JP6124844B2 - SERVER, METHOD USING DATABASE, PROGRAM, SYSTEM, TERMINAL, TERMINAL PROGRAM, AND VOICE DATA OUTPUT DEVICE - Google Patents

Description

  The present disclosure relates to a database, a method using the database, a program, a system, a terminal, a terminal program, and an audio data output device.

  Conventionally, a system capable of interacting with a user has been proposed as a dialog system. In this system, voices of a plurality of characters used for dialogue are provided, and switching of character voices is generally performed in accordance with the user's preference (Patent Document 1).

JP 2006-337432 A

  On the other hand, when the voices of a plurality of characters are made available, it is necessary to prepare a dictionary necessary for dialogue for each character.

  In this regard, it is necessary to update the dictionary of each character every time a new function dialogue pattern is added, and there is a problem that the number of characters becomes complicated.

  The present disclosure has been made to solve the above-described problem, and is a database capable of updating a dictionary in a simple manner, a method using the database, a program, a system, a terminal, a terminal program, and An object is to provide an audio data output device.

  A database according to an embodiment of the present disclosure is a database configured to be additionally writable and used for response processing to a request from a user, provided for each of a plurality of selectable characters, and for each of the characters A plurality of basic dictionaries in which response data used for response processing to the first request is stored, and used for response processing to the second request instead of the basic dictionary associated with the plurality of characters and corresponding to the selected character. And an additional dictionary in which response data is stored.

  Preferably, the additional dictionary is provided corresponding to an abstract character obtained by abstracting a plurality of characters.

  Preferably, the response data is text data used for outputting a voice as a response process to a request from the user.

  Preferably, the response data of the additional dictionary is text data that is generalized while excluding unique features while leaving the common features of the response data of the basic dictionary.

  A method according to an embodiment of the present disclosure is configured to be additionally writable, provided corresponding to a plurality of selectable characters, and a plurality of basic dictionaries in which response data is stored for each character, and a plurality of characters. A method using a database including an additional dictionary in which response data stored in response data used instead of the basic dictionary corresponding to the selected character is stored, the step of receiving a request from the user, and the selection of the character A step of accepting, and a step of executing response processing in accordance with the accepted request. The step of executing the response process includes a step of extracting response data using a basic dictionary provided corresponding to the selected character when the received request is the first request, and the received request includes If it is the second request, extracting response data using an additional dictionary associated with the selected character, synthesizing speech based on the extracted response data and the selected character And outputting voice data obtained by voice synthesis.

  A program according to an embodiment of the present disclosure is configured to be additionally writable, provided corresponding to each of a plurality of selectable characters, a plurality of basic dictionaries in which response data is stored for each character, and a plurality of characters A program for causing a computer to use a database including an additional dictionary in which response data used in place of a basic dictionary corresponding to a selected character is stored and stored in the computer. And a step of receiving a selection from the character, a step of receiving a selection of a character, and a step of executing a response process in accordance with the received request. The step of executing the response process includes a step of extracting response data using a basic dictionary provided corresponding to the selected character when the received request is the first request, and the received request includes If it is the second request, extracting response data using an additional dictionary associated with the selected character, synthesizing speech based on the extracted response data and the selected character And outputting voice data obtained by voice synthesis.

  A system according to an embodiment of the present disclosure is a system that uses a database configured to be additionally writable, and is received by a receiving unit that receives a request from a user, a selection receiving unit that receives a selection of a character, and a receiving unit. Response execution means for executing response processing according to the request is provided. A database is provided corresponding to each of a plurality of selectable characters, a plurality of basic dictionaries in which response data is stored for each character, and a basic dictionary corresponding to the plurality of characters and corresponding to the selected character. And an additional dictionary in which response data used instead is stored. When the received request is the first request, the response execution means includes a first extraction means for extracting response data using a basic dictionary provided corresponding to the selected character, and the received request is In the case of the second request, the second extraction means for extracting response data using an additional dictionary associated with the selected character, voice based on the extracted response data and the selected character Speech synthesis means for synthesizing; and output means for outputting the synthesized voice data.

  A method according to another embodiment of the present disclosure includes a plurality of basic dictionaries configured to be additionally writable, provided corresponding to a plurality of selectable characters, and including a plurality of basic dictionaries in which response data is stored for each character. A method using a second database including a database and an additional dictionary associated with a plurality of characters and storing response data used in place of a basic dictionary corresponding to a selected character, the request from a user , Receiving a character selection, and executing a response process in accordance with the received request. The step of executing the response process is a step of extracting response data using the basic dictionary of the first database provided corresponding to the selected character when the received request is the first request. When the received request is a second request, a step of extracting response data using an additional dictionary of the second database associated with the selected character, and the extracted response data are selected. And synthesizing the voice based on the character and outputting the synthesized voice data.

  A terminal according to an embodiment of the present disclosure is provided in an external device, is configured to be additionally writable, is provided corresponding to each of a plurality of selectable characters, and is used for response processing to the first request for each character. A plurality of basic dictionaries in which response data is stored, and an additional dictionary in which response data stored in response processing for the second request is stored in place of the basic dictionary associated with the plurality of characters and corresponding to the selected character A receiving unit that receives a first or second request from a user, a selection receiving unit that receives a selection of a character, and a first or second request received by the receiving unit. The voice synthesis based on the response data extracted using the database and the character selected by the selection receiving means. And output means for outputting the audio data.

  A terminal according to another embodiment of the present disclosure is provided in the main body, is configured to be additionally writable, is provided corresponding to each of a plurality of selectable characters, and is used for response processing for the first request for each character. A first database including a plurality of basic dictionaries in which response data is stored, and a response to the second request instead of the basic dictionary corresponding to the selected character provided in the external device and associated with the plurality of characters A terminal using a second database including an additional dictionary in which response data used for processing is stored, receiving means for receiving a first or second request from a user, and selection receiving means for receiving a selection of a character And response executing means for executing response processing in accordance with the request received by the receiving means. The response execution means includes an extraction means for extracting response data using the basic dictionary of the first database provided corresponding to the selected character when the reception means receives the first request; Means for acquiring response data extracted using the additional dictionary of the second database associated with the selected character when the means receives the second request, and the response data extracted or acquired And a voice synthesizing unit that synthesizes speech based on the character selected by the selection accepting unit, and an output unit that outputs the synthesized voice data.

  A terminal according to another embodiment of the present disclosure is provided in an external device, provided corresponding to each of a plurality of selectable characters, and response data used for response processing to the first request is stored for each character. A first database including a plurality of basic dictionaries and response data provided in the main body, associated with the plurality of characters, and used for response processing to the second request instead of the basic dictionary corresponding to the selected character Is a terminal that uses the second database including the additional dictionary in which is stored, and accepts the first or second request from the user, the selection accepting means for accepting the selection of the character, and the accepting means. Response executing means for executing response processing according to the received request. A response execution unit configured to acquire response data by using a basic dictionary of the first database provided corresponding to the selected character when the reception unit receives the first request; Means for extracting response data using the additional dictionary of the second database associated with the selected character when the second request is received by the means, and the response data extracted and acquired and the selection reception Speech synthesizing means for synthesizing speech based on the character selected by the means, and output means for outputting the synthesized voice data.

  A terminal according to another embodiment of the present disclosure is provided in the main body, is configured to be additionally writable, is provided corresponding to each of a plurality of selectable characters, and is used for response processing for the first request for each character. A plurality of basic dictionaries in which response data is stored, and an additional dictionary in which response data stored in response processing for the second request is stored in place of the basic dictionary associated with the plurality of characters and corresponding to the selected character A terminal that uses a database including: a receiving unit that receives a first or second request from a user, a selection receiving unit that receives a selection of a character, and a response process according to the request received by the receiving unit Response execution means. The response executing means extracts first response data using the basic dictionary of the first database provided corresponding to the selected character when the accepting means accepts the first request. And a second extracting means for extracting response data using an additional dictionary in the second database associated with the selected character when the receiving means accepts the second request, and the extracted response Speech synthesis means for synthesizing speech based on the data and the character selected by the selection accepting means, and output means for outputting the synthesized voice data.

