mode may be a parameter simply for the server 20 to determine the content of the speech, but may be made to match the character of the speech device 10. For example, when the server 20 has the fine mode and the fine mode is selected, the server 20 may determine the content of the utterance as if the speech apparatus 10 is a fine character. Also, for example, when the server 20 has the Kansai dialect mode and the Kansai dialect mode is selected, the server 20 determines the content of the utterance as if the utterance device 10 is a character from Kansai. Good.

  The modes that can be set in the server 20 are, for example, as follows.

-Tokyo valve mode-Kansai valve mode-Standard mode-Energy mode-Polite mode It is also possible to select a plurality of modes from the above-described settable modes and set them in the server 20 in an overlapping manner. For example, it is possible to set the server 20 as a mode such as a energetic Kansai valve mode in which the Kansai valve mode and the energetic mode are superimposed. Further, the mode may be changed by the operation of the user.

  Also, for each mode, the server 20 may manage the character names in correspondence. For example, the server 20 is made to correspond to character names such as "Edokko Koro-chan" for the Tokyo valve mode, "Genki ai-chan" for the fine mode, and "Lobby Nao-chan" for the polite mode, etc. This may be the name of the mode.

  Furthermore, when a character image (graphics, color, etc.) is associated with each mode and the mode is set in the server 20, the speech device 10 causes the display unit described later to display a character image corresponding to the set mode. May be

(Configuration of the speech device 10)
As shown in FIG. 1, the speech device 10 includes a communication unit (speech content acquisition unit) 11, a speech information acquisition unit 12, a speech unit 13, an operation reception unit 14, and a display unit 15.

  The communication unit 11 is connected to the server 20, and can transmit and receive data to and from the server 20.

  The voice information acquisition unit 12 obtains voices around the speech device 10, and outputs the obtained voice as input voice information.

  The utterance unit 13 acquires the utterance content, and utters a phrase included in the acquired utterance content with intonation included in the utterance content.

  The operation accepting unit 14 accepts information input by the user operating a hard key, a switch, a touch sensor or the like. In addition, the operation accepting unit 14 can also accept voices uttered by the user as information manipulated by the user. The operation accepting unit 14 outputs the accepted information as operation information.

  The display unit 15 has a function of notifying the user of various types of information. The display unit 15 includes at least one of a display panel represented by a liquid crystal display panel and a light emitting unit configured to be capable of emitting light, and notifies the user of various information via the display panel or the light emitting unit. It is preferable to do. When the display unit 15 includes the light emitting unit, the light can be notified to the user by emitting light from the light emitting unit.

  In addition, the display unit 15 may have a configuration that indicates the set mode to the user. For example, when the display unit 15 includes a display panel, the display unit 15 may select characters (such as "Edokko Coro", "Genki Ai", "Girls Nao", etc.) corresponding to the set mode. An image (graphics) is displayed on the display unit 15. This image may be a moving image, and it is more preferable to operate according to the speech. This image may be downloaded from the server 20 and displayed on the display unit 15 according to the set mode at the time of mode setting to be described later, or a storage unit (not shown) built in the speech device 10 after downloading. ) And may be displayed on the display unit 15 by reading the same from a storage unit (not shown). In addition, the display unit 15 may display an image stored in advance in a storage unit (not shown) by reading an image according to the set mode.

  When the display unit 15 includes the light emitting unit, the display unit 15 may cause the light emitting unit to emit a color according to the set mode. For example, it is blue in the case of Tokyo valve mode, red in the case of fine mode, green in the case of polite mode, and so on. In addition, the display unit 15 may blink the light emitting unit according to the utterance.

(Configuration of server 20)
As shown in FIG. 1, the server 20 includes a server communication unit (voice output unit) 21, a control unit 22, a storage unit 23, and an external information acquisition unit (voice information acquisition unit) 24.

  The server communication unit 21 is connected to the speech apparatus 10 and can transmit and receive data to and from the speech apparatus 10.

The control unit 22 centrally controls each component of the server 20. The function of the control unit 22 is, for example, CPU (Central Processing) for a program stored in a storage device such as a random access memory (RAM) or a flash memory.
Unit) is implemented. Further, as shown in FIG. 1, the control unit 22 also functions as a voice information recognition unit (recognition means) 25 and an utterance content determination unit (utterance content determination means, determination means) 26.

  The storage unit 23 stores various databases such as a database including a recognition phrase and an answer phrase corresponding to the recognition phrase, or a vague database including an ambiguous answer phrase. The details of each database will be described later.

  The external information acquisition unit 24 is connected to the network 2 and can acquire external information from the network 2 according to an instruction from the control unit 22.

(Configuration of control unit 22)
As described above, the control unit 22 also functions as the voice information recognition unit 25 and the utterance content determination unit 26.

  The voice information recognition unit 25 recognizes the acquired input voice information and outputs it as recognition information. Here, the speech information recognition unit 25 recognizing the input speech information means that the speech information recognition unit 25 discriminates a phrase and intonation included in the input speech information. For example, when the input voice information includes a flat "Ohayou" without accents, the voice information recognition unit 25 determines that the phrase included in the acquired input voice information is "Good morning" and the intonation is a flat without accents. To determine. Then, the voice information recognition unit 25 outputs recognition information indicating that the recognition phrase is "Good morning" and the recognition intonation is "Flat without accent."

  The utterance content determination unit 26 determines the utterance content to be uttered by the utterance device 10 according to the set mode and the recognition information output from the voice information recognition unit 25. Details of the process in which the utterance content determination unit 26 determines the utterance content will be described later.

(Processing of control unit 22)
The processing of the control unit 22 will be described below with reference to FIG.

  FIG. 2 is an example of a database which the utterance content determination unit 26 refers to in order to determine the utterance content in the utterance system 1 according to the first embodiment of the present invention.

  As shown in FIG. 2, each database contains a recognition phrase and a corresponding answer phrase and intonation. In addition, each database includes the recognition phrase and the answer phrase included in the same category. For example, as phrases included in the category of "Morning greeting", "Good morning" is included in the recognition phrases of Database 1 and Database 2 and "Good morning" is included in the recognition phrase of Database 3. In addition, as shown in FIG. 2, the databases 1 to 3 respectively include "I'm here", "I'm now" and "I just returned" as phrases included in the category "Greetings at home" "Good night" and "Good night" are respectively included as a phrase included in the category "Guest at bedtime". Each database corresponds to the mode that the server 20 has. For example, it is assumed that the database 1 corresponds to the standard mode, the database 2 corresponds to the fine mode, and the database 3 corresponds to the polite mode. In the present embodiment, all intonations included in the databases 1 to 3 are flat intonation without accents.

