JP2009175630A - Speech recognition device, mobile terminal, speech recognition system, speech recognition device control method, mobile terminal control method, control program, and computer readable recording medium with program recorded therein - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To attain a speech recognition device capable of performing highly accurate speech recognition. <P>SOLUTION: The speech recognition device 1 includes: a usage data base determination section 18 which selects a data base for outputting character information corresponding to speech featured value which is associated with positional information for indicating a position; and a speech recognition section 12 for performing speech recognition by using the data base selected by the usage data base determination section 18. Thereby, speech recognition using the data base corresponding to the present position of own device is performed, and the highly accurate speech recognition is attained. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a voice recognition device, a portable terminal, a voice recognition system, a voice recognition device control method, a portable terminal control method, a control program, and a computer-readable recording medium on which the program is recorded. It is about.

  In recent years, a speech recognition system capable of inputting a sentence by voice has been put into practical use and used as a practical system in various fields. This is because the voice signal of the voice uttered by the user is converted into digital data and compared with a predetermined pattern, using a voice recognition engine that recognizes the utterance content, and the user's voice is recognized as a sentence. Accept.

  And this speech recognition system has improved recognition performance year by year due to technological advances, and its functions are diversified (multifunctional). At the same time, the number of words to be recognized in the speech recognition system tends to increase. For example, when a voice recognition system is installed in a car navigation system, words to be recognized include functions such as operations, place names and facility names throughout the country, and the like. For this reason, the number of words to be recognized is enormous. It is very difficult to store such an enormous number of words in a dictionary storage device and to accurately and efficiently recognize the utterance (voice) from the user from the dictionary storage device.

  Furthermore, in speech recognition, misrecognition may occur due to the influence of noise in the surrounding environment and differences in the user's voice quality, volume, speech rate, and the like. Also, when a dialect or the like is uttered, if the word corresponding to the dialect or the like is not registered in the word dictionary, it becomes an unknown word and the input speech cannot be recognized correctly.

  As described above, all voice recognition systems have a restriction that words that are not registered in a dictionary in which words to be recognized are registered in advance cannot be recognized correctly.

Patent Document 1 includes an unspecified speaker recognition unit and a specific speaker recognition unit. In the case of a specific speaker, the specific speaker recognition unit is used, and in other cases. A system is described that uses a dictionary provided in each of the two recognition units by selectively switching the two recognition units, such as using an unspecified speaker recognition unit.
Japanese Patent Laid-Open No. 03-9399 (published on January 17, 1991)

  However, the configuration described in Patent Document 1 has the following problems. That is, in the configuration described in Patent Document 1, only the specific speaker recognition unit and the non-specific speaker recognition unit are selectively switched depending on whether or not the speaker is a specific speaker. Only dictionaries are used.

  Therefore, as long as the speakers are the same, the same dictionary is used regardless of the situation in which the speech recognition apparatus is used. Therefore, even if the situation in which the voice recognition device is used changes, the recognizable word does not change, and the accuracy of the voice recognition device does not improve.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a voice recognition device, a portable terminal, a voice recognition system, a voice recognition device control method, a portable terminal control method, and a control with high voice recognition accuracy. An object is to realize a program and a computer-readable recording medium on which the program is recorded.

  In order to solve the above-described problem, a speech recognition apparatus according to the present invention includes a database selection unit that selects a database that outputs character information corresponding to a feature amount of speech associated with position information indicating a position, and And voice recognition means for performing voice recognition using the database selected by the database selection means.

  The method for controlling a speech recognition apparatus according to the present invention includes a database selection step for selecting a database that outputs character information corresponding to a feature amount of speech, which is associated with position information indicating a position, and the database selection step. And a speech recognition step for performing speech recognition using the database selected in (1).

  According to said structure and method, speech recognition is performed using the database matched with the positional information which shows a position.

  Thereby, the voice recognition using the database according to the position indicated by the position information can be performed.

  Therefore, when a device (such as a microphone) that accepts speech input to the speech recognition device is present at a certain location, a speech database that is enriched with words and acoustic / language models that are considered to be frequently used at that location is used. Can be used. Therefore, more accurate speech recognition can be realized.

  For example, when speech is input in the Kansai region, it is likely that the speech is a Kansai dialect. Therefore, if the Kansai dialect acoustic / language model database is selected as the database used for speech recognition, the accuracy will be higher. Voice recognition can be performed. Also, when inputting speech at a station, it is highly likely that travel related topics are spoken, so if you select a dictionary with travel related words as a database for speech recognition, It becomes possible to perform highly accurate speech recognition.

