JP2002215186A - Speech recognition system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a speech recognition system which can reduce restrictions in respect of information size for a lexical dictionary used for speech recognition in a terminal station which receives a voice input. SOLUTION: The speech recognition system arranges the lexical dictionary used for speech recognition in the state where divided into a plurality of lexical dictionaries in a service center 2. Each of the divided lexical dictionaries is transmitted to an on-vehicle system 1 which really performs speech recognition and is used for speech recognition as necessary from the service center 2 through a communication line.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a speech recognition system.

[0002]

2. Description of the Related Art In recent years, navigation devices having a voice recognition system have been proposed. In such a conventional navigation device, all databases such as a vocabulary dictionary used for speech recognition are stored in advance in a database storage means in the navigation device. Therefore, in this navigation device, in order to increase the number of vocabulary words that can be recognized by speech to a certain level or more, it is necessary to prepare a vocabulary dictionary of a certain vocabulary level or more in the navigation device from the initial stage. Since the database used for recognition has a large information size and a large burden on the database storage means, the contents of the vocabulary dictionary and the like of the database are narrowed down as much as possible.

[0003]

For this reason, the conventional navigation device has a problem that the information size of the vocabulary dictionary is greatly restricted.

[0004] In view of the above problems, it is an object of the present invention to provide a speech recognition system capable of reducing restrictions on the information size of a vocabulary dictionary used for speech recognition in a terminal station or the like that accepts speech input.

[0005]

The technical means for achieving the above object comprises a base station and a terminal station, wherein the base station comprises a plurality of vocabulary information used for speech recognition performed by the terminal station. Database storage means for storing a plurality of vocabulary dictionaries, each comprising: a terminal that can be connected to the terminal station via a communication line, and in response to a transmission request sent from the terminal station via the communication line, Information processing means for reading a vocabulary dictionary from the database storage means in units of the vocabulary dictionary and transmitting the vocabulary dictionary to the terminal station via the communication line, wherein the terminal station receives voice input, Storage means capable of rewriting the storage content, and connectable to the base station via the communication line, and transmits the transmission request to the base station via the communication line, The vocabulary dictionary transmitted from the base station via the communication line in response to the transmission request is stored in the storage unit, while the voice input received by the voice input reception unit is stored in the storage unit. And a voice recognition unit that checks each vocabulary information included in the vocabulary dictionary and selects vocabulary information corresponding to the voice input as candidate vocabulary information.

Further, the technical means for achieving the above object includes a base station and a terminal station, wherein the base station includes a plurality of vocabulary information items each used for speech recognition performed by the terminal station. Database storage means for storing the vocabulary dictionary of the terminal, the terminal station can be connected to the terminal station via a communication line, and responds to a transmission request sent from the terminal station via the communication line. Information processing means for reading the vocabulary dictionary from the database storage means in units of the vocabulary dictionary and transmitting the vocabulary dictionary to the terminal station via the communication line, wherein the terminal station receives a voice input. And storage means capable of rewriting the storage content, and connectable to the base station via the communication line, and wherein the base station has been transmitted from the base station via the communication line. While the vocabulary dictionary is stored in the storage unit, the voice input received by the voice input reception unit is compared with the respective vocabulary information included in the vocabulary dictionary stored in the storage unit, and the vocabulary dictionary corresponding to the voice input is matched. When vocabulary information is present in the vocabulary dictionary stored in the storage unit, the corresponding vocabulary information is selected as candidate vocabulary information, while vocabulary information corresponding to the voice input is stored in the storage unit. When not present in the vocabulary dictionary, the request signal requesting transmission of the vocabulary dictionary that may include vocabulary information corresponding to the voice input is transmitted to the base station via the communication line, Checking each vocabulary information included in the vocabulary dictionary transmitted from the base station via the communication line in response to the request signal,
Voice recognition means for selecting vocabulary information corresponding to the voice input as candidate vocabulary information.

[0007] More preferably, the terminal station is mounted on a mobile unit, and the terminal station further includes a position detecting unit for detecting a current position of the mobile unit, and the voice recognition unit of the terminal station. Transmits the position information indicating the current position detected by the position detection means to the base station via the communication line, and from the base station via the communication line in response to the transmission of the position information. The information processing device of the base station further includes a function of storing the transmitted vocabulary dictionary in the storage unit, and the information processing device of the base station receives the position information in response to the position information being transmitted through the communication line. The information processing apparatus may further include a function of reading the vocabulary dictionary corresponding to the current position indicated by information from the database storage unit and transmitting the dictionary to the terminal station via the communication line.