  A terminal program according to an embodiment of the present disclosure is provided in an external device, configured to be additionally writable, provided corresponding to each of a plurality of selectable characters, and used for response processing to the first request for each character. A plurality of basic dictionaries in which response data is stored, and an additional dictionary in which response data stored in response processing for the second request is stored in place of the basic dictionary associated with the plurality of characters and corresponding to the selected character A terminal program that causes a computer of a terminal that uses a database including the terminal program to receive a first or second request from a user to the computer, a step of receiving a selection of a character, Extraction using database according to the first or second request received And a step of outputting the audio data speech synthesis was based on the character selected in the response data and the selection receiving unit, to function so as to execute the process.

  A terminal program according to another embodiment of the present disclosure is provided in the main body, is configured to be additionally writable, is provided corresponding to each of a plurality of selectable characters, and is used for response processing to the first request for each character. A first database including a plurality of basic dictionaries in which response data is stored, and a response to the second request instead of the basic dictionary corresponding to the selected character provided in the external device and associated with the plurality of characters A terminal program for causing a computer of a terminal that uses a second database including an additional dictionary in which response data used for processing is stored to be executed by the computer. In accordance with the received request, the step of receiving the selection of the character, the step of receiving the selection of the character, Function is to so as to execute a process comprising the step of performing a response process Te. The step of executing the response process includes a step of extracting response data using the basic dictionary of the first database provided corresponding to the selected character when the first request is received; The response data extracted using the additional dictionary of the second database associated with the selected character when the request is received, the response data extracted or acquired and the selected character And synthesizing speech based on the above and outputting speech synthesized speech data.

  A terminal program according to another embodiment of the present disclosure is provided in an external device, provided corresponding to each of a plurality of selectable characters, and response data used for response processing to the first request is stored for each character. A first database including a plurality of basic dictionaries and response data provided in the main body, associated with the plurality of characters, and used for response processing to the second request instead of the basic dictionary corresponding to the selected character Is a terminal program for causing a computer of a terminal that uses the second database including the additional dictionary stored therein to receive the first or second request from the user to the computer; , Receiving the character selection, and executing response processing according to the received request And a step to function so as to execute the process. The step of executing the response process includes a step of acquiring response data using a basic dictionary of the first database provided corresponding to the selected character when the first request is received, The response data is extracted using the additional dictionary of the second database associated with the selected character, and the response data extracted or acquired and selected by the selection receiving means Voice synthesis based on the character, and outputting voice synthesized voice data.

  A terminal program according to another embodiment of the present disclosure is provided in the main body, is configured to be additionally writable, is provided corresponding to each of a plurality of selectable characters, and is used for response processing to the first request for each character. A plurality of basic dictionaries in which response data is stored, and an additional dictionary in which response data stored in response processing for the second request is stored in place of the basic dictionary associated with the plurality of characters and corresponding to the selected character A terminal program for causing a computer of a terminal that uses a database including: a terminal program to accept a first or second request from a user to a computer; and a step of accepting selection of a character. And executing a response process in accordance with the received request. To function so as to line. The step of executing the response process includes a step of extracting response data using the basic dictionary of the first database provided corresponding to the selected character when the first request is received; The response data is extracted using the additional dictionary of the second database associated with the selected character, and the voice is generated based on the extracted response data and the selected character. A step of synthesizing, and a step of outputting the synthesized voice data.

  An audio data output device according to an embodiment of the present disclosure is an audio data output device that uses a database configured to be additionally writable, and includes a reception unit that receives a request from a user, and a selection reception unit that receives a selection of a character. Response executing means for executing response processing in accordance with the request received by the receiving means. The database includes a plurality of basic dictionaries that are respectively provided corresponding to a plurality of selectable characters and in which response data is stored for each character. The voice data output device includes an additional dictionary receiving unit that receives an additional dictionary in which response data used instead of the basic dictionary corresponding to the selected character is stored, and the additional dictionary according to the reception of the additional dictionary of the additional dictionary receiving unit And a registration unit for registering a correspondence table representing a correspondence relationship between the plurality of characters and the additional dictionary in the database. When the received request is the first request, the response execution means includes a first extraction means for extracting response data using a basic dictionary provided corresponding to the selected character, and the received request is In the case of the second request, with reference to the correspondence table, second extraction means for extracting response data using an additional dictionary corresponding to the selected character, the extracted response data, and the selected character Voice synthesis means for synthesizing speech based on the above and output means for outputting voice data synthesized by voice synthesis.

  The above and other objects, features, aspects and advantages of this disclosure will become apparent from the following detailed description of this disclosure, which is to be understood in connection with the accompanying drawings.

  It is not necessary to provide a dictionary for each character, and the dictionary can be updated by a simple method.

It is a figure explaining the audio | voice output system 1 based on Embodiment 1. FIG. It is a block diagram showing the outline | summary of the hardware constitutions of the cleaning robot 100 based on Embodiment 1. FIG. It is a block diagram showing the outline | summary of the hardware constitutions of the server 300 based on Embodiment 1. FIG. It is a sequence diagram explaining the flow of the response process in the audio | voice output system 1 based on Embodiment 1. FIG. It is a block diagram explaining the function of the server 300 based on Embodiment 1. FIG. It is a figure explaining the structure of the database 531 based on Embodiment 1. FIG. It is a figure explaining the specific example of the database 531 based on Embodiment 1. FIG. It is a figure explaining the flow of the server 300 based on Embodiment 1. FIG. It is a block diagram explaining the function of the audio | voice output system based on Embodiment 4. FIG. It is a sequence diagram explaining the flow of the response process in the audio | voice output system 1A based on Embodiment 4. FIG. It is a figure explaining the structure of the server based on Embodiment 7. FIG.

  Hereinafter, embodiments will be described with reference to the drawings. In the description of the embodiments, when referring to the number and amount, the scope of the present invention is not necessarily limited to the number and amount unless otherwise specified. In the description of the embodiments, the same parts and corresponding parts are denoted by the same reference numerals, and redundant description may not be repeated. Unless there is a restriction | limiting in particular, it is planned from the beginning to use suitably the structure shown in the structure shown to embodiment.

<Embodiment 1>
(Configuration of audio output system 1)
FIG. 1 is a diagram illustrating an audio output system 1 based on the first embodiment.

  With reference to FIG. 1, the audio output system 1 according to the first embodiment includes a cleaning robot 100, a network 5, an external device 50, and a server 300.

  The cleaning robot 100 is provided so as to be able to communicate with the server 300 via the network 5. In this example, the case of communicating with the server 300 via the network 5 is described, but a method of directly communicating with the server 300 may be used.

  As an example of the audio data output device, the audio output system 1 outputs an audio from a cleaning robot 100 to a human (user), and receives an audio from the human (user) input to the cleaning robot 100. The cleaning robot 100 outputs voice (hereinafter also referred to as “voice response”) that is recognized by the server 300 and that represents the response content to the input voice. By repeating this process, the audio output system 1 according to the first embodiment realizes a pseudo conversation between the user and the cleaning robot 100.

  In the first embodiment, as an example of the voice data output device, a cleaning robot 100 that recognizes voice and outputs a voice response to the user will be described as an example, but the present invention is limited to this. is not. For example, a doll having an interactive function, a household appliance other than the cleaning robot 100 (for example, a television, a microwave oven, etc.), a mobile phone, a smartphone, a tablet terminal, a personal computer or other information processing terminal, an electronic piano or other electronic musical instrument, It can also be realized by an automobile or other equipment.

  In the embodiment, a configuration in which the server 300 is realized by one server will be described as an example. However, the present invention is not limited to this, and at least one of the units (each function) included in the server 300 is described. A configuration may be adopted in which the unit is realized by another server.

  In this example, the server 300 can execute a predetermined function to be described later in cooperation with the external device 50. For example, the server 300 can access the external device 50 to acquire information related to the weather forecast, and execute a response process based on the acquired information.

(Hardware of audio output system 1)
FIG. 2 is a block diagram illustrating an outline of a hardware configuration of the cleaning robot 100 based on the first embodiment.

  As shown in FIG. 2, the cleaning robot 100 includes a CPU (Central Processing Unit) 610, a temporary storage unit 620, a storage unit 630, a communication unit 640, an input unit 650, and an output unit 660.