  The speech recognition apparatus according to the present invention may further include position information acquisition means for acquiring position information.

  According to said structure, speech recognition is performed using the database matched with the positional information acquired by the positional information acquisition means. Here, the position information acquisition unit may acquire position information indicating the position of the own device, or acquire position information indicating the position of the external device from an external device such as a portable terminal via a communication path. You may do. When the position information indicates the position of the own device, the voice for performing voice recognition is acquired at the own device or in the vicinity of the own device. Further, when the position information indicates the position of the external device, the voice for performing voice recognition is acquired in the vicinity of the external device or the external device.

  As a result, the position information indicates the position of the device to which the voice is input, so that voice recognition can be performed using a database corresponding to the position of the voice input.

  The voice recognition apparatus according to the present invention is a voice / position information receiving means for receiving position information indicating the position of the portable terminal and feature information of voice input to the portable terminal from portable terminals connected by a communication path. And character information transmission means for transmitting voice recognition result information, which is a result of voice recognition by the voice recognition means, to the portable terminal.

  According to said structure, speech recognition is performed using the database matched with the positional information which shows the received position from the portable terminal connected by the communication path. Then, the result of voice recognition is transmitted to the portable terminal.

  Thereby, the voice recognition using the database according to the position of the mobile terminal can be performed.

  In the speech recognition apparatus according to the present invention, when the database selection unit selects a database that uses a database different from the database currently used, a confirmation screen for confirming whether a new database may be selected on the display unit May be displayed.

  According to the above configuration, when the database to be used is changed, the confirmation screen is displayed on the display unit. This allows the user to check before the database is changed.

  In order to solve the above-described problem, a mobile terminal according to the present invention includes a position information acquisition unit that acquires position information indicating the position of its own device, and a feature amount extraction that extracts a feature amount used for speech recognition from input speech. Means, position information acquired by the position information acquisition means, and feature quantity information indicating the feature quantity extracted by the feature quantity extraction means, to the speech recognition apparatus, and by the transmission means Character information receiving means for receiving, from the voice recognition device, character information that is a result of voice recognition performed by the voice recognition device based on the transmitted feature information and position information.

  According to the above configuration, the voice input to the own apparatus and the information indicating the position of the own apparatus are transmitted to the voice recognition apparatus, and the voice recognition is performed using the database corresponding to the position of the own apparatus. Receive the result.

  Thus, voice recognition using a database corresponding to a position can be performed even if the device itself does not include voice recognition means and a database.

  Even the voice recognition system including the voice recognition server and the portable terminal can achieve the effects described above.

  In order to solve the above-described problem, the mobile terminal according to the present invention has an execution function for acquiring execution function information indicating a function being executed by the own device or a function executed most recently in the mobile terminal having a plurality of functions. An information acquisition unit; a database selection unit that selects a database that outputs character information corresponding to the feature amount of the input speech and is associated with the execution function information acquired by the execution function information acquisition unit; And voice recognition means for performing voice recognition using the database selected by the means.

  In addition, in the mobile terminal control method according to the present invention, in a mobile terminal having a plurality of functions, an execution function information acquisition step of acquiring execution function information indicating a function being executed on the own device or a function executed most recently, A database selection step for selecting a database that outputs character information corresponding to the feature amount of the input speech associated with the execution function information acquired in the execution function information acquisition step, and selected in the database selection step And a speech recognition step for performing speech recognition using the prepared database.

  According to the above configuration and method, the own apparatus acquires the execution function information indicating the function that is currently or most recently executed, and performs voice recognition using the database associated with the execution function information.

  As a result, it is possible to perform speech recognition using a database corresponding to the function currently or most recently executed by the own device.

  Thus, for example, when the device itself has a function as a music player and is executing the function as a music player, it can perform speech recognition by selecting a database enriched with music-related words, Higher accuracy speech recognition can be realized.

  The voice recognition device and the portable terminal may be realized by a computer. In this case, the voice recognition device, the portable terminal, and the voice recognition server are operated by the computer by operating the computer as the respective means. A speech recognition device, a portable terminal, a speech recognition server control program to be realized, and a computer-readable recording medium on which the control program is recorded also fall within the scope of the present invention.