  Description will be made of processing performed by the control unit 22 when the database shown in FIG. 2 is stored in the storage unit 24 and the user issues "Ohayou" with a flat intonation without accent to the speech apparatus 10. .

  First, the voice information acquisition unit 12 outputs input voice information including a flat intonation “oh you” without accents to the server 20 via the communication unit 11. The control unit 22 of the server 20 acquires input voice information via the server communication unit 21. Then, the voice information recognition unit 25 recognizes the acquired input voice information. In this case, the voice information recognition unit 25 outputs, to the utterance content determination unit 26, recognition information indicating the recognition phrase "Ohayou" and the recognition intonation "flat without accent".

  The utterance content determination unit 26 determines the utterance content to be uttered by the utterance device 10 from the acquired recognition information.

  For example, when the standard mode is set in the server 20, the utterance content determination unit 26 refers to the database 1 corresponding to the standard mode, and selects the answer phrase "Ohayou" corresponding to the recognition phrase "Ohayou". Then, the utterance content determination unit 26 determines the selected phrase "Good morning" and the flat intonation without accent as the utterance content.

  Further, for example, when the fine mode is set in the server 20, the utterance content determination unit 26 switches the database to be referred to to the database 2 corresponding to the fine mode, and the answer phrase "today" corresponding to the recognition phrase "Ohayou" Let's take it easy! " Then, the utterance content determination unit 26 determines, as the utterance content, the selected phrase “Let's get in touch with you today” and a flat intonation without an accent.

  Then, the utterance content determination unit 26 outputs the determined utterance content to the utterance device 10 via the server communication unit 21. The utterance unit 13 of the utterance device 10 acquires the utterance content via the communication unit 11, and utters the phrase included in the acquired utterance content in the intonation included in the utterance content.

  Further, for example, when the polite mode is set in the server 20, the utterance content determination unit 26 switches the database to be referred to to the database 3 corresponding to the polite mode. Here, in the database 3, the recognition phrase included in the category “Morning greetings” is “Good morning”, which is different from the recognition phrase “Good morning” acquired by the speech content determination unit 26 from the voice information recognition unit 25. Therefore, the utterance content determination unit 26 does not select the answer phrase. Therefore, the server 20 does not make any speech from the speech device 10 because the speech content is not determined.

  Note that the server 20 may be configured to refer to a phrase database that does not include a recognition phrase. In this case, the utterance content determination unit 26 can determine the utterance content by switching the phrase database with reference to the phrase database.

  Further, in this case, the server 20 may be configured to refer to a recognition database that is a recognition database associated with a phrase database and that contains a recognition phrase and does not contain an answer phrase. In the configuration, the voice information recognition unit 25 can switch the recognition database according to the phrase included in the input voice information with reference to the recognition database, and can recognize the input voice information.

  Subsequently, processing for setting intonation of a phrase to be uttered according to the mode set in the server will be described.

  First, the storage unit 24 stores the database 10 having the same recognition phrase and response phrase as the database 1 of FIG. 2 and in which each phrase of the response phrase includes an intonation different from the response phrase of the database 1. In the present embodiment, it is assumed that the database 10 includes an answer phrase of intonation in which “Yo” of “Good morning” is accented with respect to the recognition phrase “Good morning”. Then, the database 10 is made to correspond to the Kansai dialect mode. As a method of accenting a specific sound, for example, the answer phrase may be divided into characters one by one and accented on a specific character. In addition, voices actually produced by people may be prerecorded and voices may be used.

  A process performed by the utterance content determination unit 26 when the user issues "Good morning" with a flat intonation without an accent to the utterance device 10 will be described.

  First, the utterance content determination unit 26 acquires, from the voice information recognition unit 25, recognition information indicating the recognition phrase "Ohayou" and the recognition intonation "flat without accent". Then, the utterance content determination unit 26 determines the utterance content to be uttered by the utterance device 10 from the acquired recognition information.

  For example, when the Kansai dialect mode is set in the server 20, the utterance content determination unit 26 refers to the database 10 corresponding to the Kansai dialect mode, and selects the answer phrase "good morning" corresponding to the recognition phrase "good morning". . In addition, the utterance content determination unit 26 sets the intonation of “good morning” as an intonation for accenting “Y” which is the intonation of the answer phrase in the database 10. Then, the utterance content determination unit 26 determines, as the utterance content, the selected phrase "Good morning" and the intonation for accenting "Yo".

  The server 20 may refer to a plurality of phrase databases in which different intonations are set for one or a plurality of answer phrases. In such a configuration, the utterance content determination unit 26 determines the utterance content by switching the phrase database to be referred to.

  Further, in this case, the server 20 may be configured to refer to a recognition database that is a recognition database associated with the phrase database and that includes a recognition phrase for which intonation is set and does not include an answer phrase. In the configuration, the voice information recognition unit 25 can switch the recognition database according to the phrase included in the input voice information by referring to the recognition database, and can recognize the input voice information.

  As described above, in the speech system 1 according to the first embodiment, "Good morning" or "Let's get excited today" can be uttered from the speech device 10 with respect to "Good morning" from the user. In addition, it is possible to cause the speech device 10 to utter "Good morning" with a flat intonation without accent and "Good morning" with intonation accented with "yo" with respect to "Good morning" from the user. Therefore, the speech system 1 performs selection processing for selecting a phrase to be uttered from a plurality of phrases ("Ohayou" and "I will also try today!") Included in the same category ("Amorning Greeting" category). By performing at least one of the processing for setting the intonation of the phrase to be uttered, and the processing for setting the intonation of the phrase to be uttered, it is possible to cause the uttering device 10 to utter the uttered contents tailored to the user. Smooth communication can be achieved. An example of the recognition phrase and the answer phrase is shown in FIG. FIG. 3 is an example of a recognition phrase and an answer phrase in the speech system according to the first embodiment of the present invention.

  The control unit 22 and the storage unit 23 included in the server 20 may be included in the speech apparatus 10. In this case, at least one of selection processing for selecting a phrase to be uttered from the same category and setting processing for setting intonation of the phrase to be uttered in the utterance content determination unit 26 of the utterance device 10. Determine the content of the utterance by Then, the utterance unit 13 obtains the utterance content, and utters a phrase included in the utterance content with intonation included in the utterance content. Therefore, since the speech apparatus 10 can utter the speech contents tailored to the user, smooth communication can be achieved with the user.