  As a control unit, CPU 610 executes a command and controls the operation of cleaning robot 100.

  The temporary storage unit 620 temporarily holds data generated by the CPU 610, data read from the storage unit 630, and the like. Temporary storage unit 620 is realized by, for example, a RAM (Random Access Memory) or other volatile data storage medium.

  The storage unit 630 holds data generated by the CPU 610, prestored data, programs, and the like. Storage unit 630 is implemented by, for example, a hard disk device, a flash memory, or other non-volatile data recording medium.

  The communication unit 640 is connected to the network 5 and communicates with the server 300. It is also possible to communicate with a mobile phone, a smartphone or other information communication terminal. The communication unit 640 is realized by, for example, wireless communication or wired communication. The mode of communication by the communication unit 640 is not particularly limited, and is realized by packet communication, infrared communication, Bluetooth (registered trademark), NFC (Near Field Communication), or the like.

  The input unit 650 is realized by a microphone, for example. Specifically, sound input is accepted from the outside. Note that the sound data indicating the sound that the microphone receives input will be described in the case of receiving sound data (also referred to as sound data) included in the frequency band of sound mainly produced by humans. Sound data including a frequency band other than the band may be included. The microphone outputs audio data indicating the input sound to the CPU 610. As a method for detecting sound data from sound data, for example, a method for detecting sound data by extracting a frequency band (for example, a frequency band of 100 Hz or more and 1 kHz or less) of a sound uttered by a human from sound data may be mentioned. Can do. In this case, the input unit 650 may include, for example, a bandpass filter or a filter that combines a high-pass filter and a low-pass filter in order to extract a frequency band of a human-generated voice from sound data. .

  Note that the input unit 650 may include, for example, a keyboard, a mouse or other pointing device, a signal input terminal, an infrared light receiving unit, and the like.

  The output unit 660 is realized by a speaker, for example. Specifically, an audio signal representing the response content output to the outside is reproduced. Output unit 660 can include, for example, a liquid crystal monitor, an organic EL (Electro Luminescence) monitor, and an LED external output interface.

  The driving unit 670 is a wheel that the cleaning robot 100 moves and a motor that drives the wheel. For example, when the device is different from the cleaning robot 100, the drive unit 670 may include a communication circuit, a vibrator, a wheel, a compressor, an image processor, and the like.

The cleaning unit 680 includes a brush, a suction pump, and the like.
In addition, the said structure is not necessarily essential structure, For example, it is also possible to add or delete a function according to apparatuses, such as the cleaning part 680. FIG.

  FIG. 3 is a block diagram showing an outline of the hardware configuration of the server 300 based on the first embodiment.

  As illustrated in FIG. 3, the server 300 includes a CPU 510, a temporary storage unit 520, a storage unit 530, a communication unit 540, an input unit 550, and an output unit 560.

As a control unit, CPU 510 executes instructions and controls the operation of server 300.
Temporary storage unit 520 temporarily holds data generated by CPU 510, data read from storage unit 530, data given to server 300, and the like. Temporary storage unit 520 is realized by, for example, a RAM (Random Access Memory) or other volatile data storage medium.

  The storage unit 530 holds data generated by the CPU 510, data given to the server 300, data stored in advance for causing the server 300 to perform a predetermined operation, a program, and the like. Storage unit 530 is realized by, for example, a hard disk device, a flash memory, or other nonvolatile data recording medium.

  Communication unit 540 is connected to network 5 and communicates with cleaning robot 100. The communication unit 540 is realized by, for example, wireless communication or wired communication. The mode of communication by the communication unit 540 is not particularly limited, and is realized by packet communication, infrared communication, Bluetooth (registered trademark), NFC (Near Field Communication), or the like.

  The input unit 550 accepts input of commands or characters and other information to the server 300. The input unit 550 can include, for example, a keyboard, a mouse or other pointing device, a signal input terminal, an infrared light receiving unit, and the like.

  The output unit 560 outputs data generated in the server 300, a result searched by the CPU 510, and the like. The output unit 560 is realized by, for example, a liquid crystal monitor, an organic EL (Electro Luminescence) monitor, an LED external output interface, or the like.

(Response processing overview)
FIG. 4 is a sequence diagram illustrating the flow of response processing in the audio output system 1 based on the first embodiment.

  As shown in FIG. 4, the user utters (also referred to as user utterance) to cleaning robot 100 (sequence sq0).

  Cleaning robot 100 accepts an input of voice in response to the user utterance (sequence sq1). Specifically, the cleaning robot 100 receives an external sound input via a microphone.

  Next, cleaning robot 100 outputs audio data to server 300 (sequence sq2). Specifically, the received audio data is output to the server 300 via the communication unit 640.

  Next, server 300 receives the voice data transmitted from cleaning robot 100 and executes voice recognition (sequence sq3). Specifically, the audio data is received via the communication unit 540. Then, the voice content of the received voice data is recognized. Server 300 executes response processing based on the recognized audio content (sequence sq4).

  Next, server 300 transmits voice data generated by voice synthesis as a result of the response process to cleaning robot 100 (sequence sq5). Specifically, the voice data is transmitted to the cleaning robot 100 via the communication unit 540.

  Next, cleaning robot 100 executes an output process of audio data received from server 300 (sequence sq6). Specifically, the audio data from the server 300 is received via the communication unit 640.

Cleaning robot 100 reproduces the sound based on the sound data (sequence sq7).
Specifically, an audio signal is reproduced through a speaker.

  By this process, it is possible to perform a response process according to the content uttered by the user and to output a voice indicating the response content to the user.

  In this example, a method of executing the voice recognition process and the voice synthesis process on the server 300 side will be described as an example. However, the present invention is not limited to this, and the voice recognition process and the voice synthesis process are executed on the cleaning robot 100 side. Alternatively, one may be executed on the cleaning robot 100 side and the other on the server 300 side. The same applies to the following configurations.

Next, a specific configuration of the server 300 will be described.
(Functional block diagram of server 300)
FIG. 5 is a block diagram illustrating functions of the server 300 based on the first embodiment.

  As shown in FIG. 5, the server 300 includes a character selection input receiving unit 400, a character setting unit 402, a response execution unit 404, a voice input reception unit 414, a voice recognition unit 416, and an alarm execution unit 418. And a database (DB) 531.

  The character selection input receiving unit 400 receives an instruction to select and input a character from the user. In this example, as voice reproduction by response processing to a user utterance, it is possible to select and reproduce voices of a plurality of characters provided in advance.

  Here, the “character” means a character possessed by speech, and has a unique character by a characteristic or individual tone, character, voice color, or a combination thereof. For example, there is a habit in the tone and the voice is spoken in the Kansai dialect. Further, the character is not limited to a person but may be an animal or the like, or anthropomorphized person. For example, it may be a cry of a dog, an anthropomorphic dog that utters human words, or an existing animation character.

  For example, a plurality of character selection instructions are executed in accordance with an operation of a predetermined key provided on the input unit 650 of the cleaning robot 100, and the selection instructions are transmitted to the server 300 via the communication unit 640. Server 300 receives the selection instruction via communication unit 540 and receives a character selection instruction at character selection input reception unit 400. In this example, a case where four selectable characters A to D are provided will be described.

  The character setting unit 402 sets the selected character. The phoneme pattern in the speech synthesis is switched according to the setting of the character. The phoneme pattern also includes data related to the uttered voice such as loudness, utterance speed, volume, high frequency emphasis, and inflection. Audio data corresponding to the character is generated according to the setting of the character.

  The voice input receiving unit 414 receives voice data input from the cleaning robot 100 via the communication unit 540. Then, the received voice data is output to the response execution unit 404.

The response execution unit 404 executes response processing according to the received audio data.
Specifically, the response execution unit 404 includes a speech recognition unit 416, a first extraction unit 406, a second extraction unit 408, a speech synthesis unit 410, and a data output unit 412.

  The voice recognition unit 416 recognizes the voice content (sound content) indicated by the voice data received by the voice input reception unit 414 as the recognized content (recognition phrase).