Preferably, the database storage means of the base station stores geographically related vocabulary information having relevance with a geographical position among a plurality of vocabulary information included in each of the stored vocabulary dictionaries. Is capable of storing position specifying information for specifying a geographical position indicating the association in association with the geographically relevant vocabulary information, and the information processing means stores the geographically relevant vocabulary information in the terminal station. When transmitting, the location specifying information associated with the geographically related vocabulary information is preferably read out from the database storage means and transmitted to the terminal station.

Further, preferably, the database storage means can store network connection service information for acquiring information related to the geographically related vocabulary information via a communication network in association with the geographically related vocabulary information. The information processing means, when transmitting the geographically relevant vocabulary information to the terminal station, reads out the network connection service information associated with the geographically relevant vocabulary information from the database storage means together with It is good to transmit to a terminal station.

[0010] Preferably, the voice recognition system is used for voice recognition of voice input to a navigation device mounted on a vehicle.

[0011] Further, a technical means for achieving the above object is a voice recognition system used for voice recognition of voice input to a navigation device mounted on a vehicle, comprising: a base station and a terminal station. The base station includes a database storage unit that stores a vocabulary dictionary including a plurality of vocabulary information used for speech recognition, and is connectable to the terminal station via a communication line, and the terminal station is connected to the terminal via the communication line. Recognize a voice pattern of a voice signal transmitted from a station, collate the voice pattern with each of the vocabulary information included in the vocabulary dictionary stored in the database storage unit, and correspond to the recognized voice pattern. Voice recognition means for selecting vocabulary information to be performed as candidate vocabulary information and transmitting the vocabulary information to the terminal station via the communication line. Voice input receiving means for receiving input, the voice signal of the voice input which is connectable to the base station via the communication line, and which is received by the voice input receiving means, to the base station via the communication line. And information processing means for receiving the candidate vocabulary information transmitted from the base station via the communication line while transmitting.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS <First Embodiment> FIGS. 1 and 2
1 is a block diagram of a speech recognition system according to a first embodiment of the present invention. As shown in FIGS. 1 and 2, this voice recognition system includes an in-vehicle system (terminal station) 1 mounted on a vehicle and a service center (base station) 2.

The in-vehicle system 1 also has a function as a navigation device, as will be described later. The input to the in-vehicle system 1 is made by voice input via the microphone 11 or operation input via the operation input unit 15. The output from the in-vehicle system 1 is performed by audio output via the speaker 12 or display output via the display unit 16. Specific input contents include, for example, voice or operation input for setting a destination, and output contents include, for example, voice or image output for route guidance to the destination. .

As shown in FIG. 1, the in-vehicle system 1 has a microphone (voice input receiving means) 11 for receiving voice input.
A speaker 12 for performing voice output, a voice processing unit 13 for performing voice recognition and voice synthesis, a navigation unit 14 for performing communication processing and navigation processing with the service center 2, and an operation input unit 15 for receiving an operation input. And a display unit 16 for performing display output, and a communication unit (such as a mobile phone) 17. The voice processing unit 13 includes a storage unit (storage unit) 13a, a voice recognition unit 13
b and a voice synthesizer 13c. The navigation unit 14 includes a communication processing unit 14a and a navigation processing unit (position detection unit) 14b. Among such configurations, the voice recognition unit 13b, the communication processing unit 14
a and the communication unit 17 correspond to a voice recognition unit according to the present invention.

As shown in FIG. 2, the service center 2
A communication unit 21, a computer 22, and a database storage device (database storage unit) 23 are provided. Among them, the communication unit 21 and the computer 22 correspond to information processing means according to the present invention.

Here, the database used for speech recognition is generally used for recognizing a pronunciation pattern of an inputted speech input (what combination of vowels and consonants the speech input is composed of). And a vocabulary database for recognizing a word (vocabulary) corresponding to the recognized voice pattern.

In this embodiment, among these databases, the pattern recognition database is prepared in the storage unit 13a of the vehicle-mounted system 1 from the initial stage. The vocabulary database is prepared in the database storage device 23 of the center 2, and then the necessary part of the vocabulary database is downloaded from the service center 2 to the vehicle-mounted system 1 as needed.

In response to this, in the present embodiment, the vocabulary database is composed of a plurality of separable vocabulary dictionaries, and can be downloaded in vocabulary dictionary units.
Each vocabulary dictionary includes a plurality of vocabulary information. Each vocabulary information includes word information of the word corresponding to the vocabulary information (text data of the word, etc.).