  As described above, the speech recognition apparatus according to the present invention includes a database selection unit that selects a database that outputs character information corresponding to a feature amount of an input speech, which is associated with position information indicating a position, And a voice recognition unit that performs voice recognition using the database selected by the database selection unit.

  Further, the control method of the speech recognition apparatus according to the present invention includes a database selection step of selecting a database that outputs character information corresponding to the feature amount of the input speech and associated with the position information indicating the position; A speech recognition step of performing speech recognition using the database selected in the database selection step.

  Thereby, the voice recognition using the database according to the position indicated by the position information can be performed.

  Therefore, when a device (such as a microphone) that receives speech input to the speech recognition device is present at a certain location, a database enriched with words and acoustic / language models that are considered to be frequently used at the location is used. Voice recognition can be performed, and more accurate voice recognition can be realized.

  In addition, the speech recognition apparatus according to the present invention, in a speech recognition apparatus having a plurality of functions, execution function information acquisition means for acquiring execution function information indicating a function being executed by the own apparatus or a function executed most recently, A database selection unit that selects a database that outputs character information corresponding to the feature amount of the input speech, which is associated with the execution function information acquired by the execution function information acquisition unit, and is selected by the database selection unit. And a voice recognition unit that performs voice recognition using a database.

  As a result, it is possible to perform speech recognition using a database corresponding to the function currently or most recently executed by the own device.

  In addition, the mobile terminal according to the present invention transmits the speech information to the speech recognition apparatus, and the mobile terminal that receives the character information indicated by the transmitted speech information from the speech recognition apparatus acquires location information indicating the position. An acquisition unit, a feature amount extraction unit that extracts a feature amount of speech used for speech recognition from input speech, position information acquired by the position information acquisition unit, and a feature amount extracted by the feature amount extraction unit Transmitting means for transmitting feature amount information to the speech recognition device, and information acquisition means for acquiring character information as a result of speech recognition by the speech recognition device based on the feature amount information and position information transmitted by the transmitting means. It is the structure equipped with these.

  Thereby, even if the own apparatus is not equipped with the voice recognition means and the database, the voice recognition using the database corresponding to the position can be performed.

[Embodiment 1]
An embodiment of the present invention will be described below with reference to FIGS. Further, the voice recognition device 1 shown below can be provided in a mobile phone.

  FIG. 1 is a block diagram of a speech recognition apparatus 1 according to the present embodiment. As shown in FIG. 1, the speech recognition apparatus 1 includes a speech input unit 10, a feature amount calculation unit (feature amount calculation unit) 11, a speech recognition unit (speech recognition unit) 12, a display unit 13, a location information acquisition unit (position Information acquisition means) 14, map information storage section 15, GPS (Global Positioning System) 16, GPS antenna 17, use database determination section (database selection means) 18, correspondence table storage section 19, database section 20, and input section 21. It is the composition which includes.

  The voice input unit 10 receives voice input from a microphone or the like, and converts the received voice into voice data. Then, the audio data is transmitted to the feature amount calculation unit 11.

  The feature amount calculation unit 11 calculates a feature amount necessary for voice recognition by the voice recognition unit 12 from the received voice data, and transmits the feature amount information to the voice recognition unit 12 as feature amount information. Examples of feature quantities include MFCC (Mel Frequency Cepstrum Coefficient), LPC (Linear Prediction Coefficient) cepstrum and power, their primary and secondary regression coefficients, and their values are dimensionally compressed by principal component analysis and discriminant analysis. However, the present embodiment is not limited to this.

  The voice recognition unit 12 performs voice recognition from the feature amount information received from the feature amount calculation unit 11 using the database determined by the use database determination unit 18 and transmits the display data to the display unit 13 as display data. For this voice recognition, a conventional technique is used.

  The display unit 13 receives display data from the voice recognition unit 12 and displays the content indicated by the display data. The display unit 13 may be any display device that can display characters and the like, and examples thereof include an LCD (Liquid Crystal Display).

  The location information acquisition unit 14 uses the latitude / longitude information (position information) received from the GPS 16 and the map information stored in the map information storage unit 15 in the map information corresponding to the received latitude / longitude ( Location information (position information) of current position) is acquired. Then, the acquired location information is transmitted to the usage database determination unit 18. Here, the location information is information indicating the region (for example, Kansai region, Tohoku region, etc.) corresponding to the current position when performing speech recognition corresponding to the acoustic / language model, and performing speech recognition corresponding to the facility. Is information indicating the name of a facility corresponding to the current position (for example, a station, an airport restaurant, etc.). The location information can be arbitrarily set as long as the database can be specified. That is, divisions and names on the map can be set freely.