  The first extraction unit 406 refers to the basic dictionary of the corresponding character in the standard conversation function dictionary group 532 stored in the database 531 based on the acquired recognition phrase, and responds to the voice content indicated by the voice data Select (determine) the content (response information).

  Based on the acquired recognition phrase, the second extraction unit 408 refers to a dictionary in a dictionary group other than the standard conversation function dictionary group 532 stored in the database 531 and corresponds to the voice content indicated by the voice data. Select (determine) response content (response information).

  In any of the first extraction unit 406 and the second extraction unit 408, referring to the database 531 stored in the storage unit 530, if a recognition phrase for voice data cannot be obtained using a dictionary, voice recognition is performed. Is determined to have failed, and the response processing is terminated.

  The speech synthesizer 410 synthesizes speech based on the selected response content and the set phoneme pattern of the character to generate speech data.

  The data output unit 412 transmits the generated voice data to the cleaning robot 100 via the communication unit 540. The cleaning robot 100 receives the audio data via the communication unit 640 and reproduces and outputs the audio from the output unit 660.

  The alarm execution unit 418 instructs to execute the notification process at a preset time when the alarm function is set to be valid. Specifically, the alarm execution unit 418 has a time management function, and outputs a notification processing instruction to the response execution unit 404 when the set time is reached.

  The function dictionary group reception unit 440 receives a dictionary group for executing a response process when a new function to be added to the database 531 is added. Here, the function dictionary group to be accepted is a dictionary corresponding to an abstract character described later.

  The additional basic dictionary receiving unit 450 receives a basic dictionary to be added to the database 531. Here, the basic dictionary to be accepted is a dictionary to be added corresponding to a new character to be described later.

  The registration unit 442 adds the function dictionary group received by the function dictionary group reception unit 440 to the database 531. The registration unit 442 adds the basic dictionary received by the additional basic dictionary reception unit 450 to the database 531. Specifically, when adding a basic dictionary corresponding to a new character, the registration unit 442 registers the abstract character registered in the database and the new character in association with each other.

(Database configuration)
FIG. 6 is a diagram illustrating the configuration of the database 531 based on the first embodiment.

  As shown in FIG. 6, the database 531 is stored in the storage unit 530 and is configured to be additionally writable.

The database 531 stores a dictionary for executing response processing for each function.
In this example, the database 531 includes a standard conversation function dictionary group 532 for executing standard conversation response processing. The standard conversation function dictionary group 532 includes a dictionary for executing a clock response process (clock function).

  The database 531 further includes a weather forecast function dictionary group 534 for executing a weather forecast response process (weather forecast function) and an alarm function dictionary group 538 for executing an alarm response process (alarm function). And have.

  The weather forecast function dictionary group 534 includes an abstract character X additional dictionary 534A and an abstract character Y additional dictionary 534B.

  The alarm function dictionary group 538 includes an abstract character X additional dictionary 538A and an abstract character Y additional dictionary 538B.

  Further, the database 531 is provided with a correspondence table 539 indicating associations between a plurality of characters generated by the registration unit 442 and abstract characters. In the correspondence table 539, the characters A and B and the abstract character X are registered in association with each other. Characters C and D and abstract character Y are registered in association with each other.

  When a new character is added, the registration unit 422 generates a new correspondence table indicating the character association. A correspondence table in which the abstract character and the new character are associated with each other is generated.

  In this example, a weather forecast function dictionary group 534 for executing a weather forecast function and an alarm function dictionary group 538 for executing an alarm function are newly added to the database 531 as an example. Specifically, the weather forecast function dictionary group 534 and the alarm function dictionary group 538 are received by the function dictionary group reception unit 440 and stored in the database 531 by the registration unit 442.

  By adding the new dictionary group to the server 300, it is possible to add various voice output patterns and expand the functions.

  Here, the standard conversation function dictionary group 532 includes a character A basic dictionary 532A, a character B basic dictionary 532B, a character C basic dictionary 532C, and a character D basic dictionary 532D.

In other words, four basic dictionaries for characters are provided in advance for standard conversation.
Therefore, in general, when expanding (adding) a function, it is necessary to provide a new additional dictionary for each character in accordance with the four basic character dictionaries provided in advance.

  On the other hand, creating an additional dictionary tailored to all characters imposes a load on the side of adding a function, and the process becomes complicated when the number of characters is large.

  Therefore, in this example, instead of creating and adding additional dictionaries tailored to each character, an abstract that excludes unique features while leaving common features of similar characters from multiple characters. A character is defined, and an additional dictionary (text data) for the abstract character (also called an abstract character) is created.

  For example, the character A and the character B are defined as an abstract character X that leaves a common general polite tone such as “is”, “masu”, etc., and excludes a unique characteristic tone.

  Further, an abstract character Y is defined from the character C and the character D, leaving a common general friend tone and excluding a unique characteristic tone.

  Then, an additional dictionary of polite tone abstract characters X and an additional dictionary of friend tone characters Y are created.

  Then, the character A and the character B talking in a polite tone such as “is” or “mas” are associated with the polite tone abstract character X.

  Also, the friend tone character C and character D such as “Dayon” are associated with the friend tone abstract character Y.

  In this example, when the function is expanded (added) and the function is used, an additional dictionary of the abstract character X or abstract character Y is used to generate a sound using the set phoneme pattern of the character. Audio data is generated by synthesis.

  For example, when the user has set character A or B, voice synthesis is performed with the phoneme pattern of character A or B using the additional dictionary of abstract character X. Since the additional dictionary is defined as a polite tone, the voice data created by voice synthesis with the character A or B talking in a polite tone can maintain a natural conversation without causing a sense of incongruity.

  Further, when the user has set the character C or D, voice synthesis is performed with the phoneme pattern of the character C or D using the additional dictionary of the abstract character Y. Since the additional dictionary is defined as a friend tone, it is possible to maintain a natural conversation without causing a sense of incongruity in voice data created by voice synthesis with the character C or D talking in a friend tone.

  In this example, two abstract characters, a polite tone abstract character, and a friend tone abstract character have been described. However, the present invention is not limited to this. For example, a male tone abstract character and a woman tone abstract character are provided. An additional dictionary may be created by defining a character. In this example, the case where an abstract character is defined by tone is described as an example. However, the present invention is not limited to tone, and can be defined separately.

  For example, when a new function is added by this method, instead of creating and adding an additional dictionary for each character, an abstract character is defined and associated with a plurality of characters. By using this dictionary, it is possible to reduce the load on the side of creating a dictionary by adding functions. Further, the number of dictionaries to be added can be reduced and management is easy, and the capacity for storing dictionaries can be reduced.

  Further, when creating a new character (for example, character E), only the basic dictionary is added. In the database 531, a correspondence table 539 in which a new character E and an abstract character are associated is generated by the registration unit 442. For example, if the new character E is a character that speaks in a polite tone, it is registered in association with the abstract character X. Thus, when creating a character, it is not necessary to add a dictionary corresponding to all functions. For example, if a dictionary for executing response processing of a standard conversation is created, an abstract character dictionary is used for other functions. By using it, it is possible to perform response processing while maintaining a natural conversation. Therefore, it is possible to reduce the load on the side of creating a basic dictionary corresponding to a new character.

FIG. 7 is a diagram illustrating a specific example of the database 531 based on the first embodiment.
As an example, the database 531 is stored in the storage unit 530 included in the server 300 based on the embodiment.

  FIG. 7A shows a table in which the response processes for the recognition phrases “Good morning” and “What time is it?” Are arranged from the basic dictionary provided for each character in the standard conversation function dictionary group 532. The database 531 stores text data as the contents of response processing.

  When the recognition phrase “good morning” is recognized, the answer phrase is different for each character. As an example, in the case of character A, “Good morning, good morning” is determined as the response content. In the case of the character B, “Good morning” is determined as the response content. In the case of the character C, “Ohayo” is determined as the response content. In the case of the character D, “Oh-ha” is determined as the response content.

  The command when the recognition phrase “good morning” is recognized is set to “none”. Therefore, it becomes a normal standard conversation.