In this embodiment, the service center 2
In the database storage device 23, geographically-related vocabulary information (for example, vocabulary information indicating a place name or a facility name) having a relevance to a geographical position among a plurality of vocabulary information included in each vocabulary dictionary stored therein. With regard to, position specifying information (such as a coordinate value on a map and an address) for specifying a geographical position indicating the relevancy can be stored in association with the geographically relevant vocabulary information.

Further, in the present embodiment, the database storage device 23 stores the information related to the geographically related vocabulary information via a communication network (such as the Internet) in association with the geographically related vocabulary information included in each vocabulary dictionary. Network connection service information (such as address information) to be acquired can be stored.

First, each configuration of the service center 2 will be described. The communication unit 21 can be connected to the in-vehicle system 1 via a communication line (such as a communication network using a wireless telephone line) under the control of the computer 22. The database storage device 23 stores the above-mentioned vocabulary database, and location specifying information and network connection service information associated with geographically related vocabulary information in the vocabulary database.

When a transmission request described later is transmitted from the in-vehicle system 1 via the communication line of the communication unit 21, the computer 22 includes a vocabulary dictionary corresponding to the transmission request (or includes vocabulary information corresponding to the transmission request). The vocabulary dictionary (or vocabulary dictionary) is read from the database storage device 23 and transmitted to the in-vehicle system 1 via the communication line of the communication unit 21.

When position information indicating the current position of the vehicle is transmitted from the in-vehicle system 1 via the communication line of the communication unit 21 to the computer 22, the vocabulary corresponding to the current position indicated by the position information is sent to the computer 22. A dictionary (or a vocabulary dictionary including vocabulary information corresponding to the current position) is read from the database storage device 23 and transmitted to the in-vehicle system 1 via the communication line of the communication unit 21. Here, as the vocabulary dictionary corresponding to the current position transmitted from the service sensor 2 to the in-vehicle system 1, for example, vocabulary information indicating a place name or a facility name of the current position (or a word meaning related to the place name or the facility name). The vocabulary dictionary containing (or possibly containing) the information) is selected for transmission.

When the computer 22 transmits the vocabulary dictionary, if the location specifying information and the network connection information are associated with the geographically related vocabulary information included in the vocabulary dictionary, the associated location specifying information and The network connection information is also read out from the database storage device 23 and transmitted to the in-vehicle system 1.

Next, each configuration of the vehicle-mounted system 1 will be described. The storage unit 13a is a rewritable storage device, and includes a hard disk, a large-capacity semiconductor memory, and the like. In this storage unit 13a,
The above-described pattern recognition database is stored from an initial stage, and the vocabulary dictionary downloaded from the service center 2 is stored. The storage unit 13a also stores geographical position information and network connection service information described below downloaded together with the vocabulary information. Further, the storage unit 13a also stores a speech synthesis database used by the speech synthesis unit 13c for speech synthesis.

The speech recognition unit 13b performs a speech recognition process in response to a speech recognition request given from a navigation processing unit 14b described later. In the voice recognition processing, first, the voice pattern of the voice input (voice signal) received by the microphone 11 is recognized using the pattern recognition database of the storage unit 13a. The recognition of the voice pattern is performed by, for example, a method based on a hidden Markov model.

Subsequently, when the voice pattern is recognized, each vocabulary information (word) included in the vocabulary dictionary stored in the storage unit 13a is compared with the recognized voice pattern, and the recognized voice pattern is checked. Is stored in the storage unit 13a, a word corresponding to the vocabulary information is selected as a candidate word and provided to the navigation processing unit 14b. On the other hand, if the vocabulary information corresponding to the recognized voice pattern does not exist in the storage unit 13a as a result of the comparison, the vocabulary dictionary including (or possibly including) the vocabulary information corresponding to the voice pattern is transmitted. The requested transmission request is transmitted to the service center 2 via the communication processing unit 14a. The transmission request may include the recognized voice pattern.

Subsequently, when the vocabulary dictionary transmitted from the service center 2 in response to the transmission request is received via the communication processing unit 14a, the vocabulary dictionary is stored in the storage unit 13a as a part of the vocabulary database. On the other hand, the recognized voice pattern is matched with each vocabulary information included in the received vocabulary dictionary, and a word corresponding to the vocabulary information corresponding to the recognized voice pattern is sent to the navigation processing unit 14b as a candidate word. Given.

