JP2001249686A5 - - Google Patents

Description

0013
BEST MODE FOR CARRYING OUT THE INVENTION [Embodiments of the Invention]
The invention according to claim 1 of the present invention is a method of performing voice recognition based on an acoustic model and a recognition target vocabulary dictionary for input voice, and is set by a user's movement information or a user. There is a method of sequentially updating the speech recognition target vocabulary dictionary according to the information and performing voice recognition using the updated recognition target vocabulary dictionary, and by performing voice recognition using the recognition target vocabulary dictionary depending on the situation. , Has the effect of improving the recognition rate.

0014.
In the invention described in claim 2 of the present invention, the setting information set by the user is based on the navigation destination set by the user, the route information to the current location and the destination, or the user. The voice recognition method according to claim 1, which is characterized in that it is at least one of the set category information of the vocabulary to be recognized, and is a vocabulary that the user will generate in the future when updating the voice recognition target vocabulary dictionary. It has an action as selection information of.

0015.
The invention according to claim 3 of the present invention is the voice recognition method according to claim 1 or 2 , wherein the movement information of the user uses at least one of the position information, the movement speed, and the movement direction of the user. When updating the recognition target vocabulary dictionary, it has an effect of being able to construct a recognition target vocabulary dictionary according to the moving state of the user.

0016.
In the invention according to claim 4 of the present invention, the timing for sequentially updating the recognition target vocabulary is when a predetermined time has elapsed since the last update of the recognition target vocabulary dictionary, or when the route deviates from the planned route. The voice recognition method according to claim 3, which is at least one of the cases where the average movement speed changes and the setting information changes, has an effect of designating the update timing of the voice recognition dictionary.

[0017]
The invention according to claim 5 of the present invention is the voice recognition method according to claim 3 or 4, wherein the movement information of the user is information obtained from a satellite positioning system (GPS). , It has the effect of being able to be used as update information of the speech recognition target vocabulary dictionary by reflecting the situation of the user.

[0019]
The invention according to claim 6 of the present invention is a device that performs voice recognition based on an acoustic model and a recognition target vocabulary dictionary with respect to the input voice, and is based on the movement information of the user or the user. The recognition target vocabulary dictionary for sequential voice recognition is updated according to the setting information, and the input voice is recognized using the updated recognition target vocabulary dictionary, and the input voice is recognized efficiently and dynamically. It has the effect of improving the recognition rate by updating the target vocabulary dictionary and performing voice recognition.

0020
The invention according to claim 7 of the present invention is efficient because the category information gives clues such as position information when selecting the recognition target vocabulary according to the setting information by the user and the moving state from the speech recognition dictionary database. It has the effect of realizing the update of the recognition target vocabulary.

0021.
In the invention according to claim 8 of the present invention, the setting information set by the user is based on the navigation destination set by the user, the route information to the current location and the destination, or the user. It is assumed that it is at least one of the set category information, and has an effect as vocabulary selection information that the user will generate in the future in updating the speech recognition target vocabulary dictionary.

[0022]
The invention according to claim 9 of the present invention is to use at least one of the user's position information, movement speed, and movement direction as the movement information of the user, and updates the recognition target vocabulary dictionary. It has the effect of being able to grasp the moving state of the user.

0024
According to the tenth aspect of the present invention, the recognition target vocabulary dictionary is updated in the previous recognition target when the user's setting information is changed, when the user deviates from the planned route, or when the average movement speed changes. The voice recognition device according to any one of claims 6 to 9, which is at least one when a predetermined time has elapsed after updating the vocabulary dictionary, and the timing of updating the voice recognition dictionary is set. It has the effect of designating.
The invention according to claim 11 of the present invention is any one of claims 6 to 10, wherein the information acquisition means extracts the movement information of the user from the information obtained from the satellite positioning system (GPS). It is a voice recognition device, and has an effect that it can be used as update information of a voice recognition target vocabulary dictionary by reflecting the situation of the user.

0025.
The invention according to claim 12 of the present invention is a navigation device using the voice recognition device according to any one of claims 6 to 11, and can be used by voice without remembering the complicated functions of the device. It requires a function of ease of use, and further, since it constructs a recognition target vocabulary dictionary suitable for the situation of the user in updating the recognition target vocabulary dictionary, it has an effect of reducing voice recognition errors.

0026
The invention according to claim 13 of the present invention is a method of selecting a recognition dictionary used for voice recognition by the navigation device according to claim 12 , and is a method of selecting a recognition dictionary, which is a movement speed of a user, a position of a user, or a setting by a user. By updating the recognition target vocabulary dictionary with priority using at least one method of information, time, and category information of each vocabulary, it has an effect that the recognition target dictionary can be used efficiently.

[0027]
The invention according to claim 14 of the present invention is characterized in that a method by voice recognition can be used for setting setting information by a user, and navigation without the user remembering complicated operations of the device. It has the effect of being able to use the device.