  As another example, when the recognition phrase “What time is it now?” Is recognized, the answer phrase is different for each character. As an example, in the case of the character A, “... Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command when the recognition phrase “what time is it now” is set to “acquire time”. Therefore, the current time is acquired. For example, when “8 o'clock” is acquired, “8 o'clock is good” is determined as the response content.

  In the case of the character B, “... Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command when the recognition phrase “what time is it now” is set to “acquire time”. Therefore, the current time is acquired. For example, when “8 o'clock” is acquired, “It is 8 o'clock” is determined as the response content.

  In the case of the character C, “...” And “Dayon” are determined as response contents. Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command when the recognition phrase “what time is it now” is set to “acquire time”. Therefore, the current time is acquired. For example, when “8 o'clock” is acquired, “8 o'clock” is determined as the response content.

  In the case of the character D, “...” And “DA ONE” are determined as response contents. Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command when the recognition phrase “what time is it now” is set to “acquire time”. Therefore, the current time is acquired. For example, if “8 o'clock” is acquired, “8 o'clock is one” is determined as the response content.

  FIG. 7B shows a table in which the response processing for the recognition phrase is organized from the abstract character addition dictionaries provided in the alarm function dictionary group 538 and the weather forecast function dictionary group 534.

A case where the recognition phrase “What is the weather today?” Will be described.
When the recognition phrase “How is the weather today?” Is recognized, the answer phrase is different for each character. As an example, in the case of the abstract character X, “...” and “This is a forecast” are determined as response contents. Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command for recognizing the recognition phrase “What is the weather today?” Is set to “Acquire Weather Information”. Therefore, weather information is acquired. For example, if “sunny” is acquired, “sunny forecast” is determined as the response content.

  In the case of the abstract character Y, “...” And “Yes” are determined as response contents. Here, information acquired by a command is inserted into the portion of “...”. Specifically, the command for recognizing the recognition phrase “What is the weather today?” Is set to “Acquire Weather Information”. Therefore, the current time is acquired. Then, for example, when “sunny” is acquired, “sounds likely” is determined as the response content.

  As described above, instead of creating a dictionary for each character, it is possible to reduce the load on the side of creating the dictionary by creating and using a dictionary of abstract characters associated with a plurality of characters.

  In this example, the number of dictionaries is halved because abstract characters corresponding to two characters are defined.

  Next, the case where the recognition phrase “(alarm)” is recognized will be described. This function is not a response process to a user's utterance, but a response process that automatically executes a wake-up function (command) at a set predetermined time.

  Specifically, the recognition phrase “(alarm)” executes response processing on the assumption that the alarm is recognized as an utterance from the user. Specifically, it is determined that the recognition phrase “(alarm)” has been recognized when there is an instruction from the alarm execution unit 418.

  When the recognition phrase “(alarm)” is recognized, the answer phrase is different for each character. As an example, in the case of the abstract character X, “Morning, please wake up” is determined as the response content. In the case of the abstract character Y, “It's morning, get up” is determined as the response content.

  As described above, instead of creating a dictionary for each character, it is possible to reduce the load on the side of creating the dictionary by creating and using a dictionary of abstract characters associated with a plurality of characters.

  In this example, the number of dictionaries is halved because abstract characters corresponding to two characters are defined.

  Therefore, for example, as a vendor who adds a new function to the server, when an abstract character is defined, it is necessary to create a dictionary for every character by creating a dictionary for the abstract character. It is possible to reduce the burden of creating a dictionary.

  Further, not only an abstract character but also a normal character can be added to reduce the burden of creating a dictionary.

  For example, when adding a normal character, it is not necessary to create answer phrases corresponding to all functions in the basic dictionary. Specifically, for functions that have already been added, it is possible to use an additional dictionary of abstract characters associated with the character, so for example, a dictionary relating to standard conversation response processing may be created. This can reduce the burden of creating a dictionary even when adding a normal character.

  Also, when adding (installing) new functions, instead of creating and adding additional dictionaries for each character, for example, only creating dictionaries corresponding to abstract characters X and Y Since it is possible to maintain a natural conversation with no sense of incongruity, it is possible to reduce the load on the side of creating a dictionary by adding functions.

(Flow diagram)
FIG. 8 is a diagram illustrating the flow of the server 300 based on the first embodiment.

  As shown in FIG. 8, it is determined whether audio data has been received (step S1). Specifically, the voice input receiving unit 414 receives voice data.

  If it is determined in step S1 that voice data has been received (YES in step S1), then voice recognition is executed (step S2). Specifically, the voice recognition unit 416 acquires a recognition phrase according to the voice data received from the voice input reception unit 414.

  Next, the character is confirmed (step S3). Specifically, the first extraction unit 406 confirms the character set by the character setting unit 402.

  Next, a basic dictionary is set (step S4). The first extraction unit 406 sets a basic dictionary corresponding to the set character. For example, when character A is set, character A basic dictionary 532A is set.

  Next, it is determined whether or not there is a recognition phrase in the basic dictionary (step S5). The first extraction unit 406 determines whether a recognition phrase is included in the set basic dictionary. For example, it is determined whether or not the recognition phrase is included in the character A basic dictionary 532A.

  If it is determined in step S5 that there is a recognition phrase in the set basic dictionary (YES in step S5), an answer phrase is determined (step S6). The first extraction unit 406 determines an answer phrase from the set basic dictionary.

  Next, speech synthesis is performed (step S7). Specifically, the speech synthesizer 410 generates speech data by performing speech synthesis based on the determined answer phrase and the set phoneme pattern of the character.

  Next, it outputs (step S7 #). Specifically, the data output unit 412 outputs the voice data generated by the voice synthesis unit 410 to the cleaning robot 100 via the communication unit 540.

Then, the process ends (END).
As a result, the cleaning robot 100 receives the audio data generated by the server 300, and reproduces and outputs it.

  On the other hand, if it is determined in step S5 that there is no recognition phrase in the basic dictionary (NO in step S5), an abstract character is set (step S8). Specifically, the second extraction unit 408 sets an abstract character associated with the character set by the character setting unit 402. For example, when the character A is set, the abstract character X is set.

  Next, an additional dictionary is set (step S9). Specifically, the second extraction unit 408 sets an additional dictionary corresponding to the set abstract character. For example, when an abstract character X is set, an abstract character X additional dictionary is set.

  Next, it is determined whether there is a recognition phrase in the additional dictionary (step S10). Specifically, the second extraction unit 408 determines whether there is a recognition phrase in the abstract character X additional dictionary 534A.

  If it is determined that there is a recognition phrase in the additional dictionary (YES in step S10), it is determined whether there is a command (step S11). When the second extraction unit 408 determines that there is a recognition phrase in the abstract character X additional dictionary 534A, the second extraction unit 408 determines whether there is a command corresponding to the recognition phrase. For example, in the case of the recognition phrase “What is the weather today?”, It is determined that there is a command because the command is associated with “acquire weather information”.

  If it is determined in step S11 that there is a command (YES in step S11), the command is executed (step S12). When the second extraction unit 408 determines that there is a command “acquire weather information” corresponding to the recognition phrase “what is the weather today?”, The second extractor 408 acquires weather information. Specifically, according to the command, the external device 50 is accessed to acquire information related to the weather forecast.

Then, an answer phrase is determined (step S6). The second extraction unit 408 determines an answer phrase based on information related to the weather forecast acquired by command execution. For example, “Sunny forecast” is determined. Next, speech synthesis is performed (step S7). Specifically, the speech synthesizer 410 generates speech data by performing speech synthesis based on the determined answer phrase and the set phoneme pattern of the character. For example, when the character A is set, voice data is generated based on the phoneme pattern of the character A and the answer phrase “It is a clear forecast”.

  Next, it outputs (step S7 #). Specifically, the data output unit 412 outputs the voice data generated by the voice synthesis unit 410 to the cleaning robot 100 via the communication unit 540.

Then, the process ends (END).
On the other hand, if it is determined in step S11 that there is no command (NO in step S11), an answer phrase is determined without executing the command (step S6). The second extraction unit 408 determines an answer phrase from the abstract character addition dictionary.

  Next, speech synthesis is performed (step S7). Specifically, the speech synthesizer 410 generates speech data by performing speech synthesis based on the determined answer phrase and the set phoneme pattern of the character. For example, when the character A is set, voice data is generated based on the phoneme pattern of the character A and the answer phrase.

