JP2013117597A - Broadcast reception system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a "broadcast reception system" for suppressing the generation of a period when the voice recognition of a broadcasting station name of a receivable broadcasting station is not executable.SOLUTION: The broadcast reception system manages data on a broadcasting station normally receivable by a mesh for each mesh, and when the mesh including a current position changes, creates dictionary data for voice recognition of the broadcasting station designation of each broadcasting station from data of broadcasting stations of the mesh including the current position and a mesh adjacent to the mesh, and when the creation of the dictionary data completes, updates a voice recognition dictionary to be used for the voice recognition of the broadcasting station designation with the created dictionary data.

Description

  The present invention relates to an audio input technique for a broadcast receiver.

  As a voice input technique for a broadcast receiver, a broadcast station name represented by a voice uttered by a user is recognized using a voice recognition dictionary storing voice recognition data for each broadcast station name, and the recognized broadcast station name. A technique for switching the reception frequency of a broadcast receiver to the broadcast frequency of a broadcasting station is known (for example, Patent Document 1).

Japanese Patent Laid-Open No. 11-239067

In a broadcast receiver mounted on a moving body or a broadcast receiver that is carried and moved, the broadcast stations that can be received change with the movement of the broadcast receiver. The name cannot be limited to only the broadcast station names of broadcast stations that can be received within one area.
On the other hand, it is necessary to store the voice recognition dictionary that uses the voice recognition dictionary in which the data for voice recognition is registered for all broadcasting station names that may be input by the user, that is, the names of broadcasting stations in the whole country. This increases the storage capacity and increases the processing time required for speech recognition. In addition, when viewed at each time point, since many voice recognition data of broadcast station names of broadcast stations that cannot be received at that time point are registered in the voice recognition dictionary, it is inefficient and voice recognition accuracy is high. It is easy to cause deterioration.

  Therefore, for each area, a broadcast station that can be received in the area is registered, and the broadcast station name of the broadcast station registered for the area including the current position is changed according to the movement between the areas of the current position. It is conceivable to create a speech recognition dictionary storing speech recognition data and perform speech recognition of the broadcast station name spoken by the user using the created speech recognition dictionary.

  However, in this case, after moving from the area where the current position is to another area, a voice recognition dictionary that stores voice recognition data of broadcast station names of broadcast stations that can be received in the other area can be created. Since it starts, voice recognition of broadcast station names of broadcast stations that can be received in the other area can be performed for a while after moving to the other area until the completion of creation of the voice recognition dictionary. The problem of disappearing arises.

  Therefore, the present invention creates a speech recognition dictionary that stores speech recognition data of a broadcasting station name of a broadcasting station corresponding to an area including the current position, and performs a speech recognition for speech input of the broadcasting station name. In the receiving system, an object is to suppress the occurrence of a period in which voice recognition of a broadcast station name of a receivable broadcast station cannot be performed at each time point.

  In order to achieve the above object, the present invention provides a broadcast receiving system for recognizing a broadcast station name spoken by a user and switching a broadcast station to receive a broadcast station with a recognized speech station name. A voice recognition dictionary in which voice recognition dictionary data is registered, a voice recognition means for recognizing the broadcast station name spoken by the user using the voice recognition dictionary, a current position calculating means for calculating a current position, A storage unit that stores broadcast station data for each mesh in which a broadcast station name of a broadcast station that can be received at a position in the mesh is provided corresponding to each mesh that is a geographical area of a predetermined size; When the mesh including the current position calculated by the position calculation unit changes, the mesh including the current position and the meshes around the mesh including the current position The speech recognition dictionary data for speech recognition of the broadcast station name registered in the per-mesh broadcast station data corresponding to is created, and the speech recognition dictionary registered in the speech recognition dictionary with the created speech recognition dictionary data Voice recognition dictionary updating means for updating the voice recognition dictionary by replacing data.

  According to such a broadcast reception system, immediately after the current position moves from a position in a certain mesh to a position in another mesh around the mesh, the movement to the position in the other mesh is performed. Even before the voice recognition dictionary is updated, the broadcast station recognition dictionary stores voice recognition dictionary data for voice recognition of broadcast station names of broadcast stations that can be normally received in the other meshes. Therefore, the broadcast station name can be recognized without any trouble both before and after the update of the voice recognition dictionary. Therefore, it is possible to suppress the occurrence of a period during which voice recognition of the broadcast station names of receivable broadcast stations cannot be performed at each time point.