  The map information storage unit 15 stores map information used by the location information acquisition unit 14. Further, the map information includes information about which region the point belongs to and what kind of facility exists at the point.

  The GPS 16 generates latitude / longitude information from radio waves received using the GPS antenna 17. Then, the generated latitude / longitude information is transmitted to the location information acquisition unit 14.

  The use database determination unit 18 determines a database to be used in the speech recognition apparatus 1 from the database unit 20 in accordance with the correspondence table stored in the correspondence table storage unit 19.

  Further, when the user moves from place to place and the database to be used is changed, the use database determination unit 18 displays on the display unit 13 whether or not a new database to be used may be determined. Also good.

  The correspondence table storage unit 19 stores a correspondence table in which the location information and the usage database are associated with each other, which is used when the database used by the speech recognition unit 12 is determined by the usage database determination unit 18. FIG. 2 shows an example of the correspondence table. FIG. 2A shows a correspondence table 25 when performing speech recognition corresponding to an acoustic / language model, and FIG. 2B shows a correspondence table 26 when performing speech recognition corresponding to a facility. It is shown. For example, when the correspondence table 25 shown in FIG. 2A is used and the location information is “Kansai region”, the database to be used is “Kansai dialect model”.

  The database unit 20 stores a database used by the speech recognition unit 12 when performing speech recognition by the speech recognition apparatus 1. The database outputs character information corresponding to the input voice feature amount. The database is divided into an acoustic / language model correspondence database 201 and a facility correspondence database 202. The acoustic / language model correspondence database 201 includes a database AA, a database AB,..., A database AZ, and the facility correspondence database 202 includes a database BA, a database BB,. The database is not limited to this.

  The input unit 21 is a user interface of the voice recognition device 1 and includes various operation keys and accepts various instructions to the voice recognition device 1. Also, whether to perform voice recognition corresponding to the acoustic / language model or voice recognition corresponding to the facility is received, and the received content is transmitted to the use database determination unit 18. In the present embodiment, it is described whether voice recognition corresponding to an acoustic / language model or voice recognition corresponding to a facility is performed, but the present invention is not limited to this.

  In the present embodiment, either one of the voice recognition corresponding to the acoustic / language model or the voice recognition corresponding to the facility is selected, and only one of them is supported. May be implemented together.

  Next, the flow of speech processing in the speech recognition apparatus 1 will be described using FIG. FIG. 3 is a flowchart showing the flow of the voice recognition process.

  First, the voice recognition device 1 receives the start of voice recognition by the user's operation at the input unit 21. Then, whether to perform voice recognition corresponding to the acoustic / language model or voice recognition corresponding to the facility is accepted (S301). Then, the GPS 16 measures the current position of the voice recognition device 1 (S301). Next, the location information acquisition unit 14 acquires location information from the map information stored in the map information storage unit 15 using the latitude / longitude information received from the GPS 16 (S302). Then, the use database determination unit 18 determines a database corresponding to the location information as a database used by the speech recognition apparatus 1 (S304).

  For example, when performing speech recognition corresponding to an acoustic / language model, the use database determination unit 18 refers to the correspondence table 25 in FIG. 2A, and the acquired location information is information indicating “Kansai region”. For example, it is determined as a database that uses the acoustic / language model database of “Kansai dialect”, and if it is information indicating “Tohoku region”, it is determined as a database that uses the acoustic / language model database of “Tohoku dialect”. .

  In addition, when performing speech recognition corresponding to the facility, the use database determination unit 18 refers to the correspondence table 26 in FIG. 2B, and if the acquired location information is “airport”, the travel-related word is If it is “restaurant”, it is determined as a database that uses “meal dictionary” that is rich in meal-related words.

  When the database to be used is determined, the speech recognition apparatus 1 accepts speech input (S305), performs speech recognition (S306), and displays character information as a result of speech recognition on the display unit 13 (S307).