  Next, it outputs (step S7 #). Specifically, the data output unit 412 outputs the voice data generated by the voice synthesis unit 410 to the cleaning robot 100 via the communication unit 540.

Then, the process ends (END).
On the other hand, if it is determined in step S10 that there is no recognition phrase in the additional dictionary (NO in step S10), it is determined whether there is another additional dictionary (step S13). When the second extraction unit 408 determines that there is no recognition phrase in the set abstract character X additional dictionary 534A, the second extraction unit 408 determines whether there is another additional dictionary.

  If it is determined in step S13 that there is another additional dictionary (YES in step S13), the process returns to step S9. If the second extraction unit 408 determines that there is another additional dictionary, it sets another additional dictionary. Then, the above process is repeated.

  On the other hand, if it is determined in step S13 that there is no other additional dictionary (NO in step S13), the response process for the recognized phrase is ended (END).

  The alarm execution unit 418 virtually outputs the voice data “(alarm)” to the response execution unit 404 at the set predetermined time. Thereby, the speech recognition unit 416 recognizes the speech, and the answer phrase is determined by the above processing.

<Embodiment 2>
In the above embodiment, first, a basic dictionary is set by the first extraction unit 406, and when it is determined that there is no recognition phrase in the basic dictionary, an additional dictionary is set by the second extraction unit 408 and there is a recognition phrase. However, the present invention is not limited to this, but the order is changed and the second extraction unit 408 first sets an additional dictionary corresponding to the abstract character to determine whether there is a recognition phrase. When it is determined that there is no recognition phrase, the first extraction unit 406 may set a basic dictionary to determine whether or not there is a recognition phrase in the basic dictionary.

  When the additional function is frequently used, it is possible to execute processing of the function added at high speed by processing the second extraction unit 408 prior to the first extraction unit 406.

<Embodiment 3>
In the above embodiment, the case where the first extraction unit 406 and the second extraction unit 408 are continuously processed has been described. However, the present invention is not limited to this, and the processing in the first extraction unit 406 and the second extraction unit 408 is performed. You may make it perform in parallel.

  Also, a plurality of selectable modes may be provided and any one of them may be executed. Specifically, the processing may be switched for each mode. For example, a basic mode that uses only the basic dictionary, an expansion mode that uses only the additional dictionary, a specific mode that uses only a specific dictionary in the additional dictionary, and the like are provided, and the first extraction unit 406 is configured according to the designated mode. It is also possible to switch and execute response processing using the basic dictionary used or response processing using an additional dictionary using the second extraction unit 408.

<Embodiment 4>
(Functional block diagram of audio output system based on Embodiment 4)
FIG. 9 is a block diagram illustrating functions of the audio output system based on the fourth embodiment.

  As shown in FIG. 9, the cleaning robot 100 is replaced with a cleaning robot 110, and the server 300 is replaced with a server 310.

Specifically, the cleaning robot 110 includes a CPU 611 and a database 631.
CPU 611 includes a character selection input receiving unit 200, a voice input receiving unit 214, a character setting unit 202, a response execution unit 204, and an alarm execution unit 218.

  The response execution unit 204 includes an extraction unit 206, an acquisition unit 208, a speech recognition unit 216, a speech synthesis unit 210, and a data output unit 212.

Server 310 includes a CPU 511 and a database 531A.
The CPU 511 includes a data extraction unit 420 and a data output unit 430.

  Although omitted here, the functional dictionary group reception unit 440, the additional basic dictionary reception unit 450, and the registration unit 442 described in the first embodiment may be further provided.

  The configuration of the fourth embodiment is a configuration in which the cleaning robot 110 has the main functions provided in the server 300 described in the first embodiment.

  The database 631 in this example includes the standard conversation function dictionary group 532 described with reference to FIG. Further, the database 531A in this example includes the alarm function dictionary group 538 and the weather forecast function dictionary group 534 described with reference to FIG.

  Specifically, the database 631 is used for the standard conversation function, and the database 531 is used for the additional function.

  Character selection input reception unit 200, character setting unit 202, voice input reception unit 214, alarm execution unit 218, extraction unit 206, voice recognition unit 216, voice synthesis unit 210, data output unit 212, character selection input reception unit 400, Character setting unit 402, voice input reception unit 414, alarm execution unit 418, first extraction unit 406, voice recognition unit 416, voice synthesis unit 410, and data output unit 412 are basically the same, and detailed description thereof will not be repeated. .

  The acquisition unit 208 refers to the dictionary of the additional dictionary group stored in the database 531A stored in the server 310 based on the recognition phrase recognized by the voice recognition unit 216, and corresponds to the voice content indicated by the voice data. Acquires the response contents (response information).

(Response processing overview)
FIG. 10 is a sequence diagram illustrating the flow of response processing in the audio output system 1A based on the fourth embodiment.

  As shown in FIG. 10, the user utters (also referred to as user utterance) to cleaning robot 110 (sequence sq10).

  Cleaning robot 110 accepts an input of voice for the user utterance (sequence sq11). Specifically, the input unit 650 receives external sound input via a microphone. The voice input reception unit 214 receives voice input from the input unit 650 and outputs it to the response execution unit 204.

  Next, the response execution unit 204 of the cleaning robot 110 executes voice recognition (sequence sq12). Specifically, the voice recognition unit 216 recognizes the voice content from the voice input reception unit 214.

  Next, in this example, the extraction unit 206 of the cleaning robot 110 determines that there is no dictionary corresponding to the recognized phrase in the standard conversation function dictionary group of the database 631 (sequence sq13).

  Next, acquisition unit 208 of cleaning robot 110 transmits data to server 310 (sequence sq14). The data includes information on the character set together with the recognized voice content.

  Server 310 performs response processing based on the data transmitted from cleaning robot 110 (sequence sq15). Specifically, the data extraction unit 420 refers to the database 531A for the transmitted data, and extracts response content (response information) corresponding to the audio content.

  The data extraction unit 420 performs the same process as the second extraction unit 408 described in the first embodiment.

  Next, server 310 transmits the extracted response content data to cleaning robot 110 (sequence sq16). Specifically, the data output unit 430 transmits the response content data to the cleaning robot 110 via the communication unit 540.

  Next, cleaning robot 110 executes a process for outputting data received from server 300 (sequence sq17). Specifically, the acquisition unit 208 acquires response content data transmitted from the server and outputs the response content data to the speech synthesis unit 210. The speech synthesizer 210 synthesizes speech based on the answer phrase included in the received response content and the set character to generate speech data.

  Cleaning robot 100 reproduces the sound based on the sound data (sequence sq18). Specifically, an audio signal is reproduced through a speaker. By this process, it is possible to perform a response process according to the content uttered by the user and to output a voice indicating the response content to the user.

  In this example, a case has been described in which the extraction unit 206 of the cleaning robot 110 determines that there is no dictionary corresponding to the recognized phrase in the standard conversation function dictionary group of the database 631.

  On the other hand, when it is determined that there is a dictionary corresponding to the recognized phrase in the standard conversation function dictionary group of the database 631, the extraction unit 206 extracts the answer phrase by referring to the database 631 without accessing the server 310. . Then, voice data is generated by performing voice synthesis based on the answer phrase and the set character. Then, the audio is reproduced based on the audio data.

  In this configuration, by storing the basic dictionary in the database 631 of the cleaning robot 110 and storing the additional dictionary in the database 531A of the server 310, it is possible to increase the speed of the dialogue using the basic dictionary in the cleaning robot 110. It is.

  Further, by providing the additional dictionary on the server 310 side, it is not necessary to consider the storage capacity limitation on the cleaning robot 110 side, and the additional dictionary can be easily added.

<Embodiment 5>
In the fourth embodiment, the configuration in which the standard conversation function dictionary group is provided on the cleaning robot 110 side and the additional dictionary group is provided on the server 310 side has been described. On the other hand, it is also possible to reverse the configuration and provide an additional dictionary group on the cleaning robot 110 side and a standard conversation function dictionary group on the server 310 side.