Here, in such a broadcast receiving system, the meshes around the mesh including the current position may be all meshes adjacent to the mesh including the current position.
Moreover, the above broadcast receiving system may be a broadcast receiving system mounted on an automobile.

  As described above, according to the present invention, a voice recognition dictionary storing voice recognition data of a broadcasting station name of a broadcasting station corresponding to an area including the current position is created, and the name of the broadcasting station name inputted by voice is created. When performing speech recognition, it is possible to suppress the occurrence of a period during which speech recognition of a broadcast station name of a receivable broadcast station cannot be performed at each time point.

It is a block diagram which shows the structure of the vehicle-mounted system which concerns on embodiment of this invention. It is a figure which shows the map data and broadcast station recognition dictionary which concern on embodiment of this invention. It is a figure which shows the mesh of the map data which concerns on embodiment of this invention. It is a flowchart which shows the broadcast station recognition dictionary creation process which concerns on embodiment of this invention. It is a figure which shows the process example of the broadcast station recognition dictionary creation process which concerns on embodiment of this invention.

Hereinafter, embodiments of the present invention will be described.
FIG. 1 shows a configuration of an in-vehicle system according to the present embodiment.
The in-vehicle system is a system mounted on an automobile. As shown in the figure, a map data storage unit 1 that is a storage device such as an HDD that stores map data, a tuner 2 that receives radio broadcasts, a navigation unit 3, and a user An audio output device 4 having an audio amplifier and a speaker, an operation unit 5 for receiving user operation input, a microphone 6, a voice recognition unit 7, a user interface control unit 8, a broadcast station recognition dictionary 9, and a broadcast station A dictionary creation unit 10 for creating and updating the recognition dictionary 9 and a display device 11 are provided.
Here, the voice recognition unit 7 recognizes the user's voice input from the microphone 6, and the broadcast station recognition dictionary 9 is a voice recognition dictionary used by the voice recognition unit 7 for voice recognition of the broadcast station name. . The user interface control unit 8 controls a user interface using the voice output device 4, the display device 11, the operation unit 5, and the voice recognition unit 7.

  However, the in-vehicle system may be configured using a CPU circuit having a general configuration including a microprocessor, a memory, and other peripheral devices in terms of hardware. Each unit such as the navigation unit 3, the voice recognition unit 7, the user interface control unit 8, and the dictionary creation unit 10 is realized as a process that is realized by a microprocessor executing a program prepared in advance. good. In this case, such a program may be provided to the in-vehicle system via a recording medium or an appropriate communication path.

Next, FIG. 2a shows a configuration of map data stored in the map storage unit.
Here, in the map data, as shown in FIG. 3, an area obtained by dividing the geographical range (usually the whole country) of the map represented by the map data into fixed sizes (rectangles separated by dotted lines in the figure). Are managed for each mesh.
That is, as shown in FIG. 2A, the map data includes mesh management data and mesh data provided for each mesh. The mesh management data describes the correspondence between each mesh data and the geographical range of the mesh corresponding to the mesh data.

Each mesh data includes in-mesh map data representing a map in the corresponding mesh and in-mesh broadcast station data.
On the other hand, in-mesh broadcast station data has a broadcast station record provided corresponding to each broadcast station that can be normally received in the corresponding mesh, and each broadcast station record includes the name of the corresponding broadcast station. (For example, “NHK Fukushima”) and the broadcasting station name Yomi representing the reading of the broadcasting station name of the corresponding broadcasting station (for example, “Enueichi Keifukushima”) and the broadcasting of the corresponding broadcasting station The frequency is registered.

Next, FIG. 2 b shows the configuration of the broadcast station recognition dictionary 9.
As shown in the figure, the broadcast station recognition dictionary 9 has dictionary data (each line in the figure) provided corresponding to each broadcast station name to be subjected to speech recognition. The station name, the broadcast station name reading indicating the corresponding broadcast station name reading, the sound pattern that reads the corresponding broadcast station name or the sound pattern data indicating the feature pattern of the sound, and the corresponding broadcast station name The broadcast frequency of the broadcasting station is registered.