  With the above configuration, the following effects can be obtained. For example, let us consider a case where speech recognition corresponding to an acoustic / language model is performed and a user who possesses a speech recognition device is in the Kansai region. In this case, since the location information to be acquired is “Kansai region”, the speech recognition apparatus 1 selects the database of the acoustic / language model of “Kansai dialect” (FIG. 2A). When the user is in the Kansai region, the user is very likely to speak Kansai dialect. Therefore, it is possible to select a Kansai dialect acoustic / language model database that has a very high possibility of speaking, thereby realizing more accurate speech recognition.

  Also, consider a case where voice recognition corresponding to a facility is performed and a user who has a voice recognition device is at a station. In this case, since the location information to be acquired is “airport / station”, the voice recognition device 1 selects use of the “travel dictionary” (FIG. 2B). And when a user is in a station, there is a high possibility of talking about travel-related topics. Therefore, since it is possible to select a dictionary that is enriched with words related to a topic that is highly likely to be spoken, speech recognition with higher accuracy can be realized.

[Embodiment 2]
The following will describe another embodiment of the present invention with reference to FIGS. For convenience of explanation, members having the same functions as those shown in the first embodiment are given the same reference numerals, and explanation thereof is omitted.

  FIG. 4 is a block diagram of the speech recognition apparatus 2 according to the present embodiment. The speech recognition apparatus 2 differs from the first embodiment in that a use database determination unit (database selection unit) 42, an execution function transmission unit (execution function information acquisition unit) 43, an execution function storage unit 44, and a correspondence table storage unit 45. And a database unit 46. The voice recognition device 2 has a plurality of functions (GPS, music player, etc.) in addition to voice recognition.

  In this embodiment, unlike the first embodiment, a database used for speech recognition is determined in accordance with the function executed by the speech recognition apparatus 2 or the function executed most recently.

  Hereinafter, a characteristic configuration of the speech recognition apparatus 2 will be described in detail.

  The execution function transmission unit 43 determines a function currently being executed in the speech recognition apparatus 2 and transmits execution mechanism information indicating the function to the use database determination unit 42. For example, when the voice recognition device 2 functions as a music player, the content indicating that it functions as a music player is transmitted to the use database determination unit 42. If there is no currently executed function, information (executed function information) indicating the most recently executed function stored in the executed function storage unit 44 is transmitted to the use database determining unit 42.

  The execution function storage unit 44 stores functions executed by the voice recognition device 2.

  The use database determination unit 42 determines a database to be used for speech recognition from the execution function information received from the execution function transmission unit 43 using the correspondence table 51 stored in the correspondence table storage unit 45.

  The correspondence table storage unit 45 stores a correspondence table in which the location information and the use database are associated with each other, which is used when the use database determination unit 42 determines the database used by the speech recognition unit 12. FIG. 5 shows an example of the correspondence table. FIG. 5 shows a correspondence table 51 when performing speech recognition corresponding to the execution function information. For example, when the correspondence table 51 shown in FIG. 5 is used, if the execution function information is “music player”, the database to be used is “music dictionary”.

  The database unit 46 stores a database used by the speech recognition unit 12 when performing speech recognition by the speech recognition apparatus 2. The database outputs character information corresponding to the input voice feature amount. The database 46 includes a database CA, a database CB,... A database CZ. The database is not limited to this.

  Next, the flow of processing in the speech recognition apparatus 2 will be described with reference to FIG. FIG. 6 is a flowchart showing the flow of the speech recognition process.

  First, the voice recognition device 2 receives voice recognition by the user's operation at the input unit 21 (S601). Next, the execution function transmission unit 43 transmits to the use database determination unit 42 execution function information indicating a function executed by the speech recognition apparatus 2 or a function that the speech recognition apparatus 2 has executed most recently (S602). ). Then, the use database determination unit 42 determines a database to be used for speech recognition using the correspondence table 51 from the received execution function information (S603).

  For example, if the received information is “function as GPS”, the travel dictionary is determined as the use database.

  When the speech recognition apparatus 2 determines the database to be used, it accepts speech input (S604), performs speech recognition (S605), and displays character information as a result of speech recognition on the display unit 13 (S606).

  With the above configuration, the speech recognition device 2 can perform speech recognition using a database corresponding to the function currently or most recently executed by the device itself.

[Embodiment 3]
The following will describe still another embodiment of the present invention with reference to FIG. For convenience of explanation, members having the same functions as those shown in the first and second embodiments are given the same reference numerals, and explanation thereof is omitted.