The process is the same except for the arrangement of the database.
With this configuration, it is possible to increase the speed of dialogue in the additional function using the additional dictionary in the cleaning robot 110.

<Embodiment 6>
In the above configuration, the configuration in which the cleaning robot and the server process in cooperation has been described. However, the cleaning robot and the server can perform the processing only with no cooperation with the server.

  Specifically, it can be realized by including the configuration of the server 300 described in FIG. 5 in the cleaning robot 100. With this configuration, it is possible to increase the speed of the conversation in the standard conversation and the additional function by providing all the dictionaries in the cleaning robot 100.

<Embodiment 7>
FIG. 11 is a diagram illustrating a configuration of a server based on the seventh embodiment.

  Referring to FIG. 11, in this example, a case where a plurality of servers are provided is shown.

  In this example, a case where a server 300A and a server 300B are provided is shown as an example.

  In the configuration of the first embodiment, the case where the voice recognition and the process for determining the answer phrase (response mode) for the voice recognition are executed by the same server has been described. It is also possible to execute with.

  Specifically, the server 300A may perform voice recognition on the voice data, and the server 300B may output answer phrase data to the cleaning robot 100.

  For example, audio data is transmitted from the cleaning robot 100 to the server 300A (1). The server 300A performs voice recognition of the voice data (2). Then, the server 300A transmits the recognition phrase to the cleaning robot 100 (3).

  The cleaning robot 100 receives the recognition phrase from the server 300A, and transmits the recognition phrase to another server 300B (4).

  The server 300B receives the recognition phrase from the cleaning robot 100 and determines an answer phrase corresponding to the recognition phrase (5). Then, the server 300B transmits an answer phrase to the cleaning robot (6).

  In addition, in this example, although the case where the server 300A transmits the recognition phrase which performed the speech recognition of audio | voice data with respect to the cleaning robot 100 was demonstrated, if it is the information which shows not only a recognition phrase but the speech recognition result, Any thing is good. For example, it may be access information (URL (Uniform Resource Locator) etc.) necessary for accessing an answer phrase stored in the server 300B. For example, the cleaning robot 100 may receive the access information (URL) from the server 300A and acquire the answer phrase from the server 300B by accessing the server 300B. Further, not limited to access information, when an answer phrase stored in the server 300B is saved in a file format, it is information for designating a file name as information indicating the result of speech recognition from the server 300A. May be. For example, the cleaning robot 100 may receive the file name from the server 300A, specify the file name to the server 300B, and request information, thereby acquiring the file related to the answer phrase from the server 300B. Is possible.

  Similarly, text information obtained by converting a recognition phrase into text may be transmitted as information indicating the result of speech recognition from the server 300A. The cleaning robot 100 may extract a recognition phrase from the text information and access the server 300B to obtain an answer phrase. Alternatively, the cleaning robot 100 may transmit the text information to the server 300B, and the server 300B may recognize the recognition phrase. It is good also as a structure which analyzes the text information containing, determines an answer phrase based on an analysis result, and transmits to the cleaning robot 100.

  Moreover, in this example, although the case where the speech recognition was performed by the server 300 was described, the speech recognition is performed by the cleaning robot 100, the answer phrase for the result is determined in the cleaning robot 100, and the answer phrase is acquired from the server 300B. You may do it. In that case, it can be realized by providing the storage unit 630 with a URL correspondence table in which access information (URL) for accessing the reply phrase of the server 300B for the recognized phrase is associated.

  It is also possible to acquire an answer phrase for the result of voice recognition using information stored in the cleaning robot 100.

  For example, when the cache memory capable of temporarily storing information includes information on the answer phrase for the previously used recognition phrase, the answer phrase information stored in the cache memory is By using it, for example, it is possible to acquire an answer phrase without accessing the server and utter (response process) from the cleaning robot 100. As a result, it is possible to utter early using the information stored in the cache memory.

  The server configuration of the seventh embodiment can be applied to any of the first to sixth embodiments.

<Embodiment 8>
Control blocks such as a cleaning robot and a server may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or by software using a CPU (Central Processing Unit). Good.

  In the latter case, the cleaning robot and the server include a CPU that executes instructions of a program that is software that realizes each function, and a ROM (Read Only Memory) in which the program and various data are recorded so as to be readable by the computer (or CPU). Alternatively, a storage device (these are referred to as “recording media”), a RAM (Random Access Memory) that expands the program, and the like are provided. And the objective of this invention is achieved when a computer (or CPU) reads the said program from the said recording medium and runs it. As the recording medium, a “non-temporary tangible medium” such as 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 an arbitrary transmission medium (such as a communication network or a broadcast wave) that can transmit 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.

  It should be thought that embodiment disclosed this time is an illustration and restrictive at no points. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  1, 1A voice output system, 5 network, 50 external device, 100, 110 cleaning robot, 200, 400 character selection input reception unit, 202, 402 character setting unit, 204, 404 response execution unit, 206 extraction unit, 208 acquisition unit , 210, 410 Speech synthesis unit, 212, 412, 430 Data output unit, 214 Speech input reception unit, 216, 416 Speech recognition unit, 218, 418 Alarm execution unit, 300, 300A, 300B, 310 server, 406 First extraction Unit, 408 second extraction unit, 414 voice input reception unit, 420 data extraction unit, 440 functional dictionary group reception unit, 450 additional basic dictionary reception unit, 531, 531A, 631 database, 520, 620 temporary storage unit, 530, 630 Storage unit, 532 standard conversation function dictionary group, 32A, 532B, 532C, 532D Basic dictionary, 534 Weather forecast function dictionary group, 534A, 534B, 538A, 538B Additional dictionary, 538 Alarm function dictionary group, 539 Corresponding table, 540, 640 Communication unit, 550, 650 Input unit, 560 , 660 output unit, 670 drive unit, 680 cleaning unit.

Claims (17)