Second Embodiment
In the first embodiment, the speech system 1 selects a phrase to be uttered by the speech device 10 or sets an intonation of a phrase to be uttered by the speech device 10 according to the phrase contained in the input speech information. According to the above, the phrase to be uttered by the speech device 10 may be selected, or the intonation of the phrase to be uttered by the speech device 10 may be set.

  First, storage unit 23 includes database 1 shown in FIG. 2, database 2 and database 3 together with recognition phrases of database 3 including flat intonation and database 4 including intonation with accents in specific sounds of each recognition phrase of database 1. The case where is memorized is explained. In this embodiment, the intonation of the database 4 is assumed to be an intonation accented with “Yo” of “Ohayou”, and the mode corresponding to each database is the same as that of the first embodiment. Then, a case where the input voice information includes an intonation in which “Yo” of “Good morning” is accented will be described.

  First, the voice information recognition unit 25 recognizes input voice information, and utters recognition information indicating that the recognition phrase is "Toyo" and the recognition intonation accents "Yo" of "Ohayou". The content is output to the content determination unit 26.

  The utterance content determination unit 26 determines the utterance content to be uttered by the utterance device 10 from the acquired recognition information.

  For example, when the Kansai dialect mode is set in the server 20, the utterance content determination unit 26 refers to the database 4 corresponding to the Kansai dialect mode and acquires the recognition phrase of the database 4 and the intonation of the recognition phrase. It is determined whether or not it matches the recognition information. In the present embodiment, since the recognition phrase of the database 4 and the intonation of the recognition phrase coincide with the acquired recognition information, the utterance content determination unit 26 recognizes the phrase “Ohayou” corresponding to the recognition information, and “Y”. The intonation to accentuate is determined as the utterance content.

  Further, for example, when the standard mode is set in the server 20, the utterance content determination unit 26 refers to the database 1 corresponding to the standard mode, and the recognition phrase of the database 1 and the intonation of the recognition phrase are acquired. It is determined whether or not it matches the recognition information. In the present embodiment, since the intonation of the recognition phrase of the database 4 does not match the intonation of the recognition information, the speech content determination unit 26 does not determine the speech content.

  As described above, in the speech system 1 according to the second embodiment, the contents of speech can be determined according to intonation included in the input speech. Therefore, since the speech system 1 can cause the speech device 10 to utter the speech contents according to the intonation of the user, smoother communication with the user can be achieved as compared with the conventional case.

Third Embodiment
In the embodiment described above, the speech system 1 utters the speech content corresponding to the recognized input speech information from the speech device 10. However, the speech content determination unit 26 may be configured not to acquire the input speech information. The controller may execute a predetermined process when a predetermined condition is satisfied.

  For example, if the predetermined condition is "7 o'clock in the morning" and the predetermined process is "to deliver a greeting in the morning", the utterance content determination unit 26 proceeds to 7 o'clock in the morning. Select the phrase included in the greeting and decide the content of the utterance. For example, when the standard mode is set in the server 20, the utterance content determination unit 26 refers to the database 1 corresponding to the standard mode at 7 o'clock in the morning, and the phrase "Good morning" which is a morning greeting, A flat intonation without accents is determined as the utterance content. Further, for example, when the polite mode is set in the server 20, the utterance content determination unit 26 switches the database to be referred to the database 3 corresponding to the polite mode at 7 o'clock in the morning. Then, the utterance content determination unit 26 determines, as the utterance content, a phrase "Good morning" which is a morning greeting and a flat intonation without an accent.

  Thus, the speech system 1 according to the third embodiment can execute a predetermined process when a preset condition is satisfied. Therefore, the speech system 1 can communicate more smoothly with the user as compared with the related art because the speech device 10 speaks itself without acquiring the input speech information from the user.

  Further, in the third embodiment, the recognition phrase and the answer phrase do not need to correspond to each other, and the configuration may be such that a database in which the recognition phrase and the answer phrase are separated is referred to. In such a configuration, a plurality of phrase databases not including a recognition phrase are stored in the storage unit 23, and the utterance content determination unit 26 can determine the utterance content by switching the phrase databases.

Embodiment 4
As described in the third embodiment, the database stored in the storage unit 23 may be a database in which the recognition phrase and the answer phrase are separated. Therefore, the voice information recognition unit 25 may be configured to recognize the input voice information by switching the recognition database not including the answer phrase.

  For example, the case where the storage unit 23 stores the recognition database 1, the recognition database 2, and the recognition database 3 which do not include the answer phrases of the database 1, the database 2, and the database 3 shown in FIG. 2 will be described. In this case, when the input speech information includes flat intonation “Good morning”, the speech information recognition unit 25 switches the recognition database so that the phrases included in the input speech information are recognized database 1 and recognition database 2. Unlike the recognition phrase, it can be recognized that it matches the recognition phrase of the recognition database 3.

  Further, as described in the second embodiment, the recognition database may be switched by the intonation by including the intonation together with the recognition phrase in the recognition database.

  For example, a recognition database 4 including a flat intonation together with the recognition phrases of the recognition database 1, the recognition database 2 and the recognition database 3 described above, and an intonation in which a specific sound of each recognition phrase in the database 1 is accented The case where the storage unit 23 stores data will be described. In the present embodiment, the intonation of the database 4 is an intonation in which “Yo” of “Good morning” is accented.

  In this case, when the input voice information includes intonation in which "Oha" of "Ohayou" is accented, the voice information recognition unit 25 switches the recognition database to recognize the recognition database 1, the recognition database 2, and the recognition. Unlike the intonation of the recognition phrase of the database 3, it can be recognized that it matches with the intonation of the recognition phrase of the recognition database 4.

Fifth Embodiment
In the embodiment described above, when the recognition phrase that matches the acquired recognition phrase is included in the referenced database, or when the acquired recognition intonation is included in the database that is referred to, the utterance content determination unit 26 recognizes A phrase included in a category that matches a category including a phrase is determined as the utterance content. On the other hand, at least one of the case where the recognition phrase that matches the acquired recognition phrase is not included in the referenced database and the case where the intonation that matches the acquired recognized intonation is not included in the database. The utterance content determination unit 26 does not determine the utterance content. However, if the speech device 10 does not respond to the phrase issued by the user, the user may not acquire the phrase issued by the user as speech information, or may respond to the phrase issued by the user. It can not be judged whether there is no answer or failure. So, in this embodiment, even when the answer corresponding to the phrase which a user uttered does not exist in the database to refer to, the processing in which the utterance content determination unit 26 determines the utterance content will be described using FIG.