Further, when given the position information indicating the current position from the navigation processing unit 14b, the speech recognition unit 13b determines whether or not a vocabulary dictionary including vocabulary information corresponding to the position information exists in the storage unit 13a. And if it does not exist, it instructs the communication processing unit 14a to transmit its position information to the service center 2 via the communication line. Then, the vocabulary dictionary transmitted from the service center 2 via the communication line in response to the transmission of the position information is received via the communication processing unit 14 and stored in the storage unit 13a.

Further, the voice recognition unit 13 b
It also has a function of deleting unnecessary vocabulary dictionaries in “a” (for example, vocabulary dictionaries that have not been used for a certain period or more). The specification of the vocabulary dictionary to be deleted and the execution of the deletion may be instructed by voice input or operation input via the operation input unit 15 and the navigation processing unit 14b.

The voice synthesizing unit 13c performs a voice synthesizing process in response to a voice synthesis request given from the navigation processing unit 14b. In the speech synthesis processing, speech content information indicating the content of speech to be synthesized provided together with the speech synthesis request (here, character information represented by text data or the like)
Are synthesized using the voice synthesis database in the storage unit 13a and output as voice via the speaker 12.

The operation input section 15 is for receiving an operation input to the navigation unit 14. The display unit 16 is used for navigation display of a current position, a route to a destination, and the like. The communication unit 17 is configured by a mobile phone or the like, and can be connected to the service center 2 via a communication line under the control of the communication processing unit 14a.

The communication processing unit 14a can be connected to the service center 2 and various websites via a communication line by the communication unit 17. Then, the communication processing unit 14
a, when a transmission request is given from the voice recognition unit 13b,
The transmission request is transmitted to the service center 2 via the communication unit 17, and the vocabulary dictionary transmitted from the service center 2 in response to the transmission request is transmitted to the speech recognition unit 13b.
Transfer to In addition, when the communication processing unit 14a receives a website connection request from the navigation processing unit 14b, the communication processing unit 14b
Is connected to the website indicated by the address information given by the user.

Further, the communication processing unit 14a includes the voice recognition unit 1
3b, the position information indicating the current position is transmitted to the communication unit 1
7 to the service center 2 via the communication line, and receives the vocabulary dictionary transmitted from the service center 2 via the communication line in accordance with the transmission of the position information, and transfers it to the speech recognition unit 13b.

The navigation processing section 14b is a functional element for performing normal navigation processing, and has a built-in recording device (DVD-ROM and DVD-R) not shown.
In the OM driver or the like, navigation information used for the navigation processing performed by the navigation processing unit 14b is recorded. The navigation information includes display road information used to display the road shape on the display unit 16, a plurality of character information indicating the contents of voice output to be output for voice route guidance and the like, and a destination. And the like, which is used for calculating the route of the route.

The navigation processing unit 14b includes:
Although a detailed configuration is not shown, a current position detection unit that detects the current position of the own vehicle, a route search unit that searches for a route from a specific position (such as the current position) to the destination, and corresponds to input place names and facility names And a network connection unit for connecting to a web site on the net related to the input place name and facility name.

The current position detecting section detects the current position of the own vehicle on the map using at least one of the GPS and the autonomous navigation (here, both), and detects the current position of each area recorded in the recording device. Using the display road information, a map image of an area including the current position is created and displayed on the display unit 16.
And the current position is displayed on the map image. In addition, the current position detection unit performs position position information indicating the detected current position (information indicating a place name or the like of the current position).
Is given to the voice recognition unit 14a.

The route search unit searches for a route from the current position (or a specific position) to the destination using the route search road information stored in the recording device.
And a voice synthesis request for voice route guidance and character information are output to the voice synthesis unit 13c as needed. The input of the specific position and the destination is performed by a voice input through the microphone 11 or an operation input through the operation input unit 15. When a voice input is received by the navigation processing unit 14b, the navigation processing unit 14b
A voice recognition request is sent to the voice recognition unit 13b, and the input content is recognized based on the content of the candidate word given from the voice recognition unit 13b in response to the request.

The geographic guidance section performs a guidance process for guiding a geographical position corresponding to a place name or a facility name input by voice input or the like via the voice recognition section 13b. In the guidance process, for example, the location name and the location of the facility are displayed on the display unit 16 using the position identification information stored in the storage unit 13a.

The network connection unit responds to a connection instruction given by a voice input via the voice recognition unit 13b or an operation input via the operation input unit 15 to communicate with the communication processing unit 14a.
To a website connection request, and connects to a website on the net related to a place name or facility name input by voice input or the like via the voice recognition unit 13b via the communication processing unit 14a. This website is connected to the storage unit 13a
Is used.