  FIG. 7 shows a block diagram of the speech recognition system 3 according to the present embodiment. As shown in FIG. 3, in this embodiment, voice input and display are performed on the portable terminal 5, and voice recognition is performed on the voice recognition server 4.

  That is, the portable terminal 5 includes a voice input unit 10, a feature amount calculation unit (feature amount calculation unit) 11, a display unit 13, a location information acquisition unit (position information acquisition unit) 14, a map information storage unit 15, a GPS (Global Positioning). In addition to the (System) 16 and the GPS antenna 17, a transmission unit (voice / position information transmission unit) 71 and a reception unit (character information acquisition unit) 72 are provided. In addition to the voice recognition unit (speech recognition unit) 12, the use database determination unit (database selection unit) 18, the correspondence table storage unit 19, and the database unit 20, the voice recognition server 4 includes a transmission / reception unit (character information transmission unit, Voice / position information receiving means) 75.

  As shown in FIG. 7, in the speech recognition system 3, the mobile terminal 5 performs speech input, feature amount calculation, and location information acquisition, and the feature amount information and location information are transmitted to the transmission unit 71, base station 73, and transmitted to the voice recognition server 4 via the network 74. Then, the voice recognition server 4 receives the feature amount information and the location information by the transmission / reception unit 75, and transmits the feature amount information to the voice recognition unit 12 and the location information to the usage database determination unit 18.

  The use database determination unit 18 that has received the location information determines the use database by the method described above and transmits it to the voice recognition unit 12. The voice recognition unit 12 performs voice recognition by the method described above, and transmits the result data to the reception unit 72 of the mobile terminal 5 via the transmission / reception unit 75. The receiving unit 72 that has received the result data transmits the data to the display unit 13, and the display unit 13 displays the character information indicated by the received result data.

  In the present embodiment, the location information is acquired in the mobile terminal 5, but the latitude / longitude information acquired by the GPS 16 is transmitted to the speech recognition server 4, and the location information is acquired in the speech recognition server 4. There may be.

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

  Finally, each block of the speech recognition apparatuses 1 and 2 and the speech recognition system 3, particularly the speech input unit 10, the feature amount calculation unit 11, the speech recognition unit 12, the location information acquisition unit 14, the use database determination unit 18, and the use database determination The unit 42 and the execution function transmission unit 43 may be configured by hardware logic, or may be realized by software using a CPU as follows.

  In other words, the speech recognition apparatuses 1 and 2 and the speech recognition system 3 include a CPU that executes instructions of a control program that realizes each function, a ROM (read only memory) that stores the program, and a RAM (random access) that expands the program. memory), a storage device (recording medium) such as a memory for storing the program and various data. The object of the present invention is to read the program codes (execution format program, intermediate code program, source program) of the control programs of the speech recognition apparatuses 1 and 2 and the speech recognition system 3 which are software for realizing the functions described above by a computer. The recorded recording medium is supplied to the voice recognition apparatuses 1 and 2 and the voice recognition system 3, and the computer (or CPU or MPU (microprocessor unit)) reads and executes the program code recorded on the recording medium. Can also be achieved.

  Examples of the recording medium include a tape system such as a magnetic tape and a cassette tape, a magnetic disk such as a floppy (registered trademark) disk / hard disk, a CD-ROM (compact disc read-only memory) / MO (magneto-optical) / Disc systems including optical disks such as MD (Mini Disc) / DVD (digital video disk) / CD-R (CD Recordable), card systems such as IC cards (including memory cards) / optical cards, or mask ROM / EPROM ( An erasable programmable read-only memory) / EEPROM (electrically erasable and programmable read-only memory) / semiconductor memory system such as a flash ROM can be used.

  Further, the voice recognition devices 1 and 2 and the voice recognition system 3 may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited. For example, the Internet, intranet, extranet, LAN (local area network), ISDN (integrated services digital network), VAN (value-added network), CATV (community antenna television) communication. A network, a virtual private network, a telephone line network, a mobile communication network, a satellite communication network, etc. can be used. In addition, the transmission medium constituting the communication network is not particularly limited. For example, IEEE (institute of electrical and electronic engineers) 1394, USB, power line carrier, cable TV line, telephone line, ADSL (asynchronous digital subscriber loop) line Wireless such as IrDA (infrared data association) and remote control such as remote control, Bluetooth (registered trademark), 802.11 wireless, HDR (high data rate), mobile phone network, satellite line, terrestrial digital network, etc. But it is available. The present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