  FIG. 4 is an example of a fuzzy database which the speech content determination unit 26 refers to in order to determine the speech content in the speech system 1 according to the fifth embodiment of the present invention. The ambiguous database is a database including ambiguous phrases (hereinafter also referred to as ambiguous phrases). Here, the ambiguous phrase can also be expressed as a phrase included in a category different from the category of the phrase included in the input speech information. In other words, the ambiguous phrase may be expressed as a phrase included in input speech information and a phrase included in a category different from the category of the phrase selected when the phrase and the intonation match the predetermined phrase and the intonation. it can. In addition, the ambiguous phrase can also be expressed as a phrase that implies that the input speech information can not be recognized or has been recognized but there is no corresponding answer phrase.

  In the present embodiment, the database 4 corresponds to the standard mode, and the database 5 corresponds to the fine mode. An example of the vague phrase is shown in FIG. FIG. 5 is an example of the ambiguous phrase in the speech system 1 according to the fifth embodiment of the present invention.

  Subsequently, an example in which the utterance content determination unit 26 refers to the fuzzy database when the recognition phrase that matches the acquired recognition phrase is not included in the referred database will be described.

  First, the voice information recognition unit 25 recognizes input voice information, and outputs the recognition information to the utterance content determination unit 26. The utterance content determination unit 26 determines the utterance content based on the recognition information according to the mode set in the server 20. Here, when the recognition phrase included in the recognition information does not match the recognition phrase of the database to be referred to, the utterance content determination unit 26 refers to the fuzzy database and selects the fuzzy phrase.

  For example, when the standard mode is set in the server 20 and the input voice information includes the phrase "Good morning", the utterance content determination unit 26 first refers to the database 1 and "Good morning". Select the corresponding answer phrase. Here, since there is no answer phrase corresponding to "Good morning" in the database 1, subsequently, referring to the database 4, "un un" is selected as the answer phrase. In addition, when selecting an answer phrase from the database 4, the utterance content determination unit 26 selects a predetermined condition (for example, selecting an answer phrase included in the database 4 in order from the top). You may choose an answer phrase randomly. By selecting an answer phrase at random, it is possible to give the user a more natural impression of communication.

  Further, for example, when the fine mode is set in the server 20 and the input voice information includes the phrase "Good morning", the utterance content determination unit 26 first refers to the database 2, "Good morning. Select the answer phrase corresponding to Here, since there is no answer phrase corresponding to "Good morning" in the database 2, subsequently, the database 5 is referred to and "Good thing likely" is selected as the answer phrase.

  Thus, in the speech system 1 according to the fifth embodiment, when the phrase included in the input speech information from the user does not match the predetermined phrase (when there is no answer phrase corresponding to the phrase included in the speech information), the speech The device 10 utters a vague phrase. Therefore, since the user can talk with the speech device as if he / she is talking to a person, the speech system 1 can communicate more smoothly with the user.

  Next, even if the utterance content determination unit 26 includes a recognition phrase that matches the acquired recognition phrase in the referenced database, if the acquired recognition intonation does not match the referenced database, An example to be referred to will be described.

  First, the voice information recognition unit 25 recognizes input voice information, and outputs the recognition information to the utterance content determination unit 26. The utterance content determination unit 26 determines the utterance content based on the recognition information according to the mode set in the server 20. Here, when the recognition phrase included in the recognition information matches the recognition phrase of the referenced database, but the recognition intonation does not match the intonation set in the recognition phrase of the referenced database, the utterance content determination unit 26 Refers to the fuzzy database and selects a fuzzy phrase.

  For example, when the standard mode is set in the server 20 and the input voice information includes intonation accented with "Y" of the phrase "Ohayou", the utterance content determination unit 26 recognizes the recognition phrases "Ohayou" and " The answer phrase corresponding to the recognition intonation accented with "Y" is selected from the database 1. Here, although there is an answer phrase "Good morning" corresponding to the recognition phrase "Good morning" in the database 1, there is no answer phrase corresponding to the recognition intonation accented with "Good morning" and "Yo", so the utterance content The determination unit 26 refers to the database 4 and selects “say once more” as the answer phrase. Note that, as described above, the utterance content determination unit 26 selects predetermined conditions (for example, sequentially selects the answer phrases included in the database 4 from the top, etc.) when selecting the answer phrases from the database 4. It may be selected based on it, or the answer phrase may be selected at random. By selecting an answer phrase at random, it is possible to give the user a more natural impression of communication.

  Further, for example, when the fine mode is set in the server 20 and the input voice information includes intonation accented with “Yo” of the phrase “Good morning”, the utterance content determination unit 26 recognizes the recognition phrase “Good morning” and The response phrase corresponding to the recognition intonation accented with "Y" is selected from the database 2. Here, database 2 has an answer phrase "I will try to compliment you today!" Corresponding to the recognition phrase "Good morning", but an answer phrase corresponding to a recognition intonation accented with "Good morning" and "Yo". Since the utterance content determination unit 26 refers to the database 5, the utterance content determination unit 26 selects "voice is small!" As an answer phrase.

  Thus, in the speech system 1 according to the fifth embodiment, the phrase included in the input voice information from the user matches the predetermined phrase (when there is a phrase corresponding to the phrase included in the voice information), Also, when the speech does not match the predetermined intonation, the speech device 10 utters the vague phrase. Therefore, since the user can talk with the speech device as if he / she is talking to a person, the speech system 1 can communicate more smoothly with the user.

Sixth Embodiment
In the embodiment described above, the server 20 causes the utterance device 10 to utter the answer phrase corresponding to the input voice information, but acquires external information based on the input voice information, and utters the answer phrase corresponding to the external information It is good also as composition which makes 10 speak. Sixth Embodiment A configuration for acquiring external information will be described with reference to FIG.

  FIG. 6 is an example of a database which the utterance content determination unit 26 refers to in order to determine the utterance content in the utterance system 1 according to the sixth embodiment of the present invention. The database 6 of FIG. 6 includes an instruction corresponding to the recognition phrase. The instruction is a process performed by the utterance content determination unit 26. For example, the database 6 of FIG. 6 includes an instruction “to acquire weather information from the network”. Subsequently, the database 7 and the database 8 include the response phrase corresponding to the external information. The external information is information acquired by the utterance content determination unit 26 via the external information acquisition unit 24. For example, in the databases 7 and 8 of FIG. 6, weather information on weather such as "fine" and "rain" is included as external information. In the present embodiment, the database 7 corresponds to the standard mode, and the database 8 corresponds to the fine mode, and a case where the input voice information is the phrase "What is the weather today?" Will be described as an example.