With such a configuration, the vocabulary database used for speech recognition in the vehicle-mounted system 1 is prepared in the database storage device 23 of the service center 2 at an initial stage, and thereafter, the vocabulary database is provided via the communication line. In response to a transmission request transmitted from the service center 1 to the service center 2, a vocabulary dictionary corresponding to the transmission request is downloaded from the service center 2 to the vehicle-mounted system 1 as appropriate, and is used for voice recognition.

For example, when a word that does not exist in each vocabulary dictionary stored in the storage unit 13a of the in-vehicle system 1 is input by voice input, a transmission request for transmitting a vocabulary dictionary including the word is issued. Is transmitted from the in-vehicle system 1 to the service center 2, and in response to the transmission request, a corresponding vocabulary dictionary is transmitted from the service center 2 to the in-vehicle system 1, and voice recognition is performed using the vocabulary dictionary. I have.

As described above, according to the present embodiment, the vocabulary dictionary used for speech recognition is prepared in the service center 2 in a state where the vocabulary dictionary is divided into a plurality of vocabulary dictionaries. The information is transmitted as needed from the service center 2 to the vehicle-mounted system 1 that actually performs voice recognition via a line, and is stored in the storage unit 13a of the vehicle-mounted system 1 from the initial stage as in the related art. Since it is not necessary to prepare a vocabulary dictionary of a certain level or higher, the information size of the vocabulary dictionary used for speech recognition in the in-vehicle system 1 can be reduced while suppressing an increase in cost due to an increase in the capacity of the storage unit 13a of the in-vehicle system 1. Can be reduced.

The storage unit 13a of the in-vehicle system 1 can rewrite the stored contents, so that unnecessary vocabulary dictionaries and the like can be deleted, and the storage capacity can be used effectively. A single user-specific vocabulary database can be created.

Further, since the vocabulary dictionary corresponding to the current position of the vehicle on which the vehicle-mounted system 1 is mounted is automatically sent from the service center 2 to the vehicle-mounted system 1, the place names and facilities corresponding to the current position are provided. Speech can be input using names and their abbreviations, and speech recognition can be performed smoothly.

Further, since the geographical vocabulary information provided from the service center 2 to the vehicle-mounted system 1 can be provided with position specifying information for specifying a geographical position related to the geographical vocabulary information, On the in-vehicle system 1, the geographical position related to the geographically related vocabulary information can be easily specified based on the position specifying information, which is convenient.

Further, the geographic-related vocabulary information provided from the service center 2 to the in-vehicle system 1 can be provided with network connection service information for acquiring information related to the geographic-related vocabulary information via a communication network. Therefore, the in-vehicle system 1 can easily obtain information for obtaining information related to the geographically related vocabulary information via the communication network based on the network connection service information, which is convenient. .

In this embodiment, when the vocabulary information corresponding to the recognized voice pattern does not exist in the storage unit 13a, the voice recognition unit 13b outputs a transmission request to the service center 2. , At the time of initial startup, etc., the voice recognition unit 13b or the navigation processing unit 1
4b transmits a transmission request for requesting transmission of one or more vocabulary dictionaries corresponding to the type (configuration) of the vehicle-mounted system 1;
You may make it transmit to the service center 2 via the communication processing part 14a.

Thus, a plurality of vocabulary dictionaries corresponding to each type of the vehicle-mounted system 1 are prepared in the database storage device 23 of the service center 2 and the type of the vocabulary dictionary supplied from the service center 2 to the vehicle-mounted system 1 is changed. By simply doing so, it is possible to easily cope with differences in the type of the in-vehicle system 1 (including those due to differences in the type and grade of the vehicle, etc.), and a general-purpose system can be configured.

<Second Embodiment> FIGS. 3 and 4 are block diagrams of a speech recognition system according to a second embodiment of the present invention. As shown in FIGS. 3 and 4, this voice recognition system includes an in-vehicle system (terminal station) 1 mounted on a vehicle.
01 and a service center (base station) 102.

The in-vehicle system 101 also has a function as a navigation device, as will be described later, and the input / output format of the in-vehicle system 101 is the same as in the first embodiment.