A server that is used for response processing to a request from a user and is configured to be additionally writable,
A plurality of basic dictionaries provided corresponding to a plurality of selectable characters, each storing response data used for response processing to the first request for each character;
A server including an additional dictionary in which the response data stored in response to the second request is stored instead of the basic dictionary corresponding to the selected character and associated with the selected character. The server according to claim 1, wherein the additional dictionary is provided corresponding to an abstract character obtained by abstracting the plurality of characters. The server according to claim 1, wherein the response data is text data used for outputting a voice as a response process to the request from the user. The server according to claim 3, wherein the response data of the additional dictionary is text data excluding a characteristic characteristic of the response data of the basic dictionary. It is configured to be appendable, provided corresponding to each of a plurality of selectable characters, a plurality of basic dictionaries storing response data for each character, and associated with the plurality of characters, corresponding to the selected character Using a database including an additional dictionary storing the response data used instead of the basic dictionary,
Receiving a request from a user;
Receiving a character selection;
Executing a response process in accordance with the accepted request,
The step of executing the response process includes:
Extracting the response data using a basic dictionary provided corresponding to the selected character when the received request is the first request;
If the request received is the second request, and extracting the response data using the associated with the selected character added dictionary,
A method of using a database, comprising: synthesizing speech based on extracted response data and a selected character; and outputting speech synthesized speech data. It is configured to be appendable, provided corresponding to each of a plurality of selectable characters, a plurality of basic dictionaries storing response data for each character, and associated with the plurality of characters, corresponding to the selected character A program for causing a computer using a database including an additional dictionary in which the response data is used in place of the basic dictionary to be executed,
The program is stored in the computer.
Receiving a request from a user;
Receiving a character selection;
Executing a response process in accordance with the accepted request,
The step of executing the response process includes:
Extracting the response data using a basic dictionary provided corresponding to the selected character when the received request is the first request;
If the request received is the second request, and extracting the response data using the association et OF FURTHER dictionary to the selected character,
Synthesizing speech based on the extracted response data and the selected character;
A program that utilizes a database that functions to execute processing, including the step of outputting voice data that has been synthesized. A system that uses a database that can be appended to,
A receiving means for receiving a request from a user;
Selection accepting means for accepting selection of a character;
Response executing means for executing response processing in accordance with the request received by the receiving means,
The database is
A plurality of basic dictionaries provided corresponding to each of a plurality of selectable characters and storing response data for each character;
An additional dictionary in which the response data is stored in association with the plurality of characters and used instead of the basic dictionary corresponding to the selected character;
The response execution means includes
First extraction means for extracting response data using a basic dictionary provided corresponding to the selected character when the received request is a first request;
If the request received is the second request, a second extraction means for extracting the response data using the association et OF FURTHER dictionary to the selected character,
Speech synthesis means for performing speech synthesis based on the extracted response data and the selected character;
A system using a database, including output means for outputting voice data synthesized by voice synthesis. A first database including a plurality of basic dictionaries configured to be additionally writable and provided corresponding to a plurality of selectable characters and storing response data for each character, and associated with the plurality of characters, A method using a second database including an additional dictionary in which the response data used instead of the basic dictionary corresponding to the selected character is stored,
Receiving a request from a user;
Receiving a character selection;
Executing a response process in accordance with the accepted request,
The step of executing the response process includes:
Extracting the response data using the basic dictionary of the first database provided corresponding to the selected character when the received request is the first request;
If the accepted request is a second request, extracting the reply data using the additional dictionary of the second database was association et al to the selected character,
Synthesizing speech based on the extracted response data and the selected character;
A method of using a database, comprising: outputting speech synthesized speech data. A plurality of basic dictionaries provided in an external device, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And a terminal using a database associated with the plurality of characters and including an additional dictionary storing the response data used for response processing to the second request instead of the basic dictionary corresponding to the selected character There,
Accepting means for accepting a first or second request from a user;
Selection accepting means for accepting selection of a character;
Output speech data synthesized by speech based on response data extracted using the database in response to the first or second request received by the receiving means and the character selected by the selection receiving means. And a terminal. A plurality of basic dictionaries provided in the main body, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And the response data used in response processing to the second request instead of the basic dictionary associated with the plurality of characters and associated with the selected character is stored in the first database including A terminal that uses the second database including the additional dictionary,
Accepting means for accepting a first or second request from a user;
Selection accepting means for accepting selection of a character;
Response executing means for executing response processing in accordance with the request received by the receiving means,
The response execution means includes
Extracting means for extracting response data using the basic dictionary of the first database provided corresponding to the selected character when the receiving means receives the first request;
Obtaining means for obtaining response data extracted using an additional dictionary of the second database associated with the selected character when the accepting means accepts a second request;
Speech synthesis means for performing speech synthesis based on the extracted or acquired response data and the character selected by the selection receiving means;
A terminal comprising: output means for outputting voice-synthesized voice data. A first database including a plurality of basic dictionaries provided in an external device and provided corresponding to a plurality of selectable characters and storing response data used for response processing to the first request for each character And an additional dictionary that is provided in the main body, is associated with the plurality of characters, and stores the response data used for response processing to the second request instead of the basic dictionary corresponding to the selected character. A terminal that uses the database in FIG.
Accepting means for accepting a first or second request from a user;
Selection accepting means for accepting selection of a character;
Response executing means for executing response processing in accordance with the request received by the receiving means,
The response execution means includes
Obtaining means for obtaining response data using the basic dictionary of the first database provided corresponding to the selected character when the accepting means accepts the first request;
An extraction means for extracting response data using an additional dictionary of the second database associated with the selected character when the receiving means receives a second request;
Speech synthesis means for performing speech synthesis based on the extracted or acquired response data and the character selected by the selection receiving means;
A terminal comprising: output means for outputting voice-synthesized voice data. A plurality of basic dictionaries provided in the main body, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And a terminal using a database associated with the plurality of characters and including an additional dictionary storing the response data used for response processing to the second request instead of the basic dictionary corresponding to the selected character There,
Accepting means for accepting a first or second request from a user;
Selection accepting means for accepting selection of a character;
Response executing means for executing response processing in accordance with the request received by the receiving means,
The response execution means includes
When the first request received by said receiving means, first extracting means for extracting the reply data using the basic dictionary of the database provided corresponding to the selected character,
If the second request accepted by the accepting means, second extracting means for extracting the reply data using the additional dictionary of the database that is associated with the selected character,
Speech synthesis means for synthesizing speech based on the extracted response data and the character selected by the selection receiving means;
A terminal comprising: output means for outputting voice-synthesized voice data. A plurality of basic dictionaries provided in an external device, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And a terminal using a database including the additional dictionary in which the response data used in response to the second request is used instead of the basic dictionary corresponding to the selected character and associated with the plurality of characters. A terminal program for causing a computer to execute,
The terminal program accepts a first or second request from a user to the computer;
Receiving a character selection;
Outputting the voice data synthesized based on the response data extracted using the database in response to the accepted first or second request and the selected character, and executing the process. Terminal program to make it function like A plurality of basic dictionaries provided in the main body, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And the response data used in response processing to the second request instead of the basic dictionary associated with the plurality of characters and associated with the selected character is stored in the first database including A terminal program for causing a terminal computer using the second database including the additional dictionary to be executed,
The terminal program is stored in the computer.
Receiving a first or second request from a user;
Receiving a character selection;
And executing a response process according to the received request,
The step of executing the response process includes:
Extracting response data using the basic dictionary of the first database provided corresponding to the selected character when receiving the first request;
Obtaining response data extracted using the additional dictionary of the second database associated with the selected character when receiving the second request;
Synthesizing speech based on the extracted or acquired response data and the selected character;
A terminal program that causes a process to be executed, including the step of outputting voice-synthesized voice data. A first database including a plurality of basic dictionaries provided in an external device and provided corresponding to a plurality of selectable characters and storing response data used for response processing to the first request for each character And an additional dictionary that is provided in the main body, is associated with the plurality of characters, and stores the response data used for response processing to the second request instead of the basic dictionary corresponding to the selected character. A terminal program for causing a computer of a terminal using the database of 2 to be executed,
The terminal program is stored in the computer.
Receiving a first or second request from a user;
Receiving a character selection;
And executing a response process according to the received request,
The step of executing the response process includes:
Obtaining response data using the basic dictionary of the first database provided corresponding to the selected character when receiving the first request;
Extracting response data using an additional dictionary of the second database associated with the selected character when receiving a second request;
Synthesizing speech based on the extracted or acquired response data and the selected character;
A terminal program that causes a process to be executed, including the step of outputting voice-synthesized voice data. A plurality of basic dictionaries provided in the main body, configured to be additionally writable, provided corresponding to a plurality of selectable characters, and storing response data used for response processing to the first request for each character And a terminal using a database including the additional dictionary in which the response data used in response to the second request is used instead of the basic dictionary corresponding to the selected character and associated with the plurality of characters. A terminal program for causing a computer to execute,
The terminal program is stored in the computer.
Receiving a first or second request from a user;
Receiving a character selection;
And executing a response process according to the received request,
The step of executing the response process includes:
A step when receiving a first request, extracts the response data by using a basic dictionary of the database provided corresponding to the selected character,
A step when receiving the second request, extracts the response data by using the additional dictionary of the database that is associated with the selected character,
Synthesizing speech based on the extracted response data and the selected character;
A terminal program that causes a process to be executed, including the step of outputting voice-synthesized voice data. An audio data output device that uses a database configured to be appendable,
A receiving means for receiving a request from a user;
Selection accepting means for accepting selection of a character;
Response executing means for executing response processing in accordance with the request received by the receiving means,
The database includes a plurality of basic dictionaries that are respectively provided corresponding to a plurality of selectable characters and in which response data is stored for each character.
An additional dictionary receiving unit that receives an additional dictionary storing the response data used instead of the basic dictionary corresponding to the selected character;
A registration unit for registering a correspondence table representing a correspondence relationship between the plurality of characters and the additional dictionary together with the additional dictionary in the database according to reception of the additional dictionary of the additional dictionary reception unit,
The response execution means includes
First extraction means for extracting response data using a basic dictionary provided corresponding to the selected character when the received request is a first request;
A second extraction means for extracting response data using an additional dictionary corresponding to the selected character with reference to the correspondence table when the accepted request is a second request;
Speech synthesis means for performing speech synthesis based on the extracted response data and the selected character;
An audio data output device including output means for outputting the synthesized voice data.

Images