  First, the voice information recognition unit 25 recognizes the input voice information, and outputs recognition information indicating that the recognition phrase is "What is the weather today?" To the utterance content determination unit 26. The utterance content determination unit 26 switches the database to be referred to, and executes an instruction “acquire weather information from the network” corresponding to the acquired recognition information. Then, the utterance content determination unit 26 acquires weather information via the external information acquisition unit 24.

  Subsequently, the utterance content determination unit 26 selects an answer phrase corresponding to the acquired weather information.

  For example, when the acquired weather information is "fine" and the standard mode is set in the server 20, the utterance content determination unit 26 refers to the database 7 corresponding to the standard mode, and the phrase "fine" is obtained. Are determined as the utterance content.

  Also, for example, when the acquired weather information is "rain" and the fine mode is set in the server 20, the utterance content determination unit 26 refers to the database 8 corresponding to the fine mode, and "rainy! The phrase "" is determined as the utterance content.

  As described above, in the speech system 1 according to the sixth embodiment, the mode set in the external information and the server 20 is provided by providing the database including the instruction corresponding to the input voice information and the answer phrase corresponding to the external information. It is possible to cause the utterance device 10 to utter the corresponding utterance content. Therefore, the speech system 1 can cause the speech device 10 to utter speech contents according to the external information acquired in real time.

Seventh Embodiment
In the present embodiment, a process of changing the speech mode set in the server 20 by the user's voice by the server 20 will be described using FIG. 7. In the present embodiment, the utterance unit (confirmation guidance phrase utterance unit) 13 acquires the utterance content transmitted from the server 20 via the communication unit (confirmation guidance phrase reception unit) 11 of the utterance device 10, and the utterance is made The process in which the speech unit 13 utters the contents, and the process in which the voice information acquisition unit (answer reception means) 12 obtains voice information and transmits it to the server 20 via the communication unit (answer transmission means) 11 are described above. The explanation has been omitted here since the explanation has already been made in the embodiment of.

  FIG. 7 is a flowchart showing a flow of a process of changing the mode set in the server 20 in the speech system 1 according to Embodiment 7 of the present invention. When the server 20 acquires, from the speech device 10, operation information indicating that the mode is to be changed, the utterance content determination unit 26 confirms with the user whether or not to change the mode. Change it to, and so on. " Then, the utterance content determination unit 26 outputs the utterance content including the selected confirmation guidance phrase to the utterance device 10 via the server communication unit 21 (confirmation guidance phrase transmission unit) (step S1).

  When there are a plurality of speech devices 10 in the server 20, information for specifying a speech device to be a target of mode change may be included in operation information indicating that the mode is to be changed. In addition, the server 20 may obtain operation information indicating that the mode is to be changed, via an input device connected to the server 20. In addition, the speech device 10 may transmit the operation information received by the operation reception unit 14 of the speech device 10 to the server 20, so that the server 20 may acquire the operation information indicating that the mode is to be changed.

  Further, the mode may be changed according to the sound emitted by the user to the speech device 10. More specifically, a characteristic part of the character name corresponding to each mode (for example, "Ai-chan", "Nao-chan", etc.) or all of the character name (for example When the voice information acquisition unit 12 of the speech device 10 acquires voice information including “Lanly Nao-chan” and the like and the server 20 receives the acquired voice information, the server 20 changes the set mode. May be In this case, the speech apparatus 10 may transmit the speech information acquired by the speech information acquisition unit 12 to the server 20 as it is, and the server 20 may specify the corresponding mode. The speech apparatus 10 may identify each mode and character name. And the speech apparatus 10 may specify a mode corresponding to the voice information acquired by the speech information acquisition unit 12 and transmit mode information indicating the specified mode to the server 20.

  In step S1, the utterance content determination unit 25 may select a confirmation guidance phrase using a character name corresponding to the mode as the confirmation guidance phrase. As an example of the confirmation guidance phrase using the character name, there is "Becoming a healthy eye. Good?" Furthermore, when uttering the confirmation guidance phrase using the character name, the speech device 10 may display the character image of “Genki Ai-chan” on the display panel of the display unit 15. Then, the utterance content determination unit 26 starts the timer together with the output of the utterance content (step S2).

  Next, the utterance content determination unit 26 determines whether or not the started timer has passed a predetermined time and a timeout has occurred (step S3). Here, the predetermined time is a time when the user's answer to the output confirmation guidance phrase is received.

  If it is determined in step S3 that "timeout has occurred" (step S3: YES), the utterance content determination unit 26 indicates a timeout guidance phrase indicating that the process of changing the mode is to be canceled (for example, , "The mode change has been canceled due to time out", etc.). Then, the utterance content determination unit 26 outputs the utterance content including the selected timeout guidance phrase to the utterance device 10 via the server communication unit 21 (step S4).

  On the other hand, when it is determined in step S3 that "timeout has not occurred" (step S3: NO), the voice information recognition unit (acquisition means) 25 receives the response via the server communication unit (answer acceptance means) 21. It is determined whether voice information which is an answer to the confirmation guidance phrase has been acquired (step S5).

  If it is determined in step S5 that "the input speech information has not been acquired" (step S5: NO), the processing of the utterance content determination unit 26 returns to step S3 of determining whether or not a timeout has occurred.

  On the other hand, if it is determined in step S5 that "input voice information has been acquired" (step S5: YES), the voice information recognition unit 25 recognizes input voice information which is a response from the user (step S6). Then, the utterance content determination unit (mode change determination means) 26 determines whether or not to change the mode in accordance with the phrase recognized by the voice information recognition unit 25.

  If it is determined in step S6 that "the acquired voice information is a confirmation phrase (for example," good "etc.) indicating that the user approves the change" (step S6: confirmation phrase), the utterance content determination The unit 26 decides to change the mode. Then, the utterance content determination unit 26 changes the set mode, and selects a confirmation completion phrase (for example, “mode changed”) indicating that the mode change is completed. Then, the utterance content determination unit 26 outputs the utterance content including the selected confirmation completion phrase to the utterance device 10 via the server communication unit 21 (step S7).

  If it is determined in step S6 that "the acquired voice information is a wrong phrase (for example," good morning "etc.) different from the confirmation phrase" (step S6: false phrase), the utterance content determination unit 26 It is determined that the mode is not changed, and it is determined whether the number of times the wrong phrase has been acquired is the third (step S8).

  If it is determined in step S8 that "the number of times the wrong phrase has been acquired is not the third time" (step S8: NO), the utterance content determination unit 26 prompts the user to issue the confirmation phrase again. Select a reconfirmation guidance phrase (eg, "Please say one more time," etc.) indicating. Then, the utterance content determination unit 26 outputs the utterance content including the selected reconfirmation guide phrase to the utterance device 10 via the server communication unit 21 (step S9). Then, the utterance content determination unit 26 returns to step S2 to start the timer again from the beginning.