As shown in FIG. 3, the in-vehicle system 101 is a microphone (voice input receiving means) for receiving a voice input.
111, a speaker 112 for outputting voice, a voice processing unit 113 for controlling voice synthesis, and a service center 10
A navigation unit 114 that performs a communication process and a navigation process between the navigation unit 2 and an operation input unit 115 that receives an operation input; a display unit 116 that performs display output;
And a communication unit (eg, a mobile phone) 117. The voice processing unit 113 includes a voice synthesis unit 113a and a storage unit 113b. Navigation unit 114
Are the communication processing unit 114a and the navigation processing unit 1
14b. In such a configuration, the communication processing unit 114a and the communication unit 117 correspond to the communication processing unit according to the present invention.

As shown in FIG. 4, the service center 102 includes a communication section 121, a computer 122, and a database storage device (database storage means) 123. The communication unit 121 and the computer 12
Reference numeral 2 corresponds to the voice recognition means according to the present invention.

In the present embodiment, the voice signal of the voice input input by the on-vehicle system 101 is transmitted to the service center 102 via the communication line, and is transmitted to the database storage device 12.
The voice recognition is performed in the service center 102 using the pattern recognition database and the vocabulary database stored in the storage 3, and the recognition result is sent to the in-vehicle system 101 via the communication line.

First, each configuration of the service center 102 will be described. The communication unit 121 can be connected to the in-vehicle system 101 via a communication line (such as a communication network using a wireless telephone line) under the control of the computer 122. The database storage device 123 stores the same pattern recognition database and vocabulary database as those in the first embodiment.

When a voice signal is transmitted from the in-vehicle system 101 via the communication line of the communication unit 121, the computer 122 uses the pattern recognition database and the vocabulary database stored in the database storage device 123 to transmit the voice signal. It recognizes candidate vocabulary information (candidate words) corresponding to the voice signal, and transmits the recognized vocabulary information to the in-vehicle system 101 via the communication line of the communication unit 121. Note that the voice recognition method itself is the same as in the first embodiment, and a description thereof will not be repeated.

Next, each configuration of the vehicle-mounted system 101 will be described. The storage unit 113b includes a speech synthesis unit 113a
Stores a speech synthesis database used for speech synthesis. The configurations and functions of the voice synthesis unit 113a, the operation input unit 115, the display unit 116, and the communication unit 117 are the same as those of the first embodiment.
This is the same as the voice synthesis unit 13c, the operation input unit 15, the display unit 16, and the communication unit 17 according to the embodiment.

The communication processing unit 114a is connectable to the service center 102 via a communication line provided by the communication unit 117. In response to the transmission request and the reception request given from the navigation processing unit 114b, the voice received by the microphone 111 is transmitted. The input audio signal is transmitted to the service center 102 via the communication line provided by the communication unit 117, while the service center 10 is transmitted via the communication line provided by the communication unit 117.
2 is provided to the navigation processing unit 114b.

The navigation processing section 114b is a functional element for performing normal navigation processing, and has a built-in recording device (DVD-ROM and DVD-ROM) not shown.
The same navigation information as in the first embodiment is recorded in the ROM driver and the like. The basic configuration and functions of the navigation processing unit 114b relating to current position detection and route guidance are the same as those in the first embodiment.

When a voice input is accepted by the navigation processing unit 114b, the navigation processing unit 114b sends a transmission request to the communication processing unit 114a. Then, the voice signal is transmitted to the service center 102 by the communication processing unit 114a, and in response thereto, candidate vocabulary information is transmitted from the service center 102 to the communication processing unit 11
4a, the candidate vocabulary information is received from the communication processing unit 114a, and the input contents are recognized based on the contents of the candidate vocabulary information.

As described above, according to the present embodiment, the voice signal of the voice input received by the in-vehicle system 101 is sent to the service center 102 via the communication line, and the service center 102 performs voice recognition to recognize the voice. The result is returned to the in-vehicle system 101 via the communication line,
Since various databases for voice recognition need only be provided in the service center 102, a vocabulary dictionary with a sufficient vocabulary level can be prepared in the service center 102 without increasing the load on the in-vehicle system 101. Restrictions on the information size of the vocabulary dictionary used for the information processing can be reduced.

In each of the above-described embodiments, the voice recognition of the voice input for navigation is performed by the voice recognition system according to the present invention. The voice recognition of the voice input to other in-vehicle devices may be performed.

[0063]

According to the first to sixth aspects of the present invention, a vocabulary dictionary used for speech recognition is prepared in the base station in a state divided into a plurality of vocabulary dictionaries, and each of the divided vocabulary dictionaries is prepared. It is transmitted from the base station to the terminal station that actually performs voice recognition as needed via the communication line as needed. Since it is not necessary to prepare the vocabulary dictionary described above, it is possible to suppress the cost increase due to an increase in the capacity of the storage means for storing the vocabulary dictionary of the terminal station, and to reduce the cost of the vocabulary dictionary used for speech recognition in the terminal station that performs voice input. Restrictions on information size can be reduced.