0028.
The invention according to claim 15 of the present invention is a navigation device that displays a selection range of a vocabulary to be recognized on a display unit, and has an effect of visually providing a user with a state of the system.
The invention according to claim 16 of the present invention is a recognition target having vocabulary information to be voice-recognized from a pre-registered voice-recognition target vocabulary dictionary database according to the movement information of the user or the setting information by the user. It includes a voice recognition method in which the vocabulary dictionary is sequentially updated and the input voice is recognized by using the updated recognition target vocabulary dictionary based on the acoustic model and the recognition target vocabulary dictionary. It is a program to be executed by a computer, and the recognition target vocabulary dictionary for sequential speech recognition is updated according to the movement information of the user or the setting information by the user, and the updated recognition target vocabulary dictionary is used. Therefore, the input voice is recognized, and the recognition rate is improved by efficiently and dynamically updating the recognition target vocabulary dictionary to perform voice recognition.

Claims (16)

A method of performing speech recognition on input speech based on a recognition target vocabulary dictionary having an acoustic model and vocabulary information to be a target of speech recognition , which is a user's movement information or setting information by the user voice response, update sequentially the recognition target vocabulary dictionary from pre-registered speech recognition target vocabulary dictionary database, using a recognition target vocabulary dictionary that said updating, and said recognition to Rukoto the input speech Recognition method. Setting information that is set by a Subscriber, on the basis of the destination of navigation that has been set by the user, the current location and route information or category information of the vocabulary to be set recognition target by the user, to the destination The speech recognition method according to claim 1, wherein the speech recognition method is at least one of the following. Movement information of the user, location information of the user, the moving speed, that take advantage of at least one moving direction 請 Motomeko 1 or 2 speech recognition method as claimed. Timing for sequentially updating the recognition target words, when the time determined in advance after performing the updating of the recognition target words dictionary last time has elapsed, when the scheduled route or Rights prolapse, when the average moving speed is changed, sets speech recognition method according to at least one der Ru請 Motomeko 3 when information changes. Movement information of a Subscriber is a satellite positioning system: according to claim 3 or 4 speech recognition method, wherein the information obtained from (Global Positioning System GPS). For the input speech, an apparatus for performing speech recognition based on the acoustic model and recognition target vocabulary dictionary, obtain setting information by the user of the mobile information or user information acquisition unit, the information acquiring means a dictionary updating unit that sequentially updates the recognition target words dictionary using the moving information or set by the user information obtained user from using a recognition target vocabulary dictionary containing the update, sure the input speech speech recognition apparatus comprising a recognition unit you identify. 7. The speech recognition apparatus according to claim 6 , wherein the plurality of vocabularies registered in the speech recognition target vocabulary dictionary database are stored together with category names representing the features and position information thereof. The setting information set by the user of the information acquisition means is at least one of route information to the current location and the destination based on the navigation destination set by the user, or category information set by the user. der Ru speech recognition apparatus 請 Motomeko 6 or 7, wherein. Movement information of the user, location information of the user, the moving speed, the speech recognition apparatus according to any one of 請 Motomeko 6-8 you use at least one direction of movement. Updating of the recognition target words dictionary, when the setting information of a Subscriber has changed, when the de-scheduled route or rights, if the average moving speed is changed, before once performed the update of the recognition target vocabulary dictionary predetermined speech recognition apparatus according any one of 9 請 Motomeko 6 to Ru at least one der when the elapsed time since. Information obtaining means, satellite positioning system: speech recognition apparatus according to any one of 請 Motomeko 6 to 10 you extracts movement information of the user from the information obtained from the (Global Positioning System GPS). Navigation apparatus using the speech recognition apparatus of any serial mounting of claims 6 to 1 1. Recognition target vocabulary dictionary is, when the moving speed is slow, the user of the current position 置近 side of the recognition subject to vocabulary, when the moving speed is high, and the recognition target in the direction of movement and the destination neighborhood comprising vocabulary, or, if there is by that setting information to the user, using at least one vocabulary of said setting information, vocabulary to be recognized in accordance with the category information and the time available for each vocabulary The navigation device according to claim 12 , wherein the navigation device is a dictionary configured as follows . Set by the user information, the navigation apparatus 請 Motomeko 12 or 13, wherein you get by speech recognition. Further comprising claim 13 or 14 navigation device according to the display unit having a function of displaying a target range of the speech recognition. The recognition target vocabulary dictionary having vocabulary information to be the target of speech recognition is sequentially updated from the speech recognition target vocabulary dictionary database registered in advance according to the movement information of the user or the setting information by the user, and the updating is performed A program for causing a computer to execute a speech recognition method for performing speech recognition on input speech using an acoustic model and the recognition target vocabulary dictionary, using a recognition target vocabulary dictionary.