  On the other hand, when it is determined in step S8 that "the number of times the wrong phrase has been acquired is the third time" (step S8: YES), the processing of the utterance content determination unit 26 proceeds to step S10 to end the mode change. .

  When it is determined in step S6 that "the acquired voice information is an end phrase (for example," stop "etc.) indicating that the user cancels the mode change (step S6: end phrase) The utterance content determination unit 26 determines not to change the mode, and selects an end guidance phrase (for example, “mode change has been canceled” or the like) indicating that the process of changing the mode is ended. Then, the utterance content determination unit 26 outputs the utterance content including the selected end guidance phrase to the utterance device 10 via the server communication unit 21 (step S10).

  When it is determined in step S6 that "the acquired voice information is noise" (step S5: noise), the processing of the utterance content determination unit 26 determines not to change the mode, and a timeout occurs. It returns to step S3 to determine whether or not it has. In addition, as a method of determining whether audio | voice information is noise, the method of determining by whether the magnitude | size of the sound contained in input audio | voice information is contained in the predetermined | prescribed range, for example is mentioned. Here, if the range not included in the loudness of the sound emitted by the user in the conversation is taken as an example of the predetermined range, the loudness of the sound contained in the voice information is smaller than the loudness of the sound emitted by the user in the conversation Alternatively, when it is included in any of the magnitudes of the sounds emitted by the user in the conversation, the utterance content determination unit 26 determines that the input voice information is noise.

  As described above, in the speech system 1 according to the seventh embodiment, when the user changes the mode set in the server 20, after accepting the mode change operation from the user, it is further confirmed whether or not to change the mode. be able to. Therefore, it is possible to prevent the mode set in the server 20 from being erroneously changed. Further, in the speech system 1, the setting of the server 20 can be changed by the speech apparatus 10 uttering the confirmation guide phrase and the user utters the phrase. In other words, since the setting of the server 20 can be changed by the user speaking with the speech device 10, the speech system 1 can communicate smoothly with the user.

  In step S5 of FIG. 7, the speech content determination unit 26 executes the following processing based on the recognition phrase recognized by the speech information recognition unit 25. However, the speech content determination unit 26 performs the following processing based on the recognition phrase and recognition intonation. May perform the following processing.

  For example, in step S1, when the utterance content determination unit 26 acquires operation information indicating that the mode is to be changed, the database referred to for determining the utterance content is set as a database corresponding to the mode after the change. It may be More specifically, when the utterance content determination unit 26 acquires operation information indicating to change to the fine mode, the utterance content determination unit 26 determines, as a confirmation phrase, a database to be referred to for determining the utterance content. The database is switched to the database corresponding to the energy mode including "OK!" And "Regards!" (Step S1). In addition, when the information for specifying the speech apparatus used as the object of mode change is contained in operation information, a database is switched about the speech of the speech apparatus used as the object of mode change.

  Next, steps S2 to S4 executed by the utterance content determination unit 26 and step S5 executed by the voice information recognition unit 25 have already been described, and thus are omitted.

  Subsequently, in step S6, when the voice information recognition unit 25 obtains voice information including the phrase "Good" from the user, it is not a confirmation phrase corresponding to the fine mode, and therefore the utterance content decision unit 26 proceeds to step S8.

  As described above, in the speech system 1 according to the seventh embodiment, in the mode after the change, whether or not the user can utter the recognition phrase corresponding to the mode after the change as the input voice information is confirmed before changing the mode. can do.

  In addition, in the speech system 1 according to the seventh embodiment, when the server 20 acquires input voice information under a predetermined condition, the mode set in the server 20 may be changed. As a predetermined condition, for example, when the server acquires input sound information whose sound volume is larger than a predetermined value continuously a plurality of times (for example, three times), the server 20 may change to the energy mode. When the user is emitting input voice information larger than a predetermined volume, it can be determined that the user is fine, so when the server 20 changes to the fine mode, the user can have a pleasant conversation with the speech device 10.

  In addition, in the speech system 1 according to the seventh embodiment, the speech device 10 may execute the processing of FIG. 7 by including the control unit 22 and the storage unit 23 in the speech device 10. In this case, the utterance unit (confirmation guidance phrase utterance unit) 13 utters the confirmation guidance phrase determined in the utterance content determination unit (determination unit) 26 of the utterance device. Then, the voice information acquisition unit (answer reception means) 12 obtains input voice information which is a response to the confirmation guide phrase, and outputs the obtained input voice information to the voice information recognition unit 25. Then, according to the phrase recognized by the voice information recognition unit 25, the utterance content determination unit 26 determines whether to change the mode. Therefore, the speech apparatus 10 can achieve smooth communication with the user.

[Embodiment 8]
The uttered content determination unit 26 is a change unit (change means) capable of changing the voice color, the volume, the speech speed and the sound pitch of the determined uttered content in addition to the phrase and intonation according to the mode set in the server 20. It is good also as composition provided.

  For example, when the fine mode is set in the server 20, the utterance content determination unit 26 changes the voice color to the fine voice color in the change unit, and the volume is 1.2 times the standard mode and the speech speed is 1 in the standard mode. You may change it by three times.

  Also, for example, when the polite mode is set to the server 20, the utterance content determination unit 26 changes the voice color to a polite voice color in the change unit, and the volume is 0.9 times the standard mode and the speech speed is the standard mode The pitch may be changed to 0.8 times that of the standard mode and 1.2 times that of the standard mode.

  Thus, in the speech system 1 according to the eighth embodiment, the voice color, the volume, the speech speed, and the pitch of the determined speech contents can be changed according to the mode set in the server 20. Therefore, it is possible to realize the speech device 10 which is full of human touch.

[Embodiment 9]
The control unit 22 of the server 20 may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software using a CPU (Central Processing Unit).

  In the latter case, the server 20 is a CPU that executes instructions of a program that is software that implements each function, a ROM (Read Only Memory) or a storage in which the program and various data are readably recorded by a computer (or CPU). A device (these are referred to as “recording media”), a RAM (Random Access Memory) for developing the program, and the like are provided. The object of the present invention is achieved by the computer (or CPU) reading the program from the recording medium and executing the program. As the recording medium, a “non-transitory tangible medium”, for example, a tape, a disk, a card, a semiconductor memory, a programmable logic circuit or the like can be used. The program may be supplied to the computer via any transmission medium (communication network, broadcast wave, etc.) capable of transmitting the program. The present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

[Summary]
The speech control device (20) according to aspect 1 of the present invention is a speech control device that determines the contents of speech to be made to speech in the speech device (10), and selects a phrase to be uttered from a plurality of phrases included in the same category. An utterance content determination unit (utterance content determination unit 26) that determines the utterance content by performing at least one of selection processing and setting processing for setting intonation of a phrase to be uttered.