Further, the storage device of the terminal station can rewrite the stored contents, so that unnecessary vocabulary dictionaries and the like can be deleted, the storage capacity can be used effectively, and the user of the terminal station can use it. You can create your own vocabulary database.

Further, when the terminal station is mounted on a vehicle and is used to receive voice input to on-vehicle equipment mounted on the vehicle, voice input to a plurality of types of on-vehicle equipment having different types and grades is performed. A plurality of vocabulary dictionaries corresponding to each type of in-vehicle device for recognizing the information are prepared in the database storage means of the base station, and the terminal station is installed from among the vocabulary dictionaries prepared in the base station By selecting a vocabulary dictionary corresponding to the type of in-vehicle equipment mounted on the vehicle to be transmitted from the base station to the terminal station, it is possible to supply a minimum required vocabulary dictionary from the base station to the terminal station. By simply changing the type of vocabulary dictionary supplied from the base station to the terminal station, it is possible to easily deal with differences in the types of in-vehicle devices mounted on vehicles (including those due to differences in vehicle types and grades). It can be, it is possible to construct a general-purpose system.

According to the third aspect of the present invention, the vocabulary dictionary corresponding to the current position of the moving object on which the terminal station is mounted is automatically sent from the base station to the terminal station. It is possible to smoothly perform voice recognition for voice input such as a place name or a facility name corresponding to a position.

According to the fourth aspect of the present invention, the geographic-related vocabulary information provided from the base station to the terminal station is provided with position specifying information for specifying a geographical position related to the geographic-related vocabulary information. Since it is possible, the terminal station can easily specify the geographical position related to the geographically relevant vocabulary information based on the position specifying information, which is convenient.

According to the fifth aspect of the present invention, the geographically related vocabulary information provided from the base station to the terminal station has a network connection for acquiring information related to the geographically related vocabulary information via the communication network. Since service information can be added, the terminal station can easily obtain information for acquiring information related to geographically related vocabulary information via a communication network based on the network connection service information. Is convenient.

According to the present invention, the voice signal of the voice input received by the terminal station is transmitted to the base station via the communication line, the base station performs voice recognition, and the recognition result is transmitted to the communication line. Is returned to the terminal station via, and only the base station needs to have a vocabulary dictionary for speech recognition,
A vocabulary dictionary with a sufficient vocabulary level can be prepared in the base station without increasing the load on the terminal station, and restrictions on the information size of the vocabulary dictionary used for speech recognition can be reduced.

[Brief description of the drawings]

FIG. 1 is a block diagram of a voice input system according to a first embodiment of the present invention.

FIG. 2 is a block diagram of the voice input system according to the first embodiment of the present invention.

FIG. 3 is a block diagram of a voice input system according to a second embodiment of the present invention.

FIG. 4 is a block diagram of a voice input system according to a second embodiment of the present invention.

[Explanation of symbols]

 1,101 In-vehicle system 2,102 Service center 11,111 Microphone 12,112 Speaker 13a, 113b Storage unit 13b Voice recognition unit 13c, 113a Voice synthesis unit 14a, 114a Communication processing unit 14b, 114b Navigation processing unit 15,115 Operation input Unit 16, 116 Display unit 17, 117 Communication unit 21, 121 Communication unit 22, 122 Computer 23, 123 Database storage device

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI theme coat ゛ (Reference) G10L 15/28 G10L 3/00 551A 571A (72) Inventor Akinori Usami 1-7 Kikuzumi, Minami-ku, Nagoya-shi, Aichi Prefecture No. 10 Inside the Auto Network Engineering Laboratory Co., Ltd. F term (reference) 2F029 AA02 AB07 AC14 AC18 5D015 GG03 KK01 KK02 LL00

Claims (7)