  Now, in such a configuration, the navigation unit 3 uses the map data stored in the map data storage unit 1 to calculate the current position of the host vehicle and the current traveling direction of the host vehicle, and the user from the user. Accepting destination setting via the interface control unit 8, searching for a route to the set destination, a mark indicating the current position and current traveling direction on the map around the current position, and a route to the destination A process of outputting a navigation screen representing the above to the display device 11 via the user interface control unit 8 is performed. The navigation unit 3 includes a GPS receiver, a gyro sensor, a vehicle speed sensor, and the like for calculating the current position.

  On the other hand, the voice recognition unit 7 uses the broadcast station recognition dictionary 9 as the voice recognition dictionary for the user's voice input from the microphone 6 and performs voice recognition processing of the broadcast station name represented by the user's voice. The tuner 2 is instructed via the user interface control unit 8 to switch the reception frequency to the frequency registered in the broadcast station recognition dictionary 9 as the broadcast station frequency of the broadcast station name. Then, the tuner 2 switches the reception frequency to the instructed frequency. As a result, when the user speaks the broadcast station name, the broadcast station received by the tuner 2 is switched to the broadcast of the broadcast station name.

  Note that the speech recognition of the broadcast station name performed in the speech recognition unit 7 is performed by pattern matching between a speech pattern input from the microphone 6 and speech pattern data registered in the broadcast station recognition dictionary 9, for example. That is, the user selects the broadcast station name corresponding to the sound pattern data that most closely matches the sound pattern input from the microphone 6 out of the sound pattern data that matches the sound pattern input from the microphone 6 by a predetermined level or more. This is done by recognizing the name of the broadcasting station that spoke.

Next, creation of the broadcast station recognition dictionary 9 used for speech recognition of the above broadcast station names will be described.
The dictionary creation unit 10 executes a broadcast station recognition dictionary creation process shown in FIG. 4 in order to create the broadcast station recognition dictionary 9.
As illustrated, in this process, the dictionary creating unit 10 first sets an area including a mesh including the current position calculated by the navigation unit 3 and a mesh adjacent to the mesh as a target area ( Step 402).
That is, here, as shown in FIG. 5a, if the current position 500 is inside the mesh j, the area including the mesh j and the meshes e, f, g, i, k, m, n, and o adjacent to the mesh j. 501 is set as the target area.
Then, in-mesh broadcast station data of mesh data of each mesh included in the target area is acquired (step 404), and data of the broadcast station recognition dictionary 9 is created (step 406).
Here, in step 406, from the broadcast station records included in each acquired in-mesh broadcast station data, those in which the content of the broadcast station record overlaps with other broadcast station records are excluded. That is, when there are a plurality of broadcast station records having the same contents, only one of the plurality of broadcast station records remains, and other broadcast station records are excluded.

  Then, dictionary data corresponding to each remaining broadcast station record is created. The dictionary data corresponding to the broadcast station record is created by creating the voice pattern data indicating the pattern of the voice that reads the broadcast station name from the broadcast station name Yomi of the broadcast station record and the pattern of the characteristics of the voice. Then, data including the broadcast station name, the broadcast station name reading, and the broadcast frequency of the broadcast station record registered in the broadcast station record is created as dictionary data corresponding to the broadcast station record.

  Next, the execution of the speech recognition process using the broadcast station recognition dictionary 9 of the speech recognition unit 7 is temporarily stopped (step 408), and the dictionary data of the broadcast station recognition dictionary 9 is created in each dictionary created in step 406. The broadcast station recognition dictionary 9 is updated by replacing the data (step 410). Then, the execution of the voice recognition process using the broadcast station recognition dictionary 9 of the voice recognition unit 7 is resumed (step 412).

Further, thereafter, the change in the mesh including the current position calculated by the navigation unit 3 is monitored (step 414), and if the change occurs, the processing from step 402 is performed.
As a result, for example, as shown in FIG. 5a, when the current position 500 is in mesh j, area 501 including mesh j and meshes e, f, g, i, k, m, n, and o adjacent to mesh j. Speech recognition is performed using the broadcast station recognition dictionary 9 in which dictionary data of broadcast station names of broadcast stations that can be normally received is stored, and then the current position 500 is moved to the mesh f as shown in FIG. In this case, the dictionary data of the broadcast station names of the broadcast stations that can be normally received in the area 502 including the mesh f and the meshes a, b, c, e, g, i, j, and k adjacent to the mesh f are stored. Voice recognition using the stored broadcast station recognition dictionary 9 is performed. And then, as shown in FIG. 5c, when the current position 500 moves to mesh g, mesh g and meshes b, c, d, f, h, j, k, Speech recognition is performed using the broadcast station recognition dictionary 9 in which dictionary data of broadcast station names of broadcast stations that can be normally received in the area 503 including l is stored.