  According to the above configuration, the speech control apparatus performs at least one of the process of selecting the phrase to be uttered from a plurality of phrases included in the same category or setting the intonation of the phrase to be uttered. The content of the utterance to be uttered to the uttering device is determined. Therefore, since the speech control device can make the speech device utter the speech contents tailored to the user, it is possible to achieve smooth communication with the user as compared with the conventional case.

  The speech control apparatus according to aspect 2 of the present invention is configured to refer to a plurality of phrase databases including one or a plurality of phrases in the above-mentioned aspect 1, and the speech content determination means refers to the phrase database to be referred to. The content of the utterance may be determined by switching.

  According to the above configuration, the speech control device refers to the plurality of phrase databases and determines the contents of the speech by switching the phrase databases, so that speech control can be made to communicate with the user more smoothly than before. The device can be easily realized.

  The speech control apparatus according to aspect 3 of the present invention is configured to refer to a plurality of phrase databases in which different intonations are set for one or a plurality of phrases in the above aspect 1 or 2, and the speech The content determination means may determine the utterance content by switching the phrase database to be referred to.

  According to the above configuration, the speech control device refers to a plurality of phrase databases in which different intonations are set for one or a plurality of phrases, and determines a phrase for which the intonation is set as the speech content. Therefore, it is possible to easily realize a speech control device capable of achieving smooth communication with the user.

  The speech control apparatus according to aspect 4 of the present invention may be the recognition database corresponding to the phrase database in the above aspect 2 or 3, and may refer to a recognition database for recognizing input speech information.

  According to the above configuration, the speech control device can recognize the input speech information by referring to the recognition database corresponding to the phrase database.

  In the speech control device according to aspect 5 of the present invention, in the above aspect 4, a recognition unit (speech information recognition unit 25) for switching the recognition database according to at least one of a phrase included in input speech information and intonation. The speech content determination means may further determine the speech content according to at least one of the phrase and the intonation recognized by the recognition means.

  According to the above configuration, the speech control device causes the speech device to utter the speech content determined according to at least one of the phrase and the intonation included in the input speech information recognized by the recognition means. Therefore, since the user can talk to the speech apparatus as if he / she is talking to a person, the speech control apparatus can communicate more smoothly with the user than before.

  The speech control apparatus according to the sixth aspect of the present invention is the speech control device according to the second to fifth aspects, wherein the judgment means (the speech content determination unit 26) judges whether the phrase included in the input speech information matches the predetermined phrase. Furthermore, when it is determined by the determination means that the phrase included in the input voice information does not match the predetermined phrase, the utterance content determination means determines that the phrase included in the input voice information is a predetermined phrase A phrase included in a category different from the category of the phrase selected in the case of matching may be selected as the phrase to be uttered.

  According to the above configuration, when it is determined that the phrase included in the input voice information does not match the predetermined phrase, the utterance control device determines that the phrase included in the input voice information matches the predetermined phrase. A phrase included in a category different from the category of the phrase selected in can be uttered from the speaking device. Therefore, since the user can talk with the speech device as if he / she is talking to a person, the speech control device can communicate more smoothly with the user than in the past.

  The speech control apparatus according to aspect 7 of the present invention is the judging means for judging whether or not the phrase and intonation included in the input voice information match the predetermined phrase and intonation in the above-mentioned modes 3 to 5 (the speech content determination Section 26), and even if the phrase included in the input voice information matches the predetermined phrase by the determination means, the intonation included in the input voice information does not match the predetermined intonation When it is determined, the uttered content determination means sets, as the phrase to be uttered, a phrase included in a category different from the category of the phrase selected when the intonation included in the input voice information matches the predetermined intonation. You may choose.

  According to the above configuration, even if the phrase included in the input voice information matches the predetermined phrase, if the intonation included in the input voice information does not match the predetermined intonation, the predetermined phrase and intonation are included. The utterance control device can cause the utterance control device to utter a category different from the category of the phrase selected when there is a match. Therefore, since the user can talk with the speech device as if he / she is talking to a person, the speech control device can communicate more smoothly with the user than in the past.

  In the utterance control device according to an eighth aspect of the present invention, in the sixth or seventh aspect, the utterance content determination means randomly selects a phrase to be uttered from a database including a plurality of phrases included in the different categories in the selection process. You may choose to

  According to the above configuration, the speech control device can randomly select a plurality of phrases included in different categories and cause the speech device to speak. Therefore, the speech device does not answer only the same phrase, and the user can talk to the speech device as if he / she was talking to a person, so the speech control device is smoother with the user than in the past. Communication is possible.

  In the speech control apparatus according to aspect 9 of the present invention, in the above-mentioned aspects 1 to 8, the speech content determination means includes at least one of voice color, volume, speech speed and pitch of speech content to be made to speech to the speech apparatus It may further comprise changing means for changing.

  According to the above configuration, the utterance control device can change the voice color, the volume, the speech speed, and the pitch of the utterance content to be uttered to cause the utterance device to utter the utterance content. Therefore, it is possible to cause the speech apparatus to utter the speech content with the voice color, the volume, the speech speed, and the pitch according to the speech content.

  A method according to a tenth aspect of the present invention is a method of determining speech content to be made to be uttered in a speech device, comprising: selection processing for selecting a phrase to be uttered from a plurality of phrases included in the same category; The utterance content determination step of determining the utterance content by performing at least one of the setting processing of setting the intonation of the plurality of phrase databases including the plurality of phrases. By switching, the utterance content is determined.

  According to the above configuration, the method can realize an utterance device that achieves smooth communication with the user.

  A speech system according to aspect 11 of the present invention is a speech system including the speech control device according to aspect 5 and a speech device, wherein the speech device transmits input speech information to the speech control device, and performs speech control The utterance content received from the device is uttered.

  According to the above configuration, the speech system can realize the speech system having the same effect as the speech control device described in the fifth aspect.

  The speech control device according to each aspect of the present invention may be realized by a computer, and in this case, the speech control device is realized by the computer by operating the computer as each means included in the speech control device. A control program of a speech control device and a computer readable recording medium recording the same also fall within the scope of the present invention.