  Since voice recognition using a database corresponding to a place or situation can be performed, for example, it is suitable for recognition of voice input at various places.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1, showing an embodiment of the present invention, is a block diagram illustrating a main configuration of a voice recognition device. The correspondence table in the said embodiment is shown, (a) shows the correspondence table in the case of performing speech recognition corresponding to an acoustic / language model, and (b) is speech recognition corresponding to a facility. The correspondence table in the case of performing is shown. It is a flowchart which shows the flow of the speech recognition process in the said embodiment. FIG. 11 is a block diagram illustrating a main configuration of a speech recognition apparatus according to another embodiment of the present invention. The correspondence table in said other embodiment is shown. It is a flowchart which shows the flow of the speech recognition process in the said other embodiment. FIG. 24 is a block diagram illustrating still another embodiment of the present invention and illustrating a configuration of a main part of a voice recognition system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 2 Voice recognition apparatus 3 Voice recognition system 4 Voice recognition server 5 Portable terminal 10 Voice input part 11 Feature-value calculation part (Feature-value calculation means)
12 Voice recognition unit (voice recognition means)
13 Display Unit 14 Location Information Acquisition Unit (Position Information Acquisition Unit)
15 Map information storage unit 18, 42 Use database determination unit (database selection means)
19, 45 Corresponding table storage unit 20, 46 Database unit 43 Execution function transmission unit (execution function information acquisition means)
71 Transmitter (voice / position information transmitting means)
72 Receiving part (character information acquisition means)
75 Transmitter / receiver (character information transmitting means, voice / position information receiving means)

Claims (11)

Database selection means for selecting a database that outputs character information corresponding to the feature amount of the voice, associated with the position information indicating the position;
A speech recognition apparatus comprising speech recognition means for performing speech recognition using the database selected by the database selection means.   The speech recognition apparatus according to claim 1, further comprising position information acquisition means for acquiring position information. Voice / position information receiving means for receiving position information indicating the position of the portable terminal and feature amount information of voice input to the portable terminal from portable terminals connected by a communication path;
The speech recognition apparatus according to claim 2, further comprising: character information transmission means for transmitting speech recognition result information, which is a result of speech recognition by the speech recognition means, to the portable terminal.   The database selection means displays a confirmation screen for confirming whether or not a new database can be selected on the display unit when selecting a database that uses a database different from the currently used database. Item 4. The speech recognition device according to any one of Items 1 to 3. Position information acquisition means for acquiring position information indicating the position of the own device;
Feature amount extraction means for extracting feature amounts used for speech recognition from input speech;
Voice / position information transmission means for transmitting the position information acquired by the position information acquisition means and feature quantity information indicating the feature quantity extracted by the feature quantity extraction means to a voice recognition device;
Character information receiving means for receiving, from the voice recognition device, character information that is a result of voice recognition by the voice recognition device based on the feature amount information and position information transmitted by the transmission means. Mobile terminal.   A voice recognition system comprising the voice recognition device according to claim 4 and the portable terminal according to claim 5. In a mobile terminal having multiple functions,
Execution function information acquisition means for acquiring execution function information indicating a function being executed on the own device or a function executed most recently;
Database selection means for selecting a database that outputs character information corresponding to the feature amount of the input speech, associated with the execution function information acquired by the execution function information acquisition means;
A portable terminal comprising voice recognition means for performing voice recognition using the database selected by the database selection means.   A control program for operating any of the speech recognition apparatus according to claim 1 and the portable terminal according to claim 5 or 7, wherein the control program causes a computer to function as each of the above-described means.   A computer-readable recording medium on which the control program according to claim 8 is recorded. A method for controlling a speech recognition apparatus that recognizes speech using a database and outputs character information as a recognition result,
A database selection step for selecting a database that outputs character information corresponding to the feature amount of the voice, which is associated with the position information indicating the position;
And a speech recognition step for performing speech recognition using the database selected in the database selection step. In a mobile terminal having multiple functions,
An execution function information acquisition step for acquiring execution function information indicating a function being executed on the own device or a function executed most recently;
A database selection step for selecting a database that outputs character information corresponding to the feature amount of the input speech, associated with the execution function information acquired in the execution function information acquisition step;
And a voice recognition step of performing voice recognition using the database selected in the database selection step.

Images