[Claims] 1. A base station comprising: a base station; and a terminal station, wherein the base station stores a plurality of vocabulary dictionaries each including a plurality of vocabulary information used for speech recognition performed by the terminal station; The vocabulary dictionary can be connected to the terminal station via a line, and in response to a transmission request sent from the terminal station via the communication line, the vocabulary dictionary is read from the database storage unit in the vocabulary dictionary unit and Information processing means for transmitting to the terminal station via a communication line, the terminal station comprising: a voice input receiving means for receiving a voice input; a storage means capable of rewriting storage contents; and Connected to the base station via the communication line, transmits the transmission request to the base station via the communication line, and responds to the transmission request via the communication line. The vocabulary dictionary transmitted from the station is stored in the storage means, and the vocabulary information received by the voice input reception means is compared with the respective vocabulary information included in the vocabulary dictionary stored in the storage means. A speech recognition unit for selecting vocabulary information corresponding to the speech input as candidate vocabulary information. 2. A database storage means comprising: a base station; and a terminal station, wherein the base station stores a plurality of vocabulary dictionaries each including a plurality of vocabulary information used for speech recognition performed by the terminal station; The database can be connected to the terminal station via a line, and in response to a transmission request transmitted from the terminal station via the communication line, the vocabulary dictionary corresponding to the transmission request is stored in the vocabulary dictionary unit. Information processing means for reading from a storage means and transmitting to the terminal station via the communication line, the terminal station comprising: a voice input receiving means for receiving a voice input; and a storage means capable of rewriting storage contents. And allowing the storage unit to store the vocabulary dictionary transmitted from the base station via the communication line and connectable to the base station via the communication line. The voice input received by the voice input receiving unit is compared with the respective vocabulary information included in the vocabulary dictionary stored in the storage unit, and the vocabulary information corresponding to the voice input is stored in the storage unit. If the vocabulary information is present in the vocabulary dictionary, the corresponding vocabulary information is selected as candidate vocabulary information, while if the vocabulary information corresponding to the voice input does not exist in the vocabulary dictionary stored in the storage means, Transmitting the request signal requesting transmission of the vocabulary dictionary that may include vocabulary information corresponding to voice input to the base station via the communication line, and responding to the request signal via the communication line Voice recognition means for comparing the vocabulary information included in the vocabulary dictionary transmitted from the base station, and selecting vocabulary information corresponding to the voice input as candidate vocabulary information. , A speech recognition system comprising: 3. The terminal station is mounted on a mobile unit, the terminal station further includes a position detecting unit that detects a current position of the mobile unit, and the voice recognition unit of the terminal station includes: The position information indicating the current position detected by the position detecting means is transmitted to the base station via the communication line, and transmitted from the base station via the communication line in response to the transmission of the position information. Further comprising a function of storing the vocabulary dictionary in the storage means, wherein the information processing device of the base station indicates the position information in response to the position information being transmitted via the communication line. 3. The apparatus according to claim 1, further comprising a function of reading the vocabulary dictionary corresponding to the current position from the database storage unit and transmitting the vocabulary dictionary to the terminal station via the communication line. A speech recognition system as described. 4. The geographically related vocabulary information having relevance to a geographical position among a plurality of pieces of vocabulary information included in each of the stored vocabulary dictionaries, wherein the database storage unit of the base station includes: Position specifying information for specifying a geographical position indicating the relevance can be stored in association with the geographically relevant vocabulary information, and the information processing means transmits the geographically relevant vocabulary information to the terminal station. In
4. The speech recognition system according to claim 1, wherein said location specifying information associated with said geographically related vocabulary information is read out from said database storage means and transmitted to said terminal station. 5. The database storage means may store network connection service information for acquiring information related to the geographically related vocabulary information via a communication network in association with the geographically related vocabulary information, The information processing means, when transmitting the geographically related vocabulary information to the terminal station,
5. The speech recognition system according to claim 4, wherein the network connection service information associated with the geographical vocabulary information is also read out from the database storage means and transmitted to the terminal station. 6. The voice recognition system according to claim 1, wherein the voice recognition system is used for voice recognition of a voice input to a navigation device mounted on a vehicle. 7. A voice recognition system used for voice recognition of voice input to a navigation device mounted on a vehicle, comprising a base station and a terminal station, wherein the base station is used for voice recognition. Database storage means for storing a vocabulary dictionary including a plurality of vocabulary information; and a voice pattern of a voice signal which is connectable to the terminal station via a communication line and transmitted from the terminal station via the communication line. Recognizing, collating the voice pattern with the respective vocabulary information included in the vocabulary dictionary stored in the database storage means, selecting vocabulary information corresponding to the recognized voice pattern as candidate vocabulary information, Voice recognition means for transmitting to the terminal station via a communication line, the terminal station comprising: a voice input receiving means for receiving a voice input; It is connectable to the base station via a communication line, and transmits the voice signal of the voice input received by the voice input receiving unit to the base station via the communication line, while transmitting the voice signal via the communication line. Information processing means for receiving the candidate vocabulary information transmitted from the base station.