  The speech apparatus according to aspect 12 of the present invention performs at least one of a selection process of selecting a phrase to be uttered from a plurality of phrases included in the same category and a setting process of setting an intonation of a phrase to be uttered. And uttering means (uttering unit 13) for uttering the uttered contents determined by the uttered content determining means.

  According to the above configuration, it is possible to realize a speech apparatus that achieves the same effect as the speech control apparatus according to aspect 1.

  The speech control apparatus according to aspect 13 of the present invention transmits a confirmation guidance phrase for transmitting to the speech apparatus a confirmation guidance phrase for confirming to the user whether a phrase or a speech mode associated with the phrase and intonation is to be changed. Whether to change the mode according to the means (server communication unit 21), the answer accepting means (server communication unit 21) for accepting an answer from the user for the confirmation guide phrase, and the answer accepted by the answer accepting means And a mode change determination means (utterance content determination unit 26) for determining

  According to the above configuration, the speech control device can cause the user to utter a confirmation guidance phrase for confirming whether or not to change the mode. Therefore, since the speech control device can change the mode by talking with the user, it is possible to achieve smooth communication with the user.

  An utterance apparatus according to aspect 14 of the present invention is a confirmation guide phrase utterance unit (uttering unit for uttering a confirmation guide phrase for confirming to a user whether a phrase or an utterance mode associated with a phrase and intonation is to be changed 13) according to the answer acceptance means (voice information acquisition unit 12) for accepting an answer from the user to the confirmation guide phrase, and whether to change the speech mode according to the answer accepted by the answer acceptance means And a mode change determination unit (utterance content determination unit 26).

  According to the above configuration, it is possible to realize a speech apparatus that achieves the same effect as the speech control apparatus according to aspect 11.

  A speech system according to aspect 15 of the present invention is a speech system including the speech control device according to aspect 14 and a speech device, wherein the speech device receives the confirmation guide phrase from the speech control device. Confirmation guidance phrase receiving unit (communication unit 11), confirmation guidance phrase utterance unit (uttering unit 13) for uttering the confirmation guidance phrase, and answer acceptance unit (voice information acquisition for receiving an answer from the user for the confirmation guidance phrase And a response transmitting unit (communication unit 11) for transmitting the response received by the response receiving unit to the utterance control device.

  According to the above configuration, it is possible to realize a speech system that achieves the same effect as the speech control device according to aspect 11.

  A method according to aspect 16 of the present invention is a method of controlling the speech of the speech device, the confirmation guidance phrase for confirming with the user whether or not to change the speech mode associated with the phrase or the phrase and intonation. Whether to change the speech mode according to the confirmation guidance phrase transmission step of transmitting the speech to the speech apparatus, the answer reception step of receiving an answer from the user for the confirmation guide phrase, and the answer received in the answer reception step. And a decision process to decide.

  According to the above configuration, the method can realize an utterance device that achieves smooth communication with the user.

(Additional items)
In order to solve the above-mentioned subject, a speech control device concerning one mode of the present invention is a speech control device which determines the contents of speech to be made to speak in a speech device, and should speak from a plurality of phrases contained in the same category The utterance content determination means determines the above-mentioned utterance content by performing at least one of a selection process of selecting a phrase and a setting process of setting intonation of a phrase to be uttered.

  The present invention is not limited to the above-described embodiments, and various modifications can be made within the scope of the claims, and embodiments obtained by appropriately combining the technical means disclosed in the different embodiments. Is also included in the technical scope of the present invention. Furthermore, new technical features can be formed by combining the technical means disclosed in each embodiment.

  The present invention can be used in a system provided with a speech device for providing information to a user.

1 speech system 2 network 10 speech device 11 communication unit (speech content acquisition unit)
13 utterance part 20 server (utterance control device)
21 Server communication unit (voice output unit)
22 control unit 23 storage unit 24 external information acquisition unit (voice information acquisition unit)
25 speech information recognition unit 26 utterance content determination unit

Claims (7)

A speech control apparatus for determining the content of speech to be made to be uttered into a speech apparatus, comprising:
A voice information acquisition unit that obtains input voice information from the speech device;
A voice information recognition unit that recognizes the input voice information;
An utterance content determination unit that determines the utterance content by referring to one or more databases according to the mode set in the utterance control device and the recognition information recognized by the speech information recognition unit;
A voice output unit that outputs the utterance content determined by the utterance content determination unit to the utterance device;
Equipped with
The utterance content determination unit determines whether the recognition information corresponds to the mode, and does not determine the utterance content if the recognition information does not correspond to the mode.
An utterance control device characterized by the above. In the above database, the recognition phrase indicated by the recognition information and the response phrase corresponding to the recognition phrase are associated for each mode.
The utterance content determination unit selects the answer phrase by referring to the one or more databases according to the set mode.
The speech control apparatus according to claim 1, characterized in that: In the above database, the answer phrase associated with the recognition phrase indicated by the recognition information is different for each mode.
The speech control apparatus according to claim 2, characterized in that: The utterance content determination unit determines the utterance content according to the mode regardless of the setting of the utterance device.
The speech control apparatus according to claim 2 or 3, characterized in that: A method of determining an utterance content to be uttered by an utterance device, comprising:
A voice information acquisition step of acquiring input voice information from the speech device;
A voice information recognition step of recognizing the input voice information;
An utterance content determination step of determining the utterance content by referring to one or more databases according to the set mode and the recognition information recognized in the voice information recognition step.
An audio output step of outputting the utterance content determined in the utterance content determination step to the utterance device;
Including
In the utterance content determination step, it is determined whether or not the recognition information corresponds to the mode, and when the recognition information does not correspond to the mode, the utterance content is not determined.
A method characterized by A speech device,
A speech control device;
A speech system equipped with
The above speech control device
A voice information acquisition unit that obtains input voice information from the speech device;
A voice information recognition unit that recognizes the input voice information;
An utterance content determination unit that determines the utterance content by referring to one or more databases according to the mode set in the utterance control device and the recognition information recognized by the speech information recognition unit;
A voice output unit that outputs the utterance content determined by the utterance content determination unit to the utterance device;
Equipped with
The speech device is
An utterance content acquisition unit for acquiring the utterance content output from the utterance control device;
An utterance unit for uttering the acquired utterance content;
Equipped with
The utterance content determination unit determines whether the recognition information corresponds to the mode, and does not determine the utterance content if the recognition information does not correspond to the mode.
Utterance system characterized by It is a program for functioning a computer as an utterance control device according to any one of claims 1 to 4 , and the program for functioning a computer as said each part.

Images