  Therefore, for example, as shown in FIGS. 5a to 5b, immediately after the current position is moved from the position in the mesh j to the position in the mesh f, the broadcast station recognition dictionary 9 is moved with the movement to the position in the mesh f. Even before is updated, the broadcast station recognition dictionary 9 stores dictionary data of broadcast station names of broadcast stations that can be normally received in the mesh f. Recognizing it, the receiving frequency of the tuner 2 can be switched to the broadcasting frequency of the broadcasting station with the name of the broadcasting station.

Therefore, according to the present embodiment, it is possible to suppress the occurrence of a period during which voice recognition of the broadcast station names of receivable broadcast stations cannot be performed at each time point.
The broadcast station recognition dictionary creation process performed by the dictionary creation unit 10 has been described above.
The above embodiment can be similarly applied to a broadcast receiver mounted on a moving body other than an automobile such as a bicycle, a broadcast receiving system that is carried by a user, and the like.
The same applies to the case where the tuner 2 receives TV broadcasts.
In the above embodiment, the broadcast station recognition dictionary 9 is updated with the target area including the mesh including the current position and the mesh adjacent to the mesh. The broadcast station recognition dictionary 9 is updated as an area including the mesh including the mesh and the mesh adjacent to the traveling direction side of the vehicle calculated in the navigation unit 3 among the meshes adjacent to the mesh including the current position. You may make it.

  DESCRIPTION OF SYMBOLS 1 ... Map data storage part, 2 ... Tuner, 3 ... Navigation part, 4 ... Voice output device, 5 ... Operation part, 6 ... Microphone, 7 ... Voice recognition part, 8 ... User interface control part, 9 ... Broadcast station recognition dictionary 10: Dictionary creation unit.

Claims (4)

A broadcast receiving system for recognizing a broadcast station name spoken by a user and switching a broadcast station to receive the broadcast station name of the recognized broadcast station name,
A speech recognition dictionary in which speech recognition dictionary data for speech recognition of broadcast station names is registered;
Using the voice recognition dictionary, voice recognition means for voice recognition of the broadcast station name spoken by the user;
Current position calculating means for calculating the current position;
A storage unit that stores broadcast station data for each mesh in which broadcast station names of broadcast stations that can be received at positions in the mesh are provided corresponding to each mesh that is a geographical area of a predetermined size;
When the mesh containing the current position calculated by the current position calculating means changes, it is registered in the per-mesh broadcast station data corresponding to the mesh including the current position and the meshes around the mesh including the current position. The speech recognition dictionary data for speech recognition of the broadcast station name being created is created, and the speech recognition dictionary data registered in the speech recognition dictionary is replaced with the created speech recognition dictionary data. And a speech recognition dictionary updating means for updating the broadcast receiving system. The broadcast receiving system according to claim 1,
The broadcast receiving system according to claim 1, wherein the meshes around the mesh including the current position are all meshes adjacent to the mesh including the current position. The broadcast receiving system according to claim 1,
The broadcast receiving system is a broadcast receiving system mounted on an automobile. A speech recognition dictionary in which speech recognition dictionary data for speech recognition of a broadcast station name is registered; speech recognition means for recognizing a broadcast station name spoken by a user using the speech recognition dictionary; and a geographical area of a predetermined size Storage means for storing broadcast station data for each mesh in which the broadcast station names of broadcast stations that can be received at positions in the mesh are provided corresponding to each mesh, and the speech recognition means A speech recognition dictionary update method for updating the speech recognition dictionary in a broadcast receiving system for switching a broadcast station to receive a broadcast station with a speech-recognized broadcast station name,
The broadcast receiving system calculates a current position;
When the mesh including the calculated current position changes, the broadcast receiving system registers in the mesh-specific broadcast station data corresponding to the mesh including the current position and the meshes around the mesh including the current position. The speech recognition dictionary data for speech recognition of the broadcast station name being created is created, and the speech recognition dictionary data registered in the speech recognition dictionary is replaced with the created speech recognition dictionary data. Updating the voice recognition